D
Dance
▶ Dancing: A Nonverbal Language for Imagining and
Learning

Dance Learning
▶ Neurophysiological Correlates of Learning to Dance

Dancing: A Nonverbal
Language for Imagining and
Learning
JUDITH LYNNE HANNA
University of Maryland, College Park, MD, USA

Synonyms
Dance; Kinesthetic communication; Performing art

Definition
Dance is human behavior composed of purposeful,
intentionally rhythmical, and culturally influenced
sequences of communicative nonverbal body movement and stillness in time, space, and with effort.
Dance stylizes movements, some from everyday life,
with a degree of conventionality or distinctive imaginative symbolization. Each dance genre has its own
▶ aesthetic (standards of appropriateness and
competency).

Theoretical Background
Dance can engender visions of alternative possibilities
in culture, politics, and the environment. Moreover,
dance can also foster creative problem-solving and the

acquisition, reinforcement, and assessment of
nondance knowledge, emotional involvement, social
awareness, and self and group identity (Hanna 2008).
Dance is captivating nonverbal communication
that involves attention networks, motivation, and
reward. Nonverbal communication includes the bodily
conveyance of information through gesture and locomotion, proximity, touch, gaze, facial expression, posture, physical appearance, smell, and emotion.
Evolutionary biologists note that humans must attend
to motion for survival – to distinguish prey and predator, to select a mate, and to anticipate another’s
actions and respond accordingly. Humans first learn
through movement, and movement facilitates learning.
Sensory-motor activities form new neural pathways
and synaptic connections throughout life, and the
merger of body, emotion, and cognition in dance may
lead to effective communication, the medium of learning. While humans alone among species have art experiences without obvious evolutionary payoff, dance
engages innate “play” brain modules that allow us to
consider hypothetical situations so that we can form
plans in advance of difficulties, symbolically confront
current problems, and manage stress.
Speech refers to the oral/auditory medium that we
use to convey the sounds associated with human languages. Language, however, is the method of conveying
complex concepts and ideas, representations of information and relationships, and a set of rules for how
these may be combined and manipulated with or without recourse to sound (Clegg 2004:8). Multiple possible
“languages of thought” play different roles in the life of
the mind but nonetheless work together (pp. 1, 200).
Dance is a language that bears some similarities
to verbal language. For example, both have vocabulary
(locomotion and gestures in dance) and grammar
(rules for different dance traditions in putting the
vocabulary together and justifying how one movement
can follow another). And both languages have semantics (meaning). Verbal language strings together

N. Seel (ed.), Encyclopedia of the Sciences of Learning, DOI 10.1007/978-1-4419-1428-6,
# Springer Science+Business Media, LLC 2012

906

D

Dancing: A Nonverbal Language for Imagining and Learning

sequences of words, and dance strings together
sequences of movement. However, dance more often
resembles poetry, with its multiple, symbolic, and elusive meanings, than prose. Complex logical structures
are more easily communicated in verbal language.
Although spoken language can simply be meaningless
sounds, and movements can be mere motion, listeners
and viewers tend to read meaning into what they hear
and see. The vocal, auditory, and static visual channels
predominate in verbal language, whereas the motor,
moving visual, and kinesthetic channels predominate
in dance. Verbal language exists solely in a temporal
dimension, whereas dance involves the temporal plus
three dimensions in space.
Multisensory, dance heightens a perceptual awareness that expands access to clues to the meaning of
different emotions, a significant source of human motivation. Emotion may prime some goals and processes
while inhibiting others. In dance there is the sight of
dancers moving in time and space; the sound of physical movement, breathing, accompanying music and
talk; the smell of dancers’ physical exertion; the tactile
sensation of body parts touching the ground, other
body parts, people or props, and the air around the
dancers; the proxemic sense of distance between or
among dancers and dancers and spectators; and the
kinesthetic experience (a performer feeling the moving
sensual body and a spectator empathizing with the
performer’s bodily movement and energy). The eyes
indicate degrees of attentiveness and arousal, influence
attitude change, and regulate interaction. In addition,
the eyes define power and status relationships. Effective
communication, of course, depends upon the shared
knowledge between dancer and audience.
A dance learner may feel stressed by “not getting it”
or receiving negative feedback from teachers and others.
Performance anxiety affects novice and pro alike. On the
other hand, mastery of dance makes one feel satisfied,
confident, and proud. Performance can give a feeling of
the “runner’s high.” Individuals usually find strength in
the self-mastery required in learning to dance and feel
supported by others in cohesive group dancing. Performers feel accomplishment as they express the sense
of doing something and being in control, of achieving
what others want to do, try to do, but cannot do well,
and the exhilaration of performance. Of course, dance is
art and entertainment that diverts performers and audiences alike from stressors (Hanna 2006).

While dancers and their audiences can sense the feel
and command of the human body in dance, the mind
stirs the imagination, directs movement, and makes
sense of feeling. Feeling a particular emotion,
a performer may immediately manifest it through
dance, and dancing may induce emotion through energetic physical activity or interaction between or among
dancers, or dancers and spectators. Alternatively, during a performance, a dancer may recall an emotion
from earlier personal experience and use the memory
as a stimulus to express the emotion symbolically,
creating an illusion of the emotion rather than feeling
its actual presence (Hanna 1983).
A dancer’s purpose may be to provide an emotional
experience, to conceptualize through movement, or to
play with movement itself. In telling stories through
dance, troublesome themes, like fear, can be held up to
scrutiny, played with, distanced, made less threatening,
and even move people to social action.
Dance can convey meaning through devices and
spheres. The most common device for encoding meaning is through metaphor, the expression of one thought,
experience, or phenomenon in place of another that it
resembles. A common sphere in which devices operate
to convey meaning is the whole pattern of performance,
emphasizing structure, style, feeling, or drama.
Dancing involves declarative knowledge (including
concepts, history, movement vocabulary, and grammar) that may be visualized in choreography. Procedural knowledge, knowing-is-in-the-doing, or
embodied knowledge, is attained through multiple sensory perception, especially kinesthesia. This knowledge
incorporates motor skills and “muscle memory” (proprioception felt in the body), cognitive skills and strategies that enable the application of patterns in
communicating ideas and feelings in dance. People
learn by doing – action being the test of comprehension, and imagination the result of the mind blending
the old and familiar to make it new in experience. Tacit
knowledge is knowledge that cannot be articulated
verbally but may be expressed kinesthetically and emotionally through dance.
A difference between declarative knowledge and
procedural knowledge is that each likely activates different parts of the brain. Someone can know about
a dance form and yet not have the skills for performance. After a cognitive stage in which a description of
a procedure is learned, skill learning has an associative

Dancing: A Nonverbal Language for Imagining and Learning

stage in which a method for performing the skill is
worked out, and finally an autonomous stage in
which the skill becomes automatic.
The creativity of making set dances and improvising within a style requires declarative and procedural
knowledge (usually tacit) of relational rules for
matching movements with appropriate meanings.
These rules emphasize digital, analytical, and sequential processing of information. The underlying processes are hidden, rapid, multimodal, nonlinear, and
nonverbal, and the dance evolves from experimentation and exploration in the medium itself. Dancemaking involves composing movement phrases and
eventually long sequences, evaluation, changing, reevaluating, deleting, and adding.
By contrast, dance imitation, or dancing someone
else’s choreography, depends on learning a set pattern
involving analogical and spatial abilities. Observation,
inferring the mental representations that underlie
dance, and storing the representations in memory are
required. Imitation is not strict copying but
a constructed version, an interpretation through emotional expression of what is imitated. Improvisation
refers to extemporaneously creating dance out of
what is known.
Understanding dance requires reasoning, an
understanding of symbols, the ability to analyze
images, and knowing how to organize knowledge.
Given a student’s familiarity with dance elements,
dance may be a means of testing a student’s understanding of nondance subject material. Translating
emotions and ideas from one medium to another in
a different context, such as thinking metaphorically
through a physical embodiment of written or spoken
text requires an understanding of subject matter. This
creative process can reveal knowledge acquired and
what further instruction is necessary.

Important Scientific Research and
Open Questions
Dance education in the schools has been demonstrated
to be an engaging, emotional, and cognitive way of
solving problems as it communicates declarative and
procedural knowledge through various devices and
spheres of embodying the imagination (Hanna 1999).
Qualitative research in Africa found that specific
dancers perform contrastive movement patterns to
identify their distinct biological and social roles.

D

Among the Ubakala Igbo in Nigeria, when women are
life-giving mothers, and men are life-taking warriors,
women dance slowly and effortlessly in circles, whereas
men dance rapidly and forcefully in angular lines
(Hanna 1987).
Evidence of the potency of nonverbal communication comes from psychologists Goldin-Meadow (2002)
and her colleagues who focus on hand gesture, merely
one of the dancer’s communicative body parts: When
produced beside speech, gesture becomes image and
analog. However, when called upon to carry the full
burden of communication, gesture takes a languagelike form using word and sentence level structure.
Extrapolating from hand gestures to dance promises
an exponential impact: Dance utilizes a multichanneled system of gestures of various body parts
moving in time, space, and with effort, music, and
costume.
Findings about learning a second or third verbal
language seem applicable to learning a nonverbal language such as a dance genre, and to even learning more
than one kind of dance. Youngsters who grow up multilingual have more brain plasticity and multitask more
easily. Moreover, learning and knowing a second or
third language may use parts of the brain that knowing
only one’s mother tongue may not. Dancers who only
have classical ballet training often have difficulty picking up contemporary movements.
Through magnetic resonance imaging technology,
neuroscientists are discovering regions of the brain
dedicated to perceiving, reacting, remembering, thinking, creating, and judging (Grove et al. 2005). Neuroscientists discovered the communicative potential of
dance: Areas in the brain that control the hands and
gestures overlap and develop together with the areas
that control the mouth and speech. The Broca and
Wernicke areas, located in the left hemisphere, are
associated with verbal language expression and comprehension, abstract symbolic and analytic functions,
sequential information processing, and complex patterns of movement. The process of dance-making
engages some of the same components in the brain
for conceptualization, creativity, and memory as does
verbal poetry or prose, but obviously not the same
procedural knowledge. Dance is also linked to the
right hemisphere that involves elementary perceptual
tasks, nonverbal processing of spatial information,
music, and emotional reactivity. However, rigid

907

D

908

D

Data Mining

lateralization of brain function is precluded by the
transfer of inputs to each side of the brain over the
corpus callosum, the main body of nerve fibers
connecting the two hemispheres. A study of the neural
basis of the tango (using MRI and position emission
tomography) found an interacting network of brain
areas active during the performance of motor sequencing and movement intention. Dance influences the
mind causing positive plastic changes in the brain,
reorganizing neural pathways, or the way the brain
functions, for young and old alike. Physical activity
sparks biological changes that encourage brain cells to
bind to one another which reflect the brain’s fundamental ability to adapt to challenges. In complex motor
movement, the brain fires signals along the network of
cognitive functioning cells, which solidifies their connections. Extended learning in dance thus impacts how
well the brain processes other tasks. Dance has the
cognitive demands of remembering steps and executing them usually in response to music and coordination with a partner or group in space, and creating
dances.
Mirror neurons in the brain are active in someone
carrying out a particular dance movement as well as in
someone else who watches the same movement.
Greater bilateral activations occur when expert dancers
viewed movements that they had been trained to perform compared to other movements. The simulation
process of resonance between observed and embodied
action could underpin sophisticated mental functions
of empathy, sympathetic kinesthesia, and understanding in social interaction.
Triangulating knowledge from the arts and humanities, social and behavioral sciences, and cognitive/neurological science, elucidates the power of dance in
imaging and learning. Findings related to the other
arts may apply to dance because it often occurs in
combination with music, written text, poetry, and the
visual arts of set design and costume.

Cross-References
▶ Aesthetic Learning
▶ Approaches to Learning
▶ Creativity, Problem-Solving and Feeling
▶ Dance Learning: Neurological Correlates of Complex Action
▶ Joyful Learning
▶ Learning Activity

▶ Motor Learning
▶ Play and Learning
▶ Play, Exploration, and Learning
▶ Stress Management

References
Clegg, M. (2004). Evolution of language: Modern approaches to the
evolution of speech and language. General Anthropology, 10(2),
1–11.
Goldin-Meadow, S. (2002). Constructing communication by hand.
Cognitive Development, 17, 1385–1406.
Grove, R., Stevens, C., & McKechnie, S. (Eds.). (2005). Thinking in
four dimensions: Creativity and cognition in contemporary dance.
Carlton: Melbourne University Press.
Hanna, J. L. (1983). The Performer-Audience Connection: Emotion to
Metaphor in Dance and Society. Austin: University of Texas Press.
Hanna, J. L. (1987). To Dance Is Human: A of Nonverbal Communication (Rev. of 1979). Chicago: University of Chicago Press.
Hanna, J. L. (1999). Partnering dance and education: Intelligent moves
for changing times. Champaign: Human Kinetics.
Hanna, J. L. (2006). Dancing for health: Conquering and preventing
stress. Lanham: AltaMira.
Hanna, J. L. (2008). A nonverbal language for imagining and learning: Dance education in K-12 curriculum. Educational
Researcher, 37(8), 491–506.

Data Mining
▶ Learning Algorithms

Dealing with Uncertainty
▶ Complex Problem Solving

Deciphering
▶ Reading and Learning

Decision
A choice between two or more alternatives. A central
question in the social sciences, and in particular economics, is whether humans make decisions that are

Deductive Learning

optimal (global rationality) or whether they have to
aim only for decisions that are “good-enough”
(bounded rationality).

Decision Learning
▶ Rapid Response Learning in Amnesia

Declarative
▶ Prospective and Retrospective Learning in Mild
Alzheimer’s Disease

Declarative Knowledge
Declarative knowledge is knowledge of facts, concepts,
events, and objects. It can typically be directly
expressed as a proposition, image, or relational
structure. This information is stored in long-term
memory and organized into schemas that interconnect
to shape comprehension and influence semantic
interpretations.

D

Declarative Pointing
▶ Joint Attention in Humans and Animals

D

Declarative Showing
▶ Joint Attention in Humans and Animals

Decoding
▶ Reading and Learning

Deductive Learning
TRISTAN CAZENAVE
LAMSADE, Université Paris-Dauphine, Paris,
Cedex 16, France

Synonyms
Chunking; Explanation based generalization; Explanation based learning

Definition

Declarative Learning
▶ Fact Learning
▶ Meaningful Learning in Economic Games

Declarative Memory
Declarative memory contains memory for facts such as
5 þ 2 = 7 and water expands when it freezes.

Cross-References
▶ Explicit and Procedural-Learning-Based Systems of
Perceptual Category Learning

909

We can formally define three types of reasoning mechanisms: deduction, induction, and abduction. Let us
define the rule: “if we are in the morning then the sun
rises,” with the facts “we are in the morning” and “the sun
rises.” Deduction consists in asserting the fact “the sun
rises” given the rule “if we are in the morning then the
sun rises” and the fact “we are in the morning.” Induction
consists in asserting the rule “if we are in the morning
then the sun rises” from the facts “we are in the morning”
and “the sun rises.” Abduction consists in guessing the
fact “we are in the morning” from the fact “the sun rises”
and the rule “if we are in the morning then the sun rises.”
A large number of works in the field of machine
learning deal with induction. In this entry we are not
dealing with inductive learning but rather with learning from deduction.

910

D

Deductive Learning

Deduction is an important area of Artificial Intelligence. Many Artificial Intelligence systems rely on
deduction to solve problems. This is particularly true
for systems written in Prolog. Prolog enables to write
systems using first order logic. It provides a built-in
deduction mechanism, and a program can be written
declaratively as a set of facts and rules. Deduction is
also used in expert systems that also rely on fact and
rules to solve problems. Backward chaining consists in
starting with a goal expression that the system tries to
deduce and to find the rules that conclude on this goal
expression; these rules contain sets of conditions that
have to be true in order to prove the goal expression
and these conditions are set as the new goal expressions
that the system tries to deduce. On the contrary, forward chaining consists in starting from a set of facts
and to deduce new facts applying the rules that match
the set of facts (i.e., a rule matches a set of facts if all its
conditions match some facts). The main difference
between Prolog and usual expert systems is that Prolog
uses backward chaining while expert systems usually
use forward chaining.
Using first order logic to represent rules means that
the conditions of the rule can contain variables and not
only facts. The example rule of the beginning of the
entry about the sun that rises in the morning does not
contain variables. A similar rule that may contain a
variable could be: “if the variable Hour is greater than
9 and lower than 12 then the sun rises.” The goal of
deductive learning is to speed up systems that use
deduction in first order logic to solve problems.
Deductive learning (Laird et al. 1986; Mitchell et al.
1986; DeJong and Mooney 1986) consists in creating
rules that deduce rapidly some facts that have taken
time to be deduced by a rule-based system. If the system
fires these learned rules directly, he or she may find
solutions faster than the original system. The goal of
deductive learning is to speed up problem solving. One
possible drawback of learning rules this way is that the
system can be slower after learning than before learning. This undesired behavior can be caused by the time
used to match the learned rules (Minton 1990). To
avoid this, an operationality criterion (DeJong and
Mooney 1986) can be used to decide which learned
rules are kept and which ones are discarded.
Deductive learning is particularly efficient in games
that have threats (Cazenave 1998). Knowing the rules of
a game, a system can find the set of facts that enable one

player to win if he or she plays first in a particular
position. It can then generate a rule that deduces the
game can be won in similar cases where the rule matches.
Moreover, since the system knows all the conditions of
a rule that enable one player to win, he or she can analyze
these conditions and find a small subset of the possible
moves for the other player that invalidate one of the
conditions. This small subset of moves contains all the
relevant moves to prevent the first player to win, it is also
established that no other move can prevent the first
player to win. Therefore using these rules in a problem
solver for games reduces a lot the branching factor of the
games and enables to solve problems that would be
unsolvable with a brute force approach.

Theoretical Background
The correctness of deductive learning comes from the
equivalence between firing rules with variables and
proving theorems in first-order logic. A learned rule is
a generalization of a set of facts that have been deduced.
The generalization consists in replacing in the conditions of the learned rules the binded variables with true
variables. The first-order learned rule is a theorem in
the theory formed by the initial set of rules. Therefore,
learned rules are consistent with the theory and deduce
correct facts provided the initial theory is also correct.

Important Scientific Research and
Open Questions
Deductive learning can be used to find rules to direct
the search of a problem solver. It is also related to metaknowledge and to machine consciousness since the
system has to observe and analyze its own behavior in
order to create new knowledge (Pitrat 2009).

Cross-References
▶ Deductive Reasoning and Learning
▶ Inferential Learning and Reasoning
▶ Metacognition and Learning
▶ Problem Solving

References
Cazenave, T. (1998). Metaprogramming Forced Moves. In ECAI
1998, Brighton, 645–649
DeJong, G., & Mooney, R. J. (1986). Explanation-based learning: An
alternative view. Machine Learning, 1(2), 145–176.
Laird, J. E., Rosenbloom, P. S., & Newell, A. (1986). Chunking in soar:
The anatomy of a general learning mechanism. Machine Learning, 1, 11–46.

Deductive Reasoning and Learning
Minton, S. (1990). Quantitative results concerning the utility of
explanation-based learning. Artificial Intelligence, 42(2–3),
363–391.
Mitchell, T. M., Keller, R. M., & Kedar-Cabelli, S. T. (1986). Explanation-based generalization: A unifying view. Machine Learning, 1,
47–80.
Pitrat, J. (2009). Artificial beings: The conscience of a conscious
machine. London/Hoboken: ISTE/Wiley. ISBN 978-1848211018.

Deductive Reasoning and
Learning
MICHAL AYALON, RUHAMA EVEN
Department of Science Teaching, The Weizmann
Institute of Science, Rehovot, Israel

Synonyms
Logical thinking; Logic-based reasoning

Definition
According to commonly accepted notions, deductive
reasoning is the process of inferring conclusions from
known information (premises) based on formal logic
rules, where conclusions are necessarily derived from the
given information and there is no need to validate them
by experiments. Deductive reasoning can be contrasted
with inductive reasoning, in which premises provide
probable, not necessary, evidence for conclusions.

Theoretical Background
There are several forms of valid deductive argument,
for example, modus ponens (If p then q; p; therefore q)
and modus tolens (If p then q; not q; therefore not p).
Valid deductive arguments preserve truth, in the sense
that if the premises are true, then the conclusion is also
true. However, the truth (or falsehood) of a conclusion
or premises does not imply that an argument is valid
(or invalid). In addition, the premises and the conclusion of a valid argument may all be false.
Deductive reasoning is used both for constructing
arguments and for evaluating arguments. Thus, for example, based on the following information: (1) if a car can be
started, then the battery is fine, and (2) the car can be
started, then using the modus ponens rule, it is possible to
conclude that (3) the battery is fine. However, if the car

D

cannot be started, and someone claims that the problem
lies with the battery, using deductive reasoning to evaluate
this claim implies that it is impossible to infer this logically
from the given information. Thus, based on deductive
reasoning, one cannot know whether the problem lies
with the battery or not.
Since the early days of Greek philosophical and
scientific work, deductive reasoning has been considered as a high (and even the highest) form of human
reasoning. Already Aristotle, who laid down the foundations for this kind of thinking in the fourth century,
BC, perceived a person who possesses deductive ability
as being able to comprehend the Universe in more
profound and comprehensive ways. Throughout scientific development, great scientists, such as Descartes
and Popper, emphasized the importance of this kind
of reasoning to science. Although the scientific process
is based to a large extent on inductive reasoning –
developing hypotheses based on empirical observations
to describe “truths” or “facts” about our world –
deduction still plays an important role in science in
criticizing and refuting theories, using the modus
tolens rule. Deductive reasoning is also viewed as
important in technological work and in legal systems,
as well as in facilitating wise decision making related to
fields such as politics and economy.
A discipline which is most identified with deductive
reasoning is mathematics. Deductive reasoning is key
to work in mathematics, because rigorous logical proof,
which is a unique fundamental characteristic of mathematics, is constructed using deductive reasoning.
Although there are some other accepted forms of mathematical validation, deductive proof is considered as
the preferred tool in the mathematics community for
verifying mathematical statements and showing their
universality.

Important Scientific Research and
Open Questions
One line of research in the area of deductive reasoning
focuses on studying its nature and mechanism. Several
psychological theories exist nowadays, providing
explanations for people’s reasoning processes on
diverse tasks that involve the use of deductive reasoning, including accounts of the factors that influence
reasoning on such tasks (e.g., factors that affect typical
errors people make). An example for such a theory is
the mental models theory (Johnson-Laird and

911

D

912

D

Deductive Reasoning and Learning

Byrne 1991) which approaches deductive reasoning as
based on manipulations of mental models representing
situations.
Another line of research deals with the question of
the extent to which deductive reasoning is useful in
daily life. Reports of studies on argumentation from the
last decades have challenged the usefulness of deductive
reasoning in everyday situations (e.g., Toulmin 1969),
claiming that rationality, in the sense of taking the
“best” choice out of a set of options, stands in the
base of reasoning and communication in everyday
activities, and that rationality is not bound to formal
logic. Moreover, in an attempt to convince others of the
rationality of their claims and choices, people often use
various kinds of argument, which mostly do not have
the logical rigidity of deductions. Thus, other, lessstrict kinds of inferences, more of the plausible type
(e.g., inductive, abductive) are commonly used by people in everyday situations.
Still, another debated issue that has long been under
discussion relates to the evolution of the ability to
reason deductively. Some researchers, such as Piaget
(Inhelder and Piaget 1958), claimed that deductive
reasoning develops naturally and formal teaching has
no significant influence on it. Other researchers, such
as Cheng et al. (1986), maintained that teaching plays
an essential condition for the development of deductive
reasoning. Different opinions also exist regarding the
age at which the capability of deductive reasoning can
be developed or learnt. Some researchers claim that the
ability to think in a logical manner develops or may be
learnt at adolscence. Other researchers claim that children, even at the age of preschool and elementary
school, are able to use some deductive arguments.
The literature that deals with the teaching of deductive reasoning suggests that while in the past there was
commonly accepted view that general deduction rules
can be taught by formal teaching methods, in the last
decades this view became controversial. Until the
beginning of the twentieth century, classical education,
especially Latin and mathematics, has been believed to
be an effective tool for teaching deductive reasoning.
More than 2,000 years ago, Plato claimed that geometry
was the best tool for training deductive reasoning, and
the teaching of deductive reasoning became an integral
part of formal mathematics learning. At the beginning
of the twentieth century, the common opinion of stakeholders in mathematics education was that the study of

geometry could enhance deductive reasoning which
would be transferred into reasoning capacity in other
domains. An experiment conducted in geometry in the
1930s (Fawcett 1938) supported this view, suggesting an
improvement of the participating students’ deductive
reasoning. During the twentieth century, however, studies on the transfer of learning have challenged the
assumption that learning subjects such as Latin and
mathematics has an impact on acquiring general skills
of reasoning. These studies stated that transfer of learning
is not general but specific to the situations in which
learning has occurred. Thus, the goal of teaching deductive reasoning that is transferable to other domains has
been undermined, and sometimes even completely
pushed aside. Nonetheless, in the practical field, the
school environment is still often thought to be the place
where mental capabilities relating to deductive reasoning
should be developed (e.g., Ayalon and Even 2010).
Still, researchers who adopt the evolutionary psychology point of view question both the natural development of deductive reasoning and the possibility to
teach it. They contend that a conflict exists between
formal deductive reasoning and natural thinking. They
suggest that people do not naturally think in logical
terms, but instead reason using logics that are different
from the formal one.

Cross-References
▶ Argumentation and Learning
▶ Deductive Learning
▶ Development and Learning
▶ Inferential Learning and Reasoning

References
Ayalon, M., & Even, R. (2010). Mathematics educators’ views on
mathematics learning and the development of deductive reasoning. International Journal of Science and Mathematics Education,
8, 1131–1154.
Cheng, P. W., Holyoak, K. J., Nisbett, R. E., & Oliver, L. (1986).
Pragmatic versus syntactic approach to training deductive reasoning. Cognitive Psychology, 18, 293–328.
Fawcett, H. P. (1938). The nature of proof. The thirteenth yearbook of
the National Council of Teachers of Mathematics. New York:
Teachers College.
Inhelder, B., & Piaget, J. (1958). The growth of logical thinking from
childhood to adolescence. New York: Basic Books.
Johnson-Laird, P. N., & Byrne, R. M. J. (1991). Deduction. Hillsdale:
Erlbaum.
Toulmin, S. E. (1969). The uses of arguments. Cambridge, UK:
Cambridge University Press.

Deep Approaches to Learning in Higher Education

Deductive Schemas
▶ Model-Based Reasoning
▶ Pragmatic Reasoning Schemas

Deed
▶ Spatial Cognition in Action (SCA)

Deep Approaches to Learning
▶ Deep Approaches to Learning in Higher Education

Deep Approaches to Learning
in Higher Education
MICHAEL JACKSON
Department of Government and International
Relations, University of Sydney, Sydney, NSW,
Australia

Synonyms
Conceptual change; Conceptual growth; Deep
approaches to learning; Open-ended learning; Reproductive learning; SOLO taxonomy; Surface approaches
to learning; Transformational learning

Definition
A deep approach to learning concentrates on the meaning of what is learned. That concentration may involve
testing the material against general knowledge, everyday experience, and knowledge from other fields or
courses. A student taking a deep approach seeks principles to organize information. In contrast a student
using a surface approach tries to capture material in
total, rather than understand it. An example is the
student who busily copies down a diagram without
listening to the explanation of it. The emphasis is on
the sign rather than the significance.

D

913

Theoretical Background
For many years the assumption in higher education was
that there is a one-to-one relationship between what an
instructor teaches (says) and what a student learns
(hears). The only challenge to learning was to get students
into the classroom and get them to listen. To meet these
needs, effective teaching in higher education was understood to consist of a catalog of tricks and techniques to
stimulate attendance (pop quizzes, theatrical presentations, roll calling) and to command attention (calling on
silent students, assigning grades to class participation,
and so on). A model of education often parodied as the
transference of the professor’s lecture notes to the student’s notepad. It was assumed that while teachers were
trained to teach in primary and secondary schools, university professors had no need for such training. In the
1980s this complacency was challenged by a number of
researchers who empirically examined students and
teachers in higher education. They found that the transference of knowledge from teacher to student was uncertain. Two students matched for ability, attending the
same class might have quite different reactions and experiences. Why is that? Because these students might process the experience and the material in different ways.
One student might scrupulously try to capture every
word, but not grasp the themes in the material. This
student would be able to recite vast slabs of material,
but when asked to sum up its meaning, value, or implications, be at a loss (Schneps 1988). Another student
might appear detached in the class room, sitting back,
but think about the material in various contexts and as
a result grasp some of the topography of the material and
evaluate it, though be less able to regurgitate bodies of
material, still be able to assess and apply it. These differences were found in studies in Sweden, Scotland,
England, Hong Kong, and Australia. The research
method at the start was simply to interview students
after class sessions and ask them to describe in some
detail what they did in the class. That is, what they did
or thought as the class proceeded. The focus was not on
what the teacher did, though that was part of the story,
but what the student did in response or parallel with what
the teacher did. When the results of these interviews were
compared with similar interviews with teachers many
discrepancies appeared. Material that teachers presented
to stimulate thought was taken by some students as the
final word to be memorized. There was a lack of communication and a mismatch of intentions.

D

914

D

Deep Approaches to Learning in Higher Education

Important Scientific Research and
Open Questions
This body of research emerged among groups of students who met the same entrance standard at the same
institution, taking the same program of study. The
point is that the approaches to learning is not
a psychological state. None of us are born to take
deep approaches, nor to take surface approaches.
These two approaches are learned through schooling,
and for much schooling a surface approach may be
adequate. Nor does taking a deep approach guarantee
success. The efforts of some to take deep approaches to
real number theory may not suffice. For those, a surface
approach might be the best tactic to make the best of
the experience. Equally, rote learning is a tool for some
purposes, like learning the Latin names of bones, or the
conjugations of irregular verbs. But a student will only
come to a deeper understanding of material by taking
deep approaches. If the purpose of higher education is
to impart to students not just bodies of facts but principles that organize and assess knowledge, then deep
approaches have a place.
Deep approaches to learning can be fostered in
a variety of ways. First and foremost is repeatedly to
articulate the nature of value of such approaches, and
to support and encourage them in visible and practical
ways that students value. In this way institutions and
instructors can authorize students to test concepts,
theories, principles, findings, evidence, and argument
they find in the university study against their general
knowledge, common sense, prior learning, Internet
sources, and the like. Rather than forbidding students
to access Wikipedia, fostering a deep approach would
accept that it will happen, and counsel students to
evaluate and test what they find there.
More specifically, course objectives, assignments,
teaching methods, and workload can open the door to
deep approaches for students. Course objectives refer to
the purpose of studying the material. What has the
learner gained by studying this subject? In a “History
of Political Theory” course, the students will learn
about Plato, let us say, but the objective in so doing is
to learn to evaluate arguments, to grapple with wisdom
from a different world, or to test ideas against current
knowledge. The objectives are what is gained when the
course is studied, not when it is graded. If teachers want
students to be motivated by more than grades, then
what that more is has to be explicitly said.

That grades motivate students is a truism, but
teachers can focus that motivation constructively.
One empirical study found that “the majority of students reported greater use of transformational [deep]
activities for an open-ended assessment [assignment]
than for the closed examinations; and conversely less
use of reproductive [surface] activities with the openended assignments than with the short answer and
closed examinations” (Bain and Thomas 1984). An
essay, a laboratory report these give students more
scope to interpret material than true-false tests.
Teaching methods underwrite objectives and
assignments. Entwistle and Tait (1990) interviewed
undergraduate students from more than 60 departments, and found that departments with assignments
that placed a premium on factual information and gave
students less freedom (and its twin, responsibility) led
students to the surface approach to these assignments.
In addition, feedback on assignments is another crucial
element associated with the approach to learning taken
by students in these departments. If the feedback
focuses on compliance and facts, the surface approach
remained, as it did if there was no feedback apart from
the grade.
Workload is another crucial determinate of student
intention. If the work associated with a course seems
excessive, then many students will adopt a surface
approach to survive. We all do this when faced with
more than we can handle. A professor who presents
students with an 80-page syllabus, guarantees that
many will immediately take a surface approach to
cope. The irony is that some of those professors who
are most serious about teaching adopt tactics that discourage students from taking deep approaches. To
assign weekly graded assignments all but insures that
few students will do their best work on any one of them.
One implication of the discussion to this point is
that it might be more effective to manage students’
perception of the learning environment than to concentrate on special study skills sessions, essay writing
workshops, yet more PowerPoint slides, more selfpaced web material, and the like. There are few
technical solutions to human problems. But students’
“perceptions of teaching and assessment methods
[assignments] in academic departments are significantly associated with . . . students approaches to
studying” (Entwistle and Ramsden 1983). To encourage and support students to take deep approaches

Default Reasoning

clarify objectives, set assignments that permit deep
approaches, use teaching methods that place responsibility on students, and manage workload to challenge
but not discourage students. These are among the
practical implications of this body of knowledge.

Cross-References
▶ Abilities to Learn
▶ Academic Motivation
▶ Design of Learning Environments
▶ Intentional Learning
▶ Learning Objectives
▶ Motivation and Learning: Modern Theories
▶ Perceptions of the Learning Context and Learning
Outcomes References
▶ Problem-Based Learning
▶ Rote Memorization
▶ Styles of Learning and Thinking

D

Deep Learning Approaches
They encompass approaches to learning that involve
considerable rumination on the part of the student.
Deep learning will often include the application of critical thinking skills to devise a solution to a posed problem. Deep approaches are contrasted with techniques
that encourage more superficial and often less durable
learning, such as the rote memorization of facts.

Default Reasoning
GERHARD BREWKA
Computer Science Department, Institut für
Informatik, University of Leipzig, Leipzig, Germany

References
Bain, J., & Thomas, P. (1984). Contextual dependence of learning
approaches. Human learning, 3(4), 230–242.
Biggs, J. B. (2003). Teaching for quality learning at university: What the
student does (2nd ed.). Phildelphia: Society for Research into
Higher Education.
Entwistle, N., & Tait, H. (1990). Approaches to learning, evaluations
of teaching, and preferences for contrasting academic environments. Higher education, 19(3), 169–199.
Harvard-Smithsonian Center for Astrophysics. (M. Schneps, producer). (1994). A private universe: Misconceptions that block
learning [DVD]. Further information at http://www.learner.org/
resources/series28.html
Marton, F. (Ed.). (1984). The experience of learning. Edinburgh:
Scottish Academic Press.
Prosser, M., & Trigwell, K. (1999). Understanding learning and teaching: The experience in higher education. Buckingham: Society for
Research into Higher Education.
Ramsden, P. (1992). Learning to teach in higher education. London:
Routledge.

Deep Hierarchical Networks
▶ Hierarchical Network Models for Memory and
Learning

Deep Learning
▶ Learning Hierarchies of Sparse Features

915

Synonyms
Reasoning with exceptions

Definition
Default reasoning is a form of nonmonotonic reasoning where plausible conclusions are inferred based on
general rules which may have exceptions (defaults). It is
nonmonotonic in the sense that additional information may force us to withdraw earlier conclusions,
namely whenever the additional information shows
that the case at hand is exceptional.

Theoretical Background
In classical logic, adding information in the form of
additional premises never invalidates any conclusions.
Commonsense reasoning is different. We often draw
plausible conclusions based on general rules expressing
what normally is the case together with the assumption
that the world about which we reason is normal and as
expected. This is the best we can do in situations in which
we have only incomplete information. However, it can
happen that our normality assumptions turn out to be
wrong. New information can show that the situation
actually is abnormal in some respect. In this case, we
may have to give up some of our former conclusions.
For this reason, the need for logical models of
default reasoning was recognized early in Artificial
Intelligence. Interest in default reasoning was also

D

916

D

Default Reasoning

fueled by the frame problem: How to represent adequately what does not change when an action occurs.
One would like to have action formalisms where it is
sufficient to describe the changes caused by an action.
Ideally, the persistence of all other properties would be
left implicit. This can be achieved by using a default
rule like: What holds in a situation normally holds in the
situation after performing an action. The frame problem
has provided a major impetus to research in default
reasoning.

Important Scientific Research and
Open Questions
One of the most important formalizations of default
reasoning is Ray Reiter’s default logic (Reiter 1980).
In default logic, one has to specify a set W of classical
(propositional or first order) formulas which represent
what is known to be the case. In addition, default rules
are represented in the form
P : Q1 ; . . . Qn = R:
Here P is a premise which must hold for the rule
to be applicable, R is the conclusion. In addition,
Q1, . . .Qn are formulas which must be consistent with
what is derived (or, equivalently, their negations must
not be derived). For instance, the default birds normally
fly can be represented as
BirdðxÞ : FliesðxÞ=Flies ðxÞ:
Thus, in contrast to standard inference rules,
Reiter’s default rules allow us to refer in their preconditions not only to what is derived, but also to what is
not derived. This has drastic consequences. In particular, there is no longer a single set of theorems as in
classical logic. Rather, conflicting default rules may give
rise to multiple acceptable sets of formulas, called
extensions by Reiter.
To qualify as an extension, a set of formulas E must
(1) be closed under classical inference, (2) all applicable
defaults must have been applied, and (3) if E contains
a formula, then this formula must have a noncircular
derivation from W together with applicable default rules.
Based on the extensions one can then define a skeptical
inference relation (F is inferred if it is contained in all
extensions) or a credulous inference relation (F is
inferred if it is contained in some extension).
Important alternative formalizations of default
reasoning are circumscription (McCarthy 1980) and

autoepistemic logic (Moore 1985). The former is
based on the observation that some logical models
represent more normal situations than others. McCarthy defines a logical consequence relation which takes
into account the most normal models only. Defaults are
represented as implications with an explicit abnormality predicate, for example
BirdðxÞ^:AbðxÞ ! FliesðxÞ:
The most normal models then are those which
minimize the abnormal objects.
Autoepistemic logic represents consistency of
a formula through a modal operator. As in default
logic, conflicting default rules may give rise to different
acceptable sets of formulas. For a detailed overview of
these and other approaches, see (Brewka et al. 2007).
One of the key problems with formalizations of
default reasoning is their computational complexity.
Consistency checking is notoriously expensive. It
cannot be done locally and involves the whole knowledge base. The first-order variant of default logic is
not even semi-decidable. For this reason, it is important to identify special cases which are expressive
enough to be useful in practice, yet restricted enough
to allow for efficient computation. With respect to
implementations of default reasoning, there now exist
highly efficient systems for logic programming under
answer set semantics. Logic programming can be
viewed as a restricted version of default logic where
the formulas in the rules are propositional atoms.
Answer set solvers for logic programming, like the
clasp system of Potsdam University (see http://www.
cs.uni-potsdam.de/clasp/), can handle restricted forms
of default reasoning in a highly efficient manner.

Cross-References
▶ Abductive Reasoning
▶ Analogical Reasoning
▶ Cognitive Robotics
▶ Deductive Learning
▶ Deductive Reasoning and Learning
▶ Knowledge Representation
▶ Model-Based Reasoning
▶ Schema-Based Reasoning

References
Brewka, G., Niemelä, I., & Truszczynski, M. (2007). Nonmonotonic
reasoning. In V. Lifschitz, B. Porter, & F. van Harmelen (Eds.),

Deliberate Practice and Its Role in Expertise Development
Handbook of knowledge representation (pp. 239–284). Burlington:
Elsevier.
McCarthy, J. (1980). Circumscription – a form of non-monotonic
reasoning. Artificial Intelligence, 13(1–2), 27–39.
Moore, R. C. (1985). Semantical considerations on nonmonotonic
logic. Artificial Intelligence, 25(1), 75–94.
Reiter, R. (1980). A logic for default reasoning. Artificial Intelligence,
13, 81–132.

D

Deliberate Practice
Framework proposing that the efficient acquisition of
domain-related knowledge and skills, and thus exceptional performance, is primarily due to goal-directed,
effortful, and inherently not enjoyable practice.

Cross-References
▶ Volition for Learning

Defaults
Defaults are specifications of schemas and parts of
a schema that are assumed unless there is specific
information to the contrary. Defaults define the prototypical instantiation of a schema.

Deliberate Practice and Its Role
in Expertise Development
FERNAND GOBET
Department of Psychology, Brunel University,
Uxbridge, Middlesex, UK

Definition Schema
In computer programming, a “definition schema” is the
organization or structure for a database. More specifically,
the activity of data modeling leads to a schema. The term
is used in discussing both relational databases and objectoriented databases. Two common types of database
schemas are the star schema and the snowflake schema.

Deformation
▶ Memory Dynamics

Degradation
▶ Memory Dynamics

Deixis
Something where the meaning has to be understood
from context. Example: A pointing gesture changes
meaning based on what is being pointed at.

917

Synonyms
Practice; Training

Definition
According to the deliberate-practice framework, the way
to reach high levels of expertise is to carry out practice
that is consciously intended to improve one’s skills.
▶ Deliberate practice involves goal-directed activities,
which tend to be repetitive and to enable rapid feedback. Preferably performed individually, these activities tend to be effortful and not enjoyable. They can be
carried out just for a few hours a day (but not so often
that they become inefficient or even hurtful). This
framework limits the role of inherited factors to motivation, general activity levels, and height in sports. No
role is given to talent with respect to cognitive abilities.
The deliberate-practice framework has been influential
in the field of expertise research, and a large number of
studies have been conducted to understand the role of
practice in areas such as sports, games, the arts, and the
professions. There have also been controversies surrounding this framework.

Theoretical Background
The importance of practice has been recognized for
decades, first by proponents of behaviorism and then
by psychologists more interested in cognitive

D

918

D

Deliberate Practice and Its Role in Expertise Development

mechanisms (e.g., De Groot 1946/1978). Practice was
given particular emphasis in Simon and Chase’s (1973)
classic study of expertise in chess, which concluded
that grandmaster level can only be obtained after
about 10 years of dedicated practice and study. This
corresponds to between 10,000 and 100,000 h of
hard work. The deliberate-practice framework
(Ericsson et al. 1993) has taken this position to its
extreme by proposing that innate individual differences
do not limit the top levels of performance, but that
these can be further increased by dedicated practice.
The deliberate-practice framework rejects innate talent
as an explanation for cognitive abilities, arguing that
the evidence for it is flimsy at best. Rather, it proposes
that expert performance is a monotonic function of
the amount of practice. Thus, this framework takes the
clear and extreme position that deliberate practice is
not only a necessary, but also a sufficient condition
for expertise.
Deliberate practice consists of training activities.
The goal is to improve performance by optimizing
feedback and thus the correction of errors. These activities are typically effortful and not enjoyable. Thus, it is
not sufficient to play the piano just for enjoyment, even
if one devotes a considerable number of hours doing so.
It is crucial to use training techniques whose deliberate
goal is to improve one’s performance. Moreover, these
training activities can be carried out for just a few
hours a day. Excessive practice increases the risk of
injuries and burnout (especially in sports). Another
important prerequisite is the presence of a favorable
environment, and in particular strong family support.
The proponents of deliberate practice do acknowledge
the involvement of inherited factors, but these are
limited to motivation, general activity levels, and, in
some sports, height. Importantly, the involvement of
genetic factors is explicitly excluded for explaining
individual differences in high levels of cognitive
abilities. Finally, emphasis is given to individual
practice rather than group practice, as the former
increases the efficiency of the activities characteristic
of deliberate practice.
One of the clearest evidence for the role of deliberate practice (and concomitantly, the lack of involvement of talent) was provided by longitudinal
experiments where college students with average memory span were trained in the digit-span task. When they
devoted sufficiently effort and practice, these students

could perform better than individuals that were
thought in the literature to enjoy special, inherited
talent. The role of deliberate practice has also received
support from many domains, including games, music,
science, and medicine (Ericsson et al. 2006). A large
amount of data has also been collected in sports such as
karate, soccer, hockey, skating, and wrestling. In these
studies, participants are typically asked to estimate
retrospectively how many hours they have spent in
diverse types of activities, and the results are correlated
with their skill level. The results typically show that the
higher skilled individuals engage more in deliberate
practice.
However, some studies, while partly supporting
the role of deliberate practice, also suggest the importance of other factors. A good example of this is the
study carried out by Gobet and Campitelli (2007) with
chess players. (The game of chess has the great advantage that there is an official, reliable, and quantitative
measure of skill – the Elo rating.) They asked players
to estimate the number of hours devoted in their career
to a number of activities, including studying chess
alone and practicing with others (this comprised
playing competitive games). As predicted by deliberate
practice, a strong correlation was found between chess
skill and the number of hours of individual practice.
However, contrary to prediction, an even stronger
correlation was found between chess skill and the
number of hours of group practice. Another result
inconsistent with the deliberate-practice framework
was the high level of variability. Some players became
masters with relatively few hours of deliberate practice
(as low as about 3,000 h), while others needed considerably more time (up to about 24,000 h) – that is a 1:8
ratio. Finally, some players devoted more than 25,000
h to chess study and practice, but failed to become
masters. This study also uncovered two results that
suggest that other factors in addition to deliberate
practice are in play. First, there was a correlation
between final skill level and starting age, in that the
players starting younger were more likely to become
masters. This correlation held even after the contribution of deliberate practice was controlled for. Second,
the proportion of mixed-handedness was higher with
chess players than in the general population. These
results tend to suggest that practice is a necessary,
but not sufficient condition for reaching high levels
of expertise.

Delinquency and Learning Disabilities

Important Scientific Research and
Open Questions
The deliberate-practice framework has generated much
research but also much controversy (e.g., Sternberg
1996). For example, there is considerable evidence
from the fields of personality and intelligence that
there exist large individual differences between individuals, and it is plausible that at least some of these
differences might affect the acquisition of high levels
of expertise. Similarly, individual differences exist with
respect to learning, attention, and working memory. In
addition, the research on deliberate practice is mostly
correlational and rarely uses control groups (i.e., individuals that tried but failed to be experts), and it is thus
difficult to draw conclusions about the causal role of
talent and (deliberate) practice. For example, it could
be the case that, following self-selection, more gifted
individuals remain in the domain and thus log in large
numbers of hours of practice. Challenges for the deliberate-practice framework include the development of
better methods for estimating the respective contributions of practice and talent, and a more differentiated
theoretical account for explaining the large interindividual variability and other factors such as starting
age that affect the development of expertise.

Cross-References
▶ Chunking Mechanisms and Learning
▶ Development of Expertise
▶ Individual Differences in Learning
▶ Learning in Practice
▶ Learning in the CHREST Cognitive Architecture

References
De Groot, A. D. (1978). Thought and choice in chess (De Groot, A.
Trans.). The Hague: Mouton. (Original work Het denken van den
schaker published 1946).
Ericsson, K. A., Krampe, R. T. H., & Tesch-Romer, C. (1993). The role
of deliberate practice in the acquisition of expert performance.
Psychological Review, 100, 363–406.
Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman, R. R.
(2006). The Cambridge handbook of expertise and expert performance. New York: Cambridge University Press.
Gobet, F., & Campitelli, G. (2007). The role of domain-specific
practice, handedness and starting age in chess. Developmental
Psychology, 43, 159–172.
Simon, H. A., & Chase, W. G. (1973). Skill in chess. American
Scientist, 61, 393–403.
Sternberg, R. J. (1996). Costs of expertise. In K. A. Ericsson (Ed.), The
road to excellence (pp. 347–354). Mahwah: Erlbaum.

D

919

Delinquency and Learning
Disabilities
ANNEMAREE CARROLL1, STEPHEN HOUGHTON2
1
School of Education, The University of Queensland,
Brisbane, QLD, Australia
2
Graduate School of Education, The University of
Western Australia, Nedlands, WA, Australia

Synonyms
Academic difficulties; Crime; Illegal activity; Offending

Definition
Juvenile delinquency is not a special education category, rather it is a legal term. Delinquency is defined
as participation in illegal behavior by a minor who falls
under a statutory age limit. According to Lorion et al.
(1987), delinquent behavior refers to a continuum of
behaviors that deviate from mainstream social standards in ways that have related, or could result in
serious disciplinary or adjudicatory consequences.
These behaviors can be socially unacceptable to school
authorities (e.g., classroom disruption), illegal and
problematic by virtue of the age of the offender (e.g.,
status offenses such as running away, substance use),
and illegal, criminal acts independent of the offender’s
age (e.g., assault, arson, robbery, rape).
Broadly defined, ▶ learning disabilities (LDs) refer
to problems in many academic areas such as reading,
comprehension, and mathematical abilities. Recent
movement away from a discrepancy model as posited
by DSM-IV-TR has resulted in LD being defined as
“lower than expected scores on tests of achievement
(typically significantly below average) given an individual’s age and educational opportunities” (see Rucklidge
et al. 2009).

Theoretical Background
Throughout history, LD has been perceived as
a contributing factor to criminality. As early as 1911,
Terman proposed that “No investigator denies the fearful role of mental deficiency in the production of vice,
crime, and delinquency. . . Not all criminals are feebleminded but all feeble-minded are at least potential
criminals.” This perception was held up until the early
twentieth century even though there was an absence (or

D

920

D

Delinquency and Learning Disabilities

at best limited) quantitative data linking criminality
and LD. At this time, evidence was based predominantly on observations and clinical reports.
The development and introduction of quantitative
methods has demonstrated a strong relationship
between delinquency and LD (Grigorenko 2006). For
example, prevalence rates of LD among delinquent
youths have been reported to range from 26% to 75%
(Rucklidge et al. 2009). Specifically, the types of LD
identified have included reading, writing, spelling and
mathematics difficulties, dyslexia, and comprehension.
Other studies have identified hearing impairment/
deafness and speech and language impairment. The
effects of LD are considerable following the young
person’s release from incarceration. Individuals with
LD are twice as less likely to engage in schooling or
work during the initial 12 months following release and
are approximately 2.3 times more likely to be reincarcerated compared to delinquents without LD
(Bullis et al. 2004).
So what reasons are put forward for this connection?
Research has posited three main hypotheses to explain
the LD-delinquency relationship. The susceptibility
hypothesis, the school failure hypothesis, and the differential treatment hypothesis, all pointed to LD as the
single cause of delinquency. Specifically, the susceptibility hypothesis suggested that neurological and intellectual difficulties including language deficits directly
contribute to the likelihood that an individual who has
LD may become delinquent. In the school failure
hypothesis, the individual has a strong desire to achieve
but his/her delinquency-prone temperament interacts
with feelings of frustration, rejection, and selfcriticalness that leads to school failure, dropping out of
school, and thereby elevating risk of delinquency. The
differential treatment hypothesis proposed that individuals with LD are detected more easily by the police and
their social skills deficiencies contribute to their differential treatment. Some researchers refer to this as the
differential arrest hypothesis and then add a further two
hypotheses (i.e., adjudication and disposition). In the
case of the differential adjudication hypothesis, the poor
self-control, irritability, abrasiveness, and inability to
respond effectively to questions increase the risk of
being adjudicated. For the differential disposition
hypothesis, the totality of the reasons in all of the
hypotheses creates a greater probability of delinquents
with LD being committed to a detention facility.

Contemporary research has confirmed that one
characteristic endemic to adolescents involved in delinquency is a low level of academic achievement. Moreover, the type of delinquent activity committed can be
differentiated according to level of academic achievement (e.g., those who commit more serious aggressive
crimes have lower academic achievement scores than
those committing property crimes; Grigorenko 2006).
Of those who underachieve academically, externalizing
problems (e.g., aggression, antisocial behavior, defiance, impulsivity, hyperactivity, and attention modulation) are evident (Willcutt and Pennington 2000).
Lower levels of IQ, particularly verbal intelligence
along with deficits in executive functioning have also
been demonstrated among those with LD who commit
delinquent acts.
The interaction between a myriad of factors including personal factors (e.g., neurological dysfunction,
social relationship difficulties, lack of empathy) and
other factors such as family risk factors (e.g., born to
teenage mother, parent or guardian criminality/alcoholism, parent psychiatric disorder, father absence,
harsh or inconsistent discipline style, child abuse),
school risk factors (e.g., poor student–teacher relationships, lack of engagement, school failure, negative
school climate), peer risk factors (e.g., association
with deviant peers), and community risk factors (e.g.,
low socioeconomic status, single parent family) multiplies the probability that those with LD will become
delinquent (Carroll et al. 2009). Although there are
a variety of failures to succeed, school failure is often
the first to be experienced.

Important Scientific Research and
Open Questions
Historically, the rates of LD among juvenile offenders
have been high although recent studies suggest prevalence rates may in fact be two to three times lower. There
are a number of explanations for this. First, the definitions of LD have varied across studies, tending to be
broad (e.g., problems in many academic areas), through
to specific (e.g., reading comprehension, writing ability).
Similarly, contentious issues have surrounded the definition of juvenile delinquency (JD). For example, questions regarding JD being defined as a specific behavior
(detected/undetected) or as a court-defined legal status
have given rise to concerns regarding the concept of JD
and actually who is delinquent. Furthermore, these

Delinquency and Learning Disabilities

definitions have been applied to young people in different forensic and non-forensic settings, including mainstream schools, special schools, and detention facilities
(detained individuals versus convicted individuals),
suggesting inconsistent use of the term across different
samples. Those in school settings attract the label delinquent as a result of their academic under-performance
while those in detention facilities have a delinquent status as a result of their suspected or proven more serious
offending.
Second, different methods and instruments have
been applied for diagnosing LD. For example, DSMIV-TR defined LD according to a discrepancy between
IQ and scores on standardized tests of achievement
whereas other sources use the low achievement model
whereby lower than expected scores on tests of achievement (typically significantly below average) given an
individual’s age and educational opportunities
(Rucklidge et al. 2009). In cases where categorical
thresholds have been imposed on a continuous measure (e.g., estimating rates of reading difficulties), the
prevalence rates will vary according to where the
threshold was established.
Third and inextricably linked to the first two points,
has been the failure by some early researchers to control
for the many confounding effects (e.g., socioeconomic
status, ethnicity, and family criminal history). In cases
where detained and non-detained individuals are
matched on these confounding variables that artificially inflate rates of LD (e.g., age, sex, SES, and IQ),
differences on multiple measures of achievement tend
to equalize, or in some cases disappear.
Fourth, the overlap between emotional and disruptive behavior disorders (E/DBDs) must be considered
because research has shown that E/DBDs tend to be the
most prevalent disability among young offenders and
that these have been associated with more serious outcomes in young adulthood (Carroll et al. 2009).
School graduation incentives are the most costeffective strategies for preventing delinquent behavior
(Grigorenko 2006). Therefore, of particular importance is understanding what preventive and intervention strategies are effective for juvenile delinquents
with LD. Addressing the frustration and failure these
young people experience from the outset of their
school experience, may reduce their involvement in
early disruptive behavior that escalates to more serious
delinquent activities. This may involve the way in

D

which teachers and principals exercise authority
through to the educational interventions that need to
take into account the wide range of LDs presenting
within JDs. While there is evidence that basic education
during detention can reduce reoffending, educators
must realize that greater benefits can be achieved
through more specialized programs (e.g., self-regulation, problem solving, interpersonal skills) that are
specific to the difficulties associated with the everyday
performance of young people with LD.
In summary, the evidence points to: LD causing
delinquency; delinquency causing LD; delinquency and
LD being caused by a combination which may include
a neurological condition and/or social and/or environmental factors; LD and delinquency developing independently but one is perceived as more severe by others; and
LD and delinquency being independent but chance associations occurring in study samples seem to suggest some
relationship. There is no doubt about the high rates of LD
among young people involved in the juvenile justice
system nor the established link between delinquency,
schooling experience, and academic achievement. While
recent studies have suggested that rates of LD may be
lower than first observed, the huge financial implications
along with the associated family and personal human
costs are compelling reasons for the continuing study of
the contribution of LD to delinquency.

Cross-References
▶ At-Risk Learners
▶ Development and Learning
▶ Individual Differences
▶ Styles of Engagement in Learning
▶ Vulnerability for Learning Disorders

References
Bullis, M., Yovanoff, P., & Havel, E. (2004). The importance of getting
started right: Further examination of the facility-to-community
transition of formally incarcerated youth. Journal of Special Education, 38, 80–94.
Carroll, A., Houghton, S., Durkin, K., & Hattie, J. (2009). Adolescent
reputations and risk: Developmental trajectories to delinquency.
New York: Springer. ISBN 978-0-387-79987-2.
Grigorenko, E. L. (2006). Learning disabilities in juvenile offenders.
Child and Adolescent Psychiatric Clinics of North America, 15,
353–371.
Lorion, R. P., Tolan, P. H., & Wahler, R. G. (1987). Prevention. In
H. C. Quay (Ed.), Handbook of juvenile delinquency (pp. 383–
416). Oxford, England: Wiley.

921

D

922

D

Delivery Systems

Rucklidge, J. J., McLean, A. P., & Bateup, P. (2009). Criminal
offending and learning disabilities in New Zealand youth. Does
reading comprehension predict recidivism? Crime and Delinquency. doi:10.1177/0011128708330852.
Willcutt, E. G., & Pennington, B. F. (2000). Comorbidity of reading
disability and attention deficit hyperactivity disorder: Differences by gender and subtype. Journal of Learning Disabilities,
33, 179–191.

Derivational Analogy
▶ Analytic Learning

Descriptive and Interpretative
Learning Analysis
Delivery Systems

▶ Qualitative Learning Research

▶ Media and Learning

Descriptive Theory of
Knowledge Development
Democratization of Education

▶ Naturalistic Epistemology

▶ Compulsory Education and Learning

Desensitization
Deontic Reasoning

▶ Habituation and Sensitization

▶ Pragmatic Reasoning Schemas

Dependence

Desiderius Erasmus
(1466/69–1536)

▶ Contingency in Learning
DAMIAN GRACE
Department of Government and International
Relations, The University of Sydney, Sydney, Australia

Depression: Clinical Depression
▶ Goals and Goalsetting: Prevention and Treatment of
Depression

Depth of Processing
The level to which information is processed mentally,
with processing depth being determined by associations with existing memory, or cognitive effort.

Life Dates
Erasmus was born in Rotterdam most probably in 1467
(1466 and 1469 are also possible) to unmarried parents,
Roger, a priest, and Margaret, daughter of a physician.
Educated at a school that was progressive for the time,
Erasmus became an Augustinian canon. He was later
dispensed from his duties as a priest to devote himself
to philology, scriptural exegesis, and reform in Church
and society. His enthusiasm for reform did not extend
to Luther’s Reformation. While there were many points
of agreement between him and the Reformers, the

Desiderius Erasmus (1466/69–1536)

harshness of Luther’s approach was foreign to Erasmus’
moderate style of thought and conduct. He died of
dysentery in Basel in 1536.

Contribution(s) to the Field of
Learning
Erasmus was the most considerable scholar of northern
Renaissance Europe, writing many influential books
and leaving an unrivaled collection of correspondence
with early sixteenth-century humanists, such as
Thomas More and Guillaume Budé. Humanists were
devotees of classical literature and the moral values
embodied in it. In the fifteenth century, Italian humanists effectively broke from the worldview and methods
of scholasticism, which revered Aristotle as “the Philosopher” and placed a premium on logical analysis
and argument in education. The humanists, by contrast, advocated the studia humanitatis, a curriculum
based on classical literature and the study of grammar,
rhetoric, history, and moral philosophy. Humanism in
the age of Erasmus flowered just as the printing press
was beginning to make its impact on education and
learning. The standard method of instruction when
Erasmus was at school was rote memorization of texts
and relentless drills in styles of argument. The printing
press was part of the reform of instruction by making
books more accessible, and even Erasmus’ early
unpolished productions were greeted enthusiastically.
All of Erasmus’ works – even satires such as his most
widely read work, The Praise of Folly – were written to
enlighten, to educate in some way, but he wrote a string
of specifically pedagogical texts on liberal education,
composition, and teaching, starting with De ratione
studii (On the Method of Study), addressed to the
education of young adolescents. This was eventually
published in 1512 with De copia rerum ac verborum
(Foundations of the Abundant Style), written at the
request of John Colet for his St. Paul’s School. The two
works belong together, for the latter builds on the
foundations of the former. Assuming that the basics
of language have already been acquired, the Copia
addresses the fluency, propriety, and precision of the
writer or speaker and the finding of appropriate language to persuade an audience. It goes on to offer an
armory of resources, such as examples, comparisons,
and contrasts for the amplification of arguments. The
Copia became a cornerstone of sixteenth-century education (Callahan 1978, p. 100).

D

De ratione studii was also widely influential, especially in the pedagogical works of others which now
came from the presses. One such is a remarkable
work – “Lily’s Grammar” – written by John Colet’s
high master at St Paul’s School, William Lily. This grammar became the standard text for Latin instruction in
sixteenth-century England and remained in use into the
eighteenth century. Apart from his own works, such
collaborations with other humanists ensured that Erasmus’ views became part of the fabric of education in the
sixteenth century. His Colloquies, a collection of dialogs
on a range of contemporary issues intended to introduce students to colloquial Latin, continued to be used
for this purpose even when it had been transformed in
later editions into a mature work of literature. More
conventional was The Education of a Christian Prince,
a work in the tradition of mirror-of-princes writing. The
aim of this work was to restrain the customary excesses
of princes – taxes and war were prominent among
them – by education. This optimism about the moral
power of education sprang from a real concern about
the alternative: if a prince is not educated well, then the
evil tendencies of human nature will warp the prince’s
character with disastrous consequences for his kingdom.
Erasmus was not a radical educational reformer and
he has even been called conservative because of his
defense of a traditional curriculum in the classics
(Thompson 1978, p. xx; Sowards 1988, p. 124). This
is true up to a point, but it is important to appreciate
the central role of Erasmus in the defense of the classical curriculum against scholastic teachers indifferent
alike to good Latin and good letters – especially the
pagan authors. Greek works, in particular, excited
opposition from those (mainly in the religious orders)
who wished to keep Christian studies free of their taint.
The humanists tended to overstate this hostility, but
there is no doubt that Erasmus brought a deep detestation for scholastic methods and Aristotelian philosophy to his defense of good letters. This did not make
him less of a Christian. His humanism was Christianized and authors like Plato were viewed through Christian assumptions. Moreover, Erasmus did not believe
that pagan texts inculcated virtue directly, but that they
set the Christian student “going in that direction”
(Thompson 1978, p. xxx).
If Erasmus was more a defender than reformer of
the curriculum, he was conspicuous as an advocate of
teaching reform, especially in his younger years when

923

D

924

D

Desiderius Erasmus (1466/69–1536)

he was optimistic that education could drive reform in
religion and society. At a time when teaching tended to
be regarded as an occupation fit only for those with no
other prospects (Tracy 1972, p. 64), Erasmus promoted
the idea of the good teacher as the maker of good
citizens. In 1516, he encouraged a despondent schoolmaster, Johannes Sapidus, by characterizing teaching as
a singular service to society. There is no greater contribution that a learned and honorable man can make to
his country than shaping “its unformed young people”
he declared (Sowards 1988, p. 124).

Important Scientific Research and
Open Questions
Innovation for Erasmus lay in teaching the classics with
an understanding that they cultivated character, morality, and civility. It was important for teachers to
strengthen and not harm with punishments those naturally possessed of a humane or gentle disposition. As
for pupils of a more passionate nature, he doubted that
they could be taught much at all, so punishments
would probably fail with them too (Tracy 1972,
p. 11). These precepts were not entirely novel in Erasmus’ day for they mirror those of Quintilian in the
Institutes (Book 1, Chap. 3, 14), but he had to defend
them against those with a scholastic view of education
and a penchant for invective and the rod.
It might also seem unremarkable that, in De pueris
Instituendis (On the Education of Children), he advocated education from a very early age and the use of
pictures to instruct young children in matters – such as
exotic animals – that they have not experienced. Erasmus recommends using various playful devices such as
toy letters of the alphabet to encourage learning and the
use of similes and analogies so that young children may
work from their knowledge of familiar objects to an
understanding of unfamiliar ones. Thus should they
acquire a knowledge of language and a knowledge of
things simultaneously, but also some intimation that
words are not univocally related to things. Older students should, according to De ratione studii, be
instructed in the practice of discourse, but this practice
is to be rooted in usage, not in rules of grammar
learned by heart. Even though the stress for older students has moved to a knowledge of language as a path
to the knowledge of things, Erasmus insists that that
knowledge comes from best usage, not from definitions
(Margolin 1978, p. 226 f.). The best usage was to be

found in Cicero, but he was not to be followed slavishly,
as Erasmus makes clear. For this Ciceronian was also
the author of the most famous satire of Cicero-worship,
the Ciceronianus of 1528.
The title of Erasmus’ late work, De civilitate morum
puerilium libellus (A Handbook on Good Manners for
Children), published first in 1530, describes its contents. It is a manual of good manners ranging from
personal hygiene to conduct at play and etiquette in
company. It aims to educate the child in civility and the
acquisition of the self-discipline necessary to become
an adult. It is not so much a handbook of etiquette as
a way for the child to develop an appropriate disposition. Erasmus believed that good manners had moral,
as well as civic value, and sowed in children the seeds of
piety, the love of learning, and duty.
As the Reformation gathered pace and his hopes for
moderate reform became remote, Erasmus became
more pessimistic about the power of education to effect
change (Tracy 1972, p. 232). Certainly the humanitas
he valued receded in a climate of dispute, but it did not
disappear. Erasmus continued to believe in
a conversational rather than a dogmatic style of
instruction; in the power of an appealing argument to
be more convincing than a logically correct but unfeeling one; in the futility of placing too much trust in the
powers of reason to clarify the human condition. Erasmus had learnt from Cicero the wisdom of examining
both sides of an argument and then, perhaps,
suspending judgment. He did not waver from his
view that bombastic language of any kind – scholastic,
Lutheran, pseudo-Ciceronian – was less likely to engage
an opponent than the conversational style he
recommended in the range of his works and exemplified as well as he could in his life.

Cross-References
▶ Epistemology and Learning in Medieval Philosophy
▶ History of the Science of Learning

References
Callahan, V. W. (1978). The De Copia: The bounteous horn. In R. L.
DeMolen (Ed.), Essays on the works of Erasmus. New Haven and
London: Yale University Press.
Margolin, J. C. (1978). The method of “words and things” in
Erasmus’s De Pueris Instituendis (1529) and Comenius’s Orbis
sensualium pictus (1658). In R. L. DeMolen (Ed.), Essays on the
works of Erasmus. New Haven/London: Yale University Press.

Design Experiments
Sowards, J. K. (1988). Erasmus as a practical educational reformer. In
J. S. Weiland, W. Frijoff, (Eds.), Erasmus of Rotterdam: The man
and the scholar: Proceedings of the symposium held at the Erasmus
University, Rotterdam, November 9–11, 1986. Leiden: E.J. Brill.
Thompson, C. R. (1978). Introduction. In Literary and educational
writings. Collected works of Erasmus (Vol. 23). Toronto: University of Toronto Press.
Tracy, J. D. (1972). Erasmus: The growth of a mind. Genève: Librairie
Droz.

Design Experiments
NORBERT M. SEEL
Department of Education, University of Freiburg,
Freiburg, Germany

Synonyms
Design-based research

Definition
Design experiments can be considered as a special case
of field experiments for improving external validity.
Design experiments aim at particular forms of educational interventions that create novel conditions for
learning and instruction. They can be used for replication studies, and thus contribute to causal inferences
and a gradual increase in the ecological validity of
experimental designs in the area of instruction.

Theoretical Background
The idea of design experiments has its origins in criticism of instructional research and its applicability in
educational practice. According to Reeves (2006), the
quality of published research in the field of instructional research is generally poor, and Stokes (1997)
argues that instructional research contributes little to
the basic theories that underpin teaching and therefore
has little value for solving practical problems. Both
authors reiterate a comment made by Cronbach in
1975 about instructional research in general. To bridge
the gap between instructional research and practice,
Brown (1992) introduced the concept of design experiments into the methodological discussion. Design
experiments have their roots in traditional experimental and quasi-experimental research on learning and
instruction, but go beyond the laboratory. Design

D

experiments are developed explicitly as a means of
formative research for testing and refining educational
problems, solutions, and methods. Referring explicitly
to experimental research in education, Brown (1992)
focused on the central question of how to transfer the
results of this research into the classroom. She
envisioned a dynamic relationship between the classroom and laboratory research capable of bridging the
gap between theory and practice. In terms of traditional (quasi-) experimental research, this question
corresponds to the issue of external validity of
experimentation.
Brown’s central idea is developing theories of learning and instruction that work in practice. In order to
create effective conditions for learning in the classroom, learning experiences should be engineered with
an eye to the application of innovative techniques of
learning and instruction. Therefore, engineering
a working (or learning) environment can be seen as
the heart of design experiments. Engineering a working
environment presupposes the realization of a particular
learning theory and in turn contributes to the maintenance of this theory. Other constituents of design
experiments have been illustrated by Brown (1992), as
shown in Fig. 1.

Constituent 1: Contributions to
Learning Theory
The fundamental basis of design experiments is a strong
mutual relationship between the engineered working
(learning) environment and its “contribution to learning theory.” On the one hand, a learning theory is the
fundamental basis for the design of a learning environment, which may be considered as a realization of the
central assumptions of the learning theory. On the other
hand, the designed environment can make a strong contribution to the validation of the learning theory. In
order to illustrate this constituent of design experiments,
we can refer to examples from instructional research in
the past 25 years, such as “anchored instruction”
(Pellegrino and Brophy 2008), “goal-based scenarios”
(Schank et al. 1993/94), and “model-based learning and
instruction” (Seel 2003).

Constituent 2: Practical Feasibility
and Dissemination
A central argument concerning the usability of design
experiments is their feasibility and dissemination.

925

D

926

D

Design Experiments

Input
Classroom ethos
teacher/student as researcher
curriculum
technology etc.

Contributions
to learning theory

Engineering a
working environment

Output
Assessment of the right things
accountability

Practical
feasibility
(Dissemination)

Design Experiments. Fig. 1 The constituents of design experiments as described by Brown (1992, p. 142)

Again, we can refer to the aforementioned examples of
anchored instruction and goal-based scenarios. Indeed,
the Cognition and Technology Group at Vanderbilt
demonstrated over a period of 15 years with both the
Jasper Woodbury project and STAR-Legacy that it is
possible to transform instructional theory into educational practice (Pellegrino and Brophy 2008). These
examples demonstrate that good theories and their
instructional realizations have a real chance of being
situated in the classroom as well.

Constituent 3: Input Variables
Input variables of design experiments correspond to
a large extent to the independent variables of instructional experimentation. Among other things, Brown
(1992) focused on class ethos, curriculum, and technology as important input variables. Beyond these variables, a comprehensive meta-analysis by Wang et al.
(1993) identified learner characteristics (i.e., personality traits and learning styles) as the most influential
factor of instructional effectiveness. The second major
factor influencing effective instruction has been identified as the quality of instructional practice. However,
Brown (1992) does not mention individual differences
in personality traits and learning styles as input variables. Rather, she emphasizes the idea that teachers and
students could act as “researchers.” This idea corresponds to the approach of model-based learning and
instruction as well as to ▶ Learning by Design.

Constituent 4: Output Variables
Brown (1992) argued for measuring the “right things,”
such as critical thinking, explorative learning, and
problem solving. Although this may be a complicated
task, it is necessary to invest more time and effort in the
development and validation of appropriate instruments for the assessment of learning outcomes because
the quality of measurement and assessment determines
the quality of research and its contribution to theory
and practice (Pellegrino et al. 2001). In this context, the
phenomenon known as testing effect should be mentioned. It states, for instance, that taking a memory test
enhances later retention. Several studies lead to the
conclusion that “testing is a powerful means of improving learning, not just assessing it” (Roediger and
Karpicke 2006, p. 249).

Important Scientific Research and
Open Questions
Developing and disseminating a new research methodology is a slow process that more often proceeds stepby-step than by sudden decisive changes of previous
paradigms. This general observation also holds true
with regard to design experiments, which compete
against the paradigms of both quasi-experimental and
action research on learning and instruction. In addition, due to the fact that Brown did not describe design
experiments in a systematic manner but rather in
a personal narrative style, the follow-up approaches of

Design Experiments

design-based research focus on divergent aspects of
design experiments. Thus, we can observe that the
aspect of systematic experimentation is often ignored
and action research is considered as the most relevant
component of educational research. However, Brown’s
statements indicate that she did not contrast design
experiments and quasi-experimental studies but rather
viewed them as two sides of the same coin: Experimental and quasi-experimental research aims at optimizing
the internal validity of research, whereas design experiments aim at attaining a greater ecological validity of
educational interventions.
Proponents of design experiments and designbased research see their origins in applied research
designed to solve practical problems. Actually, design
experiments aim explicitly at testing theories about
how changes in knowledge and skills take place as
a result of teaching (in the classroom). At first glance,
it seems that design experiments test hypotheses in
a manner that is similar but not identical to traditional
experimental research. However, traditional experimental research follows the falsification principle,
whereas design experiments aim at “conjecture-driven
tests” (Cobb et al. 2003) that correspond to the realization principle of constructivism. This principle states
that reality is not mapped onto theories, but rather
particular realities are intentionally selected and created by experiments in order to confirm the theory
(Seel 2009).
With regard to the intended systematic experimentation in educational fields of interest, design experiments correspond to the methodology of field
experiments aiming at investigating the effectiveness of
educational interventions. Brown clearly defines design
experiments as “intervention research designed to
inform practice . . . that should be able to migrate
from the experimental classroom to average classrooms
operated by and for average students and teachers,
supported by realistic technological and personal support” (p. 143). Whereas experimental studies in the
laboratory are regularly very artificial and their results
often lack applicability to real world problems, experimental field studies take place in this real world and
generalization is therefore not a problem. Therefore,
field experiments are said to have high external or
ecological validity (Cook and Campbell 1979).
Like laboratory experiments, field experiments generally randomize the sampling of subjects and their

D

assignment into treatment and control groups and
compare outcomes between these groups. However,
randomization is typically unavailable in field settings
because the researcher is not able to manipulate treatment conditions at the level of the individual participant. In addition, it is difficult to control the relevant
variables in a field experiment, while we can control
these variables rigorously in a laboratory experiment.
Nevertheless, the systematic control of variables is
a hallmark of the laboratory experiment, another being
the manipulation of independent variable(s). In fact,
from a methodological perspective, the lack of control
over certain elements is the greatest problem with field
studies and design experiments. For example, in free
learning environments, the researcher has little or no
control over information sources the participants may
be using, and even though participants may be advised
to avoid this, there is no guarantee that they will conform. Another problem with field studies and design
experiments is that they tend to be more time consuming and therefore more expensive and demanding than
laboratory studies.
Design experiments, and especially the related
approaches of design-based research, also refer to
action research as a reflective process of progressive
problem solving conducted by individuals working
with others in teams or as part of a “community of
practice” to improve the way they address issues and
solve problems. Although Brown (1992) and other proponents of design experiments do not refer to the work
of Lewin, who coined the term “action research” in the
1940s, there are obvious parallels in the major lines of
argumentation. In 1946, Lewin described action
research as “a spiral of steps, each of which is composed
of a circle of planning, action, and fact-finding about
the result of the action.” Its practice can best be characterized as research for social management or social
engineering.
Nevertheless, inspired by several concrete examples
provided by Brown, practitioners often associate design
experiments with qualitative research methods (see,
for example, Botha et al. 2005). However, Brown
emphasized the application of both methodologies:
A “traditional use of static pretests and posttests, combined with appropriate control data, provides us
with clear evidence of the effectiveness of the intervention and is easy to share with school personnel as
well as fellow scientists” (p. 157). In sum, Brown’s

927

D

928

D

Design of Computer Games for Education

argumentation amounts to an appeal for combining
qualitative and quantitative methodologies.
Design experiments correspond, to some extent, to
the methodology of matched sampling for deriving
causal effects. Matched sampling is often used to
help assess the causal effect of some intervention or
exposure, typically when randomized experiments are
not available or cannot be conducted. Matched samples
can arise in a situation in which the same variable or
attribute is measured twice on each subject under
different circumstances. In accordance with this methodology, Seel (2009) suggested using design experiments as a heuristic for both designing synthetic
learning environments and conducting systematic
research on them. More specifically, he argues that
design experiments can be used for replication studies
and thus contribute to causal inferences and ecological
validity.
The methodology of design experiments may provide an excellent framework for replication studies.
Although the traditional view of replication entails
the collection of new data – including data on additional cases or additional measures – statisticians and
social scientists have suggested alternative replication
strategies. One is to build replication into a study from
the start. For example, a researcher can draw a sample
large enough to allow random partitioning into two
subsamples. Data from one subsample can then be used
to check conclusions drawn on the basis of analyses of
data from the other. Another approach requiring the
intensive use of computing resources is to draw multiple random subsamples from already collected data and
then use them to cross-validate results. Still another
elaboration of the basic idea of replication is the general
approach called meta-analysis.

Cross-References
▶ Action Research on Learning
▶ Effects of Testing on Learning
▶ Experimental and Quasi-Experimental Designs for
Research on Learning
▶ Field Experiments in Learning Research
▶ Field Research on Learning
▶ Learning by Design
▶ Longitudinal Learning Research
▶ Methodologies of Learning Research: Overview
▶ Teaching Experiments and Professional Learning

References
Botha, J., van der Westhuizen, D., & De Swardt, E. (2005). Towards
appropriate methodologies to research interactive learning:
Using a design experiment to assess a learning programme for
complex thinking. International Journal of Education and Development using Information and Communication Technology, 1(2),
105–117.
Brown, A. L. (1992). Design experiments: Theoretical and methodological challenges in creating complex interventions in classroom
settings. The Journal of the Learning Sciences, 2(2), 141–178.
Cobb, P., Confrey, J., DiSessa, A., Lehrer, R., & Schauble, L. (2003).
Design experiments in educational research. Educational
Researcher, 32, 9–13.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation:
Design and analysis issues for field settings. Chicago: Rand
MacNally.
Pellegrino, J. W., & Brophy, S. (2008). From cognitive theory to
instructional practice. Technology and the evolution of
anchored instruction. In D. Ifenthaler, P. Pirnay-Dummer, &
J. M. Spector (Eds.), Understanding models for learning and
instruction. Essays in honor of Norbert M. Seel (pp. 277–303).
New York: Springer.
Pellegrino, J., Chudowsky, N., & Glaser, R. (Eds.). (2001). Knowing
what students know. The science and design of educational assessment. Washington, DC: National Academy Press.
Reeves, T. C. (2006). Design research from the technology perspective. In J. V. Akker, K. Gravemeijer, S. McKenney, & N. Nieveen
(Eds.), Educational design research (pp. 86–109). London:
Routledge.
Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning.
Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255.
Schank, R. C., Fano, A., Bell, B., & Jona, M. (1993/94). The design of
goal-based scenarios. The Journal of the Learning Sciences, 3(4),
305–345.
Seel, N. M. (2003). Model-centered learning and instruction. Technology, Instruction, Cognition, and Learning, 1(1), 59–85.
Seel, N. M. (2009). Bonjour Tristesse: Why don’t we research as we
have been taught? Methodological considerations on simulationbased modelling. Technology, Instruction, Cognition and Learning, 6(3), 151–176.
Stokes, D. E. (1997). Pasteur’s quadrant – basic science and
technological innovation. Washington, DC: Brookings Institution
Press.
Wang, M. C., Haertel, G. D., & Walberg, H. J. (1993). Toward
a knowledge base for school learning. Review of Educational
Research, 63(3), 249–294.

Design of Computer Games for
Education
▶ Designing Educational Computer Games

Design of Learning Environments

Design of Learning
Environments
DIRK IFENTHALER
Institut für Erziehungswissenschaft,
Albert-Ludwigs-Universität Freiburg, Freiburg,
Germany

Synonyms
Instructional design; Instructional systems design;
Instructional technology

Definition
The design of learning environments is the systematic
analysis, planning, development, implementation, and
evaluation of physical or virtual settings in which learning takes place.

Theoretical Background
Learning environments are physical or virtual settings
in which learning takes place. Learning theory provides
the fundament for the design of learning environments.
However, there is no simple recipe for designing learning environments (Bransford et al. 2000). Additionally,
the design of learning environments will always change
in alignment with the change of educational goals.
Hence, the design of learning environments in the
1800s or 1900s was extremely different to the twentyfirst century design of learning environments. Generally, the design of learning environments includes the
three simple questions: What is taught? How is it
taught? How is it assessed? Yet, the design of learning
environments is not simply asking the above stated
three questions. Rather, it includes a systematic analysis, planning, development, implementation, and evaluation phases (Gagné 1965; Merrill 2007). The analysis
phase includes needs analysis, subject matter content
analysis, and job or task analysis. The design phase
includes the planning for the arrangement of the content of the instruction. The development phase results
in the tasks and materials that are ready for instruction.
The implementation phase includes the scheduling of
instruction, training of instructors, preparing time
tables, and preparing evaluation parts. The evaluation
phase includes various forms of formative and

D

summative assessments. The above described model
for the design of learning environments presents a
general heuristic. However, it is also often
criticized being too narrow and inflexible (Dijkstra
and Leemkuil 2008).
Bransford and colleagues (2000) differentiate
four perspectives for the design of learning
environments: learner-centered, knowledge-centered,
assessment-centered, and community-centered learning
environments. The design of learner-centered learning environments needs to take notice of learner’s knowledge, skills, attitudes, beliefs, and cultural practices as
well as including instructors and/or virtual tutors who
are aware of the learners’ characteristics (see Bransford
et al. 2000). The design of knowledge-centered learning
environments highlights the prior knowledge of
learners. Accordingly, learners’ preconceptions about
a specific phenomenon in question are vitally important. Additionally, the design of knowledge-centered
learning environments includes authentic problem
situations for learners. The design of assessmentcentered learning environments aims at combining
assessment of content knowledge and necessary skills
for specific tasks or problems. Providing feedback is the
main objective for designing assessment-centered
learning environments. Feedback could be any type of
information provided to learners. Moreover, feedback
is considered a fundamental component for supporting
and regulating learning processes. The nature of
feedback plays a critical role in learning and instruction
especially in technology-based and self-regulated
learning environments. The design of communitycentered learning environments combines several
aspects of community, including classrooms, schools,
universities, workplaces, homes, cities, states,
countries, and the virtual world. Hence, the sense of
community is involved in the design of communitycentered learning environments where instructors
and learners share their understanding of norms
and values.
Current, the design of learning environments involves
the idea of using computers for logical reasoning and
intelligent agents – an old dream of artificial intelligence:
Such applications will be designed to execute operations
of logical thinking using a multitude of rules which
express logical relationships between terms and data in
the Web. In view of the countless unfulfilled promises of
artificial intelligence in the 1980s and 1990s, however, one

929

D

930

D

Design of Learning Environments

would be well advised to remain skeptical on this point.
Nevertheless, the integration of artificial intelligence into
the Internet is a goal of the development of Web 3.0.
Should this vision actually, as has been prophesied,
be realized within the next 10 years, it would bring
about another Internet revolution and open up new
horizons for the design of learning environments. The
first step in this process might be to enable users of the
Web to modify learning environments and information
resources and create their own structures. In this way,
Web 3.0 could provide the basis for free or personal
learning environments, which have been regarded by
educational theorists as the quintessential form of
learning environment for decades (Morris 2011).
Accordingly, the definition of the design of learning
environments needs to be adapted over and over
again to the current results of instructional research
and technological progressions.

Important Scientific Research and
Open Questions
Versatile research has been conducted in the field of
learning and instruction which motivated instructional
designers to redefine the principles of teaching and
learning (Bransford et al. 2000). However, the days of
preprogrammed learning environments are numbered,
in which the learner – as in the classical paradigm of
programmed instruction – is viewed more as an audience than as an active constructor. In the near future,
learners will be the constructors of their own environments and create the structures of the content units on
their own (Morris 2011). Web-based systems designed
to optimize or supplement learning environments are
cropping up everywhere. The rapid pace of these technological developments makes it nearly impossible to
integrate them into comprehensive systems. Therefore,
so-called personal learning systems (PLS) are being
designed to enable learners to select various Web applications individually to meet specific learning goals
(Ifenthaler 2010; Seel and Ifenthaler 2009). The
requirements and features for deigning PLS are:
Portal: Rather than an isolated island, a PLS is an open
portal to the Internet which is connected with various applications and collects and structures information from other sources. The content can be
created by both learners and teachers using simple
authoring tools.

Potential for integration: Information is offered in
standard formats which learners can subscribe to
and synchronize with their desktop applications. In
this way, the learning environment is integrated
into the user’s daily working environment and
connected to it.
Neutrality of tools: Tasks in the learning environment
are designed in such a way that the learners themselves can choose which application they wish to
use to work on them. The portal can make recommendations and provide support. The media competence acquired in this manner can also be useful
in everyday life.
Symbiosis: Instead of creating new spaces, a PLS uses
existing resources. The portal works with existing
free social networks, wikis, blogs, etc.
All in all, personal learning systems require
increased personal responsibility, both from the learner
and from the instructor. At the same time, however,
they offer more freedom for individual learning. Yet, no
empirical studies are available which account for the
efficiency of PLS. Hence, much research is needed in
near future to investigate the strength and weaknesses
of these newly designed learning environments.
It is of course difficult to predict new developments
or trends in the domain of the design of learning
environments with any kind of precision, but one
thing is certain: They will continue to be dictated to
a great extent by the general development of information and communication technology.

Cross-References
▶ Blended Learning
▶ Computer-Based Learning
▶ e-Learning
▶ Interactive Learning Environments
▶ Learning Environment
▶ Online Learning

References
Bransford, J. D., Brown, A. L., & Cocking, R. R. (Eds.). (2000). How
people learn: Brain, mind, experience, and school. Washington,
DC: National Academy Press.
Dijkstra, S., & Leemkuil, H. (2008). Developments in the design of
instruction. From simple models to complex electronic learning
environments. In D. Ifenthaler, P. Pirnay-Dummer, & J. M. Spector
(Eds.), Understanding models for learning and instruction. Essays
in honor of Norbert M. Seel (pp. 189–210). New York: Springer.

Designing Educational Computer Games
Gagné, R. M. (1965). The conditions of learning. New York: Holt,
Rinehart, and Winston.
Ifenthaler, D. (2010). Learning and instruction in the digital age. In
J. M. Spector, D. Ifenthaler, P. Isaı́as, Kinshuk, & D. G. Sampson
(Eds.), Learning and instruction in the digital age: Making
a difference through cognitive approaches, technology-facilitated
collaboration and assessment, and personalized communications
(pp. 3–10). New York: Springer.
Merrill, M. D. (2007). The future of instructional design: The proper
study of instructional design. In R. A. Reiser & J. V. Dempsey
(Eds.), Trends and issues in instructional design and technology
(pp. 336–341). Upper Saddle River: Pearson Education, Inc.
Morris, R. D. (2011). Web 3.0: Implications for online learning.
TechTrends, 55(1), 42–46.
Seel, N. M., & Ifenthaler, D. (2009). Online lernen und lehren.
München: Ernst Reinhardt.

Design-Based Research
▶ Design Experiments

Designing Educational
Computer Games
CYRIL BROM1, VIT SISLER2
1
Department of Software and Computer Science
Education, Charles University in Prague, Prague 1,
Czech Republic
2
Institute of Information Studies and Librarianship,
Charles University in Prague, Prague 5, Czech Republic

Synonyms
Design of computer games for education

Definition
The design of any product is one part of the product’s
development process. The design of an educational
computer game (ECG) is a specific extension of
▶ instructional design in that it employs a computer
game as a means for accomplishing part of the educational objective. Conventional instructional design is
a process of (1) identifying learners’ needs; (2) identifying pertinent useful knowledge or skills; (3) analyzing
possible means of achieving this knowledge or skills,

D

considering the environmental context; (4) designing
an appropriate educational process based on the results
of this analysis; and (5) specifying assessments of the
target knowledge or skills. In ECG design, possible
games or game genres suitable for achieving defined
learning objectives are analyzed in Step 3, a particular
game is designed in Step 4, and possibly also considered
as a means in Step 5. Furthermore, designing the particular educational game has to follow general game
design principles. Therefore – on a conceptual level –
rules, game play and content of the game are specified;
and on a technological level, software architecture and
the game’s main algorithms are described. On the conceptual level, the goal of successful game design is the
creation of engaging and meaningful ▶ play; on the
technological level, it is the creation of functioning
and smoothly operating software. The exact ways of
verifying the game (i.e., checking whether the software
works) and its validation (i.e., determining whether the
game achieves its intended purpose, which is to help
fulfill the educational objective) should also be specified in the design phase. Depending on the educational
objectives, supplementary non-gaming teaching activities must also be designed, usually in Steps 3 and 4.

Theoretical Background
The field of educational computer games is very young,
and most research questions have not yet been
answered, including the general question how to design
effective ECGs. However, synthesis from literature on
general instructional design (e.g., Reif 2008), game
design (Salen and Zimmerman 2005), recent findings
on ECGs (e.g., Sisler and Brom 2008; Klopfer 2008),
and reviews of past research on ▶ educational games
and ▶ simulations, including non-computer ones (e.g.,
Hays 2005; Gredler 2004), starts to reveal the following
design framework:
1. A target group of learners and teachers should be
identified in advance, and the educational context
(e.g., requirements posed by the duration of fixed
lessons in secondary education) and accepted educational practice (e.g., national curricula) should
be known.
2. Learning objectives must be identified, clearly formulated, and set out in accordance with accepted
educational practice before development starts.
Learning objectives should be formulated in

931

D

932

D

Designing Educational Computer Games

operational terms, i.e., in terms of concrete,
observable abilities a student will acquire after
playing the game and conducting supplementary
nongame activities. In other words, it should be
known what the student will be able to actually do
after finishing the game. For instance, with respect
to a second language acquisition, it should be
specified that students should be able “to translate
texts of a particular complexity,” as opposed to
stating that they will “understand 500 particular
words.”
3. After learning objectives are formulated, a detailed
analysis of trade-offs among alternative
approaches should be conducted before the ECG
approach is possibly chosen. The game genre must
be chosen to meet the educational goals and to fit
the subject. Will the game be a multiplayer one?
Will it be an adventure-style game or a puzzle? The
important design choice is whether the learning
experience will be delivered by the game directly,
or whether the learning materials and the game
will be two separate systems and the game will be
played merely as a reward. The latter model is an
example of behavioristic “▶ drill and practice”
approach. While it is often criticized in ECG literature, it may actually work well for acquisition of
facts, e.g., foreign-word rehearsal. Yet, it seems that
for generating a deeper understanding of certain
key principles of given topics and for acquisition of
high-level skills, complex games and simulations
which allow open-ended exploration and integrate
learning materials with the play are more useful.
Nevertheless, it has to be emphasized that the
empirical research on effectiveness of both types
of games is far from conclusive.
4. The development of any computer game is an
interdisciplinary task. Game designers, graphics,
programmers, testers, etc., must be involved.
Additionally, when developing an ECG, educationalists, content developers, teachers, writers of
supplementary educational texts, etc., must be in
the team. The involvement of teachers in the early
design phases is of critical importance.
5. On the conceptual level, the goal of successful
game design is the creation of meaningful play.
Meaningful play in a game emerges from a clear
relationship between player action and game’s system outcome. The result of the game action has to

6.

7.

8.

9.

be communicated to the player in a perceivable
way. Simultaneously, the relationship between
action and outcome should be integrated into
the game’s larger context: i.e., the players should
know what outcomes their actions will result in as
well as which actions to take in order to win the
game.
Tentative game rules and game play must be
designed before the software development starts,
and a prototype of the game should be created.
This prototype should be evaluated on target users
and the game specification should be further
refined accordingly. Similarly, game content (e.g.,
graphics, supplementary texts, etc.) should be
evaluated as soon as possible in the development
phase and adjusted accordingly. Continuous
reevaluations are vital. The specifics of the target
group must be considered; e.g., in schools, an ECG
will usually have to be appealing to both males and
females.
The game should be challenging enough but not
too hard or complex to play since the latter might
result in students’ frustration. The dynamics of the
game should be simultaneously understandable by
the target group of players and engaging at the
same time. Winning should be based on knowledge or skills, not random factors. Sanctioning
students for making wrong choices must be balanced. In multiplayer games, “the rich get richer
while the poor get poorer” principle should be
avoided.
The complexity of the game’s user interface must
be adjusted to non-players (these may be both
teachers as well as learners). Ideally, the players
should become gradually familiar with the game,
its rules and interface directly through playing the
game, rather than being forced to resort to manuals or tutorials. In other words, the game has to
be integrated with instructional objectives and
other instructions.
Technological and personal limitations should be
considered. For instance, will the game work on
the school hardware? Would the fixed lessons’
durations pose barriers to their usage? If any of
these questions poses a problem, alternative solutions should be considered. For example, it is
possible to design a game as a supplementary
activity for home play or to consider PDA

Designing Educational Computer Games

technology. Client-server applications may help
with removing the need for end-user installations.
It might be necessary to simplify the graphic content to fit the graphic cards and internet connection available, etc.
10. No educational game can be viewed as a standalone activity. Supplementary collaborative activities, educational materials, and teaching methodologies need to be prepared in advance but in
cooperation with end-users. Further improvements in the course of development by means of
evaluation studies are vital.
11. Teachers, and often also students, should know the
educational objectives in advance. Their experience with the game should be visibly relevant to
these objectives. Research has shown that learners
should often be provided with debriefings and
feedback that explain how their experiences with
the ECG help them to fulfill these objectives.
Simultaneously, instructional courses and longterm technical and methodological support for
teachers are essential. Practical case studies have
highlighted the necessity of long-term programs
facilitating “teaching of the teachers” for successful implementation of ECGs on a broader scale.
12. The exact ways of verifying and validating the
game should be specified early in the design
phase. The evaluation of the game’s learning effect
is crucial. Yet preparation of such evaluation and
its methodology poses some difficulties given the
relative novelty of the ECG field. Typically, such
▶ evaluation should employ qualitative as well as
quantitative measures. In the early phases of evaluation, it is necessary to answer the question: has the
game been accepted by the target audience at all and
have technical problems been eliminated? Ultimately, it is often an advantage to run a controlled
study in which the experimental group playing the
game is compared to one or more control groups
receiving other forms of instruction. Such controlled study must be designed based on hypotheses
the development team aims to support or refute
(e.g., will the knowledge acquired be retained longer
by the experimental group?). A study often requires
various instruments, including educational materials used by the test group instead of the ECG;
pretests, posttests, and delayed posttests and indepth interviews with teachers and learners.

D

933

Important Scientific Research and
Open Questions
The idea of using computer games in education
is a topic that has been discussed at length over the
past 3 decades. A number of theoretical treatises
and anecdotal findings praising ECGs exist in the field
of game-based learning, but the number of sound
empirical studies is relatively limited, and many of
them are plugged with methodological problems
(see e.g., Hays 2005; Gredler 2004; Klopfer 2008; Sisler
and Brom 2008). The empirical findings can perhaps be
best summarized as follows: the games are usually at
least as effective as other instructional methods in
terms of cognitive learning outcomes, and they are
sometimes more motivating, but they can be also
detrimental to learning if they are used in the wrong
way. Because development of an ECG is a costly
process, it is of paramount importance to identify
which features, if any, can make ECGs more effective
than cheaper instructional methods. To a large
degree, potential problems with ECGs can be tackled
by a careful ECG design, preparation of additional
learning materials and supplementary activities, and
meaningful support for teachers. Some unresolved
questions follow:
1. What kind of supplementary activities should take
place in relation to the game? How should these
activities be interconnected to the game?
2. How to promote the ▶ transfer of skills learned via
the game to the real world context?
3. Integration of an ECG into the formal schooling
system is sometimes problematic. What technologies and methods make games most easily usable in
schools?
4. Development of computer graphics is expensive. Is
there a trade-off between the level of graphic
expressiveness and learning outcomes? What is the
“minimal amount of graphics” ensuring acceptable
learning outcomes?
5. In which contexts are competitive principles an
advantage and in which contexts are they
detrimental?
6. Do the answers to the abovementioned questions
differ for different age groups and by gender?
These are not new questions: some of them had
already been asked three decades ago (e.g., Malone
1981). However, the answers to most of the questions

D

934

D

Designing Lessons

are still limited. Yet some points are commonly agreed
upon, most notably:
1. The game’s story should be meaningful to the target
audience. Elements enhancing the motivation of
the students (e.g., an appropriate level of challenge
and control) should be included in the game.
2. The game’s rules should be easily understandable
for the target audience. Oftentimes, the cognitive
load induced by the game has to be reduced in
comparison to mainstream computer games.
3. The goals set out in the game should be known to
the teachers and often also to the students (compare
this with ▶ experiential learning) and should be
visibly relevant to the educational objectives.
4. The game’s tasks should be gradually challenging
enough (to avoid boredom) but not overly hard (to
avoid frustration).
5. The game should not be a stand-alone activity.
Guidance has to be provided to both students and
teachers and the game has to be integrated with
other educational activities.

Salen, K., & Zimmerman, E. (2005). Game design and meaningful
play. In J. Raessens & J. Goldstein (Eds.), Handbook of computer
game studies (pp. 59–79). Cambridge, MA: MIT Press.
Sisler, V., & Brom, C. (2008). Designing an educational game: Case
study of “Europe 2045”. In Z. Pan et al. (Eds.), Transactions of
edutainment (pp. 1–16). Berlin: Springer.

Designing Lessons
▶ Didactics, Didactic Models and Learning

Desire
▶ Goals and Goalsetting: Prevention and Treatment of
Depression
▶ Motivational Variables in Learning

Cross-References
▶ Computer-Based Learning Tools
▶ Design of Learning Environments
▶ Drill and Practice in Learning (and Beyond)
▶ Evaluation of Student Progress in Learning
▶ Experiential Learning Theory
▶ Games-Based Learning
▶ Play and its Role in Learning
▶ Play, Exploration, and Learning
▶ Simulation-Based Learning in Microworlds
▶ Transfer of Learning
▶ Virtual Reality Learning Environments

Desuggestopedia
KAZ HAGIWARA
School of Languages and Linguistics, Griffith
University, Nathan, QLD, Australia

Synonyms
Desuggestopedia; Reserves capacity communicative;
Reservopedia

References

Definition

Gredler, M. E. (2004). Games and simulations and their relationships
to learning. In D. H. Jonassen (Ed.), Handbook of research on
educational communications and technology (pp. 571–581).
Mahwah: Lawrence Erlbaum.
Hays, R. T. (2005). The effectiveness of instructional games:
A literature review and discussion. Technical Report 2005-004.
Naval Air Warfare Center Training Systems Division.
Klopfer, E. (2008). Augmented learning: Research and design of mobile
educational games. Cambridge, MA: MIT Press.
Malone, T. W. (1981). Toward a theory of intrinsically motivating
instruction. Cognitive Science, 4, 333–369.
Reif, F. (2008). Applying cognitive science to education. Thinking and
learning in scientific and other complex domains. Cambridge, MA:
MIT Press.

Suggestopedia is a teaching method that was developed
in Bulgaria from the 1960s to the 1990s. The method
was derived from Suggestology, a medical study of
suggestion in human communications and its role in
the development of personality.
Suggestopedia was developed with the aim of
establishing creative and highly efficient learning in
accordance with the natural learning style of the
brain. In doing so, it attempts to liberate the learners
from limiting social norms that have cumulatively been
created in their personalities by experiencing negative
suggestions in their social life. A very careful approach

Desuggestopedia

is taken to eliminate potential causes of harmful hypnosis. Also, mental hygiene and brain health are well
considered in the method. The method requires
a teacher and a learning group. It does not work in
a self-learning environment.

Theoretical Background
Theoretical development of the method was made at
the National Institute of Suggestology led by Georgi
Lozanov (1926–), medical doctor, professor in psychiatry, psychotherapy, and brain physiology at University
of Sofia. Also, Evelina Gateva (1939–1997), Ph.D. in
education, significantly contributed to the establishment of the current form of the method. The method
was declared final in 1994.
At an early stage of development, the method was
known to the world as an efficient memorization technique with which one can recognize the meanings of
1,000 foreign words in one session. Later, it was developed into more practical forms that are applicable not
only to language teaching, but also to the general curriculum in schools. The effectiveness of the method was
tested and reported to UNESCO in 1978, and
UNESCO also formed its own working party in 1979
to examine the method and produced a final report in
1980 that recommended that the method be widely
applied to areas such as elimination of illiteracy. The
method’s success inspired other countries to develop
their own “accelerated learning” methods.
“Suggestopedia” is a combination of Latin origin
word “suggestion” and Greek origin word “pedagogy.”
Lozanov has called this method by three other names to
distinguish his method from other similar methods.
“Desuggestopedia” has been used since 1994 to indicate
that this method puts an importance on “desuggestive
suggestion.” He has also called it “Reserves Capacity
Communicative (Re-Ca-Co)” since 1998 to indicate
the method is conscious about the potential of
“reserves of mind” and accesses its capacity through
communication. Later in 2006, Lozanov started to call
it “Reservopedia” to imply that his method is a part of
the yet to be established science of “Reservology,” the
study of “reserves of mind.”

Suggestology
In developing Suggestopedia, Lozanov attempted to
apply knowledge of Suggestology to the method of
education. Suggestology consists of (1) the research to

D

discover the nature of suggestion and (2) the construction of a brain function model as a receptor/processor
of the suggestion. Lozanov’s aim was to create a method
with which the brain can work most healthily and
efficiently with the help of properties of suggestion.
As a by-product of psychotherapy, mental hygiene
was always an issue for Suggestopedia. In
Suggestopedia, the promotion of mental hygiene did
not contradict the promotion of brain function efficiency (Lozanov 1978, p. 222).

Nature of Suggestion
Following is a short summary of a part of the outcomes
of Suggestology about the nature of suggestion that is
applied in Suggestopedia.
1. A personality, consciously or unconsciously, constantly and holistically influences, and is influenced
by, general suggestions. Such suggestions can create
a belief in social norms through the experience of
following the social orders and the “commonsense.”
Such beliefs in social norms can limit or inhibit the
person’s ability.
2. Many phenomena that used to be thought of as
unique to those people who are in hypnosis, or
master of ascetical training, or with psychic disorder, or natural genius, can appear in ordinary people by being exposed to general suggestions. Such
phenomena are hypermnesia and a control over the
autonomic functions of the body.
3. Some phenomena that never appear in a hypnotized
person can appear in an ordinary person with
a particular type of general suggestions. Such phenomena are promotion of creativity, inhibition of
suggestibility, lowering of the risk of being hypnotized, and minimizing or removing the influences of
social suggestive norms. Suggestology calls such
types of suggestions “desuggestive suggestions.”
4. The mental state of a person is always changing
from time to time. Also, different types of personality can coexist in a person. The forms of reaction
to the suggestive stimuli are different and unique to
each person.
5. Every person has a mental protection system to
prevent the unwanted influences of suggestions.
The protection system consists of three “suggestive
barriers”: (1) logical (or reasoning) barrier that
eliminates unreasonable suggestions to maintain

935

D

936

D

Desuggestopedia

logical consistency, (2) ethical barrier that eliminates immorality by checking against personal
morals, and (3) affective barrier that intuitively
stops information to maintain emotional stability.
The states and heights of these barriers are diverse
and unique to each person. They are also constantly
changing their forms over time.
6. Nevertheless, there are general tendencies in suggestibility: (1) Suggestions from more prestigious
sources are more acceptable, (2) suggestions given
in more trustful human relationships are more
acceptable, and (3) suggestions given in less aggressive and less defensive communication are more
acceptable.

Brain Model in Suggestology
Following are some of the characteristics of the brain/
mental model in Suggestology:
1. The brain inherently desires to learn and feels happiness when it learns.
2. It is natural for the brain to memorize all information
at once regardless whether the information is given to
the central or the peripheral area of consciousness. It
is also natural that the acquired information is stored
in the brain for a long time in near-perfect form. In
this respect, the general term “good memory” does
simply mean a strong ability to “recall.”
3. Mental activities in the brain occur on two conceptual planes: (1) conscious plane and
(2) paraconscious plane. Both planes coexist in
parallel and constantly exchange information each
other. Conscious mental activity requires support
from a large mass of information stored in the
paraconscious area. Brain activities on the
paraconscious level are more or less automatic,
emotional, and unlimited. When the brain is
demanded to do intensive conscious mental activities without sufficient reserves in the
paraconscious area, the brain becomes frustrated.
4. Information given to a part of the brain is immediately shared by other parts of the brain. It is impossible to stop the information from spreading
around the brain. Therefore, for example, the
brain is not good at separating logic from emotion.
It is good at association.
5. In general, to some extent, the brain likes changes
and surprises. It does not like mechanical

repetitions or very predictable consequences. At
the same time, the brain likes a safe and consistent
environment.
6. The brain tends to create multiple personalities.
Many personalities appear from time to time
in many aspects of the life of a normal healthy
person.
7. The brain as an information processor has an integrated structure of holographic and hierarchic
functions. In such a structure, while each element
of the brain can represent the whole brain system, it
processes certain types of information in certain
ways on demand of the integrated core personality.
In accordance with the above-mentioned brain
model, Suggestopedia established a unique teachinglearning style. First it creates a great reserve of information in the learner’s “paraconscious” area and then it
activates the acquired information in a cheerful environment where the learners enjoy games, songs, music,
arts, changes, surprises, and their different selves. The
teacher guarantees the learners with a safe, protective,
and encouraging environment that allows them to
make mistakes, express themselves, and even hide
behind others. An integrated dynamic balance of emotion, logic, part, and the whole is well maintained
throughout the course.

Elimination of Hypnosis
Suggestopedia tends to be believed as a method that
uses the state of hypnosis, and its techniques and activities are made so that they induce hypnosis.
However, these are misunderstandings. Rather, one
central mission of Suggestopedia from the beginning of
its development is to eliminate all possible elements
that induce hypnosis from all educational activities
(Lozanov 1978, p. 269). Lozanov, a hypnotist himself,
strongly opposes using hypnosis in nonclinical practices. He shares apprehensions with other hypnosis
researchers such as Weitzenhoffer (1989) that hypnosis
not only prevents a person from being spontaneous,
but can also cause serious mental and physiological
disorders even after wakening.
According to Lozanov, the state of hypnosis is
harmful for the personality because:
1. A hypnotized person cannot make a decision.
2. A hypnotized person cannot think critically.
3. A hypnotized person cannot show creativity.

Desuggestopedia

4. Once hypnotized, the same person can be more
easily hypnotized again.
5. Nonclinical hypnosis can include contaminated
suggestions that can later cause a serious psychological and physiological disorder in the person.
For the reasons mentioned above, a genuine
Suggestopedia course NEVER includes such activities
(Lozanov 2009) as:

D

between. One chapter of the course book usually
takes 4–5 days. The duration of the course, the cycle,
and the activity parts can be flexibly decided according
to the nature of the learning group.
An orientation is held prior to the course. In the
orientation, learners are advised not to worry if they
don’t understand the teacher’s talk in the target language, but just to enjoy what is happening. Learners are
also advised that no study at home is necessary.

● Instruction or training for learners to get a state of

Introduction

relaxation
● Visualization or meditation
● Breathing exercises
● Use of biofeedback and similar techniques that aim
●
●
●
●

to obtain a specific brain wave
Use of commanding and ordering expressions
Simple repetition in the task
Slow and monotonous music
Repetitive use of recorded voices

Lozanov believes that such activities can induce the
learners into a harmful state of hypnosis. Lozanov also
points out the possibility that teachers in conventional
education can, without knowing it, induce their students into hypnosis.

Course Structure
A Suggestopedia course has a basic activity cycle of
(1) “introduction,” (2) “concert sessions,” (3) “elaboration,” and (4) “summary.” The cycle is repeated from
one chapter to another, but care is taken not just to
copy and repeat the cycle so that the whole course
forms a dynamic, artistic, up-streaming spiral. All
elements in the cycle are carefully allocated to follow
the theory and the core concept of the method.
Approximate time durations of each part in the
course structure are as follows:
First day
only

Rest of the
course

Introduction:

30–45 min

Active concert session:

60 min

Passive concert session:

30 min

Elaboration, summary,
introduction:

630 min
(14  45 min)

Active concert session:

60 min

Passive concert session:

30 min

A typical day of an intensive course consists of two
90-min sessions with a 30 min break inserted in

The first day introduction is a prelude to the whole
course. It can be considered as a kind of stage performance in which the teacher as a performing artist
involves all learners in his/her communication in
order to quickly soak them in the target language
world (Hagiwara 1993). In this session, the course is
given the key atmosphere in which the learners do not
have to worry about making mistakes, and their creativity is always welcomed. The introduction also gives
learners the direction, the goal, and the reason to learn
the language in implicit, nonverbal, or “desuggestive”
ways. Taking the opportunity to make a strong impression during the first encounter of the course, some
important elements in the target language are introduced to the learners. Introductions to other chapters
are given either within the “elaboration” or as an independent session depending on the structure of each
course book.

Concert Sessions
In the “concert reading session (or concert session),”
the teacher reads the textbook with the selected background music (Lozanov and Gateva 1988) in order to
expose the learners to a mass volume of target language
information. The session consists of two readings,
“active session” and “passive session,” and the same
part is read through in each reading. Whereas music
pieces with relatively high dynamism are selected for
the “active” readings, cheerful, lively but less dynamic
pieces are selected for the “passive” readings. In the
“active reading,” the teacher reads the passage dynamically and slowly so that the intonation harmonizes
with the rhythm and melody of the music pieces that
are selected from symphonies and concertos in the
Classical to Romantic period. In the “passive reading,”
the teacher reads the passage with normal speed and
intonation that is used in daily life with the music

937

D

938

D

Desuggestopedia

selected from the Baroque period. In this way, the
“active reading” can associate words with their meaning. And the “passive reading” can associate words with
the target language’s phonetic system.
The background music is used to salvage learners
from drowsiness and to stimulate their creativity. Also,
the music is expected to have an off-focusing effect that
turns the learner’s attention away from the amount of
vocabulary so that the teacher can send a large volume
of language information to the learner’s paraconscious
area without creating anxiety, a mental block of “suggestive barrier” in the learner’s mind. The main purpose of the “concert session” is to create a mass reserve
of information that is to become a firm foundation for
the learner to later process the new language. Therefore, not all the linguistic information given in the
“concert session” is the teaching target. However, in
Suggestopedia, the reserves given in the “concert session” play a significant role in terms of promoting
health in brain functions. Usually more than 800
unique words (lexicon) are exposed to the learner in
the first day “concert session.”

learning content. This activity is often called “summary.” The summary is often incorporated into the
elaboration part. The teacher asks learners to do the
summary task. However, the task is never enforced, and
the teacher can wait until the learner is ready.

Important Scientific Research and
Open Questions

The “elaboration” starts on the following day of the
concert sessions. “Elaboration” is a series of sessions in
which learners read through the chapter that has been
read in the prior concert sessions. The learners are taught
“four macro language skills” in the various activities in
“elaboration.” Superficially, it looks similar to an ordinary communicative language class that gives learners
such tasks as oral practice, reading comprehension,
grammar introduction, grammar task, games, songs,
role play, and story-telling. However, in the
Suggestopedia, the concept employed by the teacher
when preparing classroom tasks is very different from
conventional language instructions. The teacher of
Suggestopedia prepares the classroom so that the plenty
and diverse information can reach learner’s
“paraconscious area” through “peripheral perceptions.”
The teacher considers the use of focusing/off-focusing
techniques so that the central learning target will not
unnecessarily catch the learner’s attention. Elaboration
often includes the introduction to the next chapter.

Most basic research for Suggestology was conducted in
the 1960s and the 1970s. In particular, intensive experimental research has stopped since 1980 when the
National Research Institute for Suggestology was closed
as a result of political changes in communist Bulgaria.
Suggestology now needs reinterpretation within the
latest terms of such fields of science as modern neurology, brain science, cognitive science, and the study of
semiosis. For example, such key concepts as “reserves of
mind,” “paraconscious,” and “desuggestive suggestion”
may need to be explained in the framework of modern
science with respect to Lozanov’s definition.
Suggestopedia as a philosophy of education can be
applied to other teaching/learning methods.
Suggestopedia is a complete system and each part and
elements are not separately applicable to other
methods. However, it is possible for other methods to
apply the principle of Suggestopedia that is, for example, the use of “desuggestive suggestions” to maintain
learner’s motivation to the subject, or keeping learner’s
mental hygiene by eliminating hypnotic elements from
the classroom and so on. Such application of the
Suggestopedic principle can be explained in the context
of the modern terminology of education, such as
“authenticity,” “learned optimism,” “collaborative
learning,” and “student-centered learning.”
As recommended in the 1980 UNESCO report,
Suggestopedia needs the establishment of a comprehensive system to train its teachers and maintain
teaching quality. A concrete testing system is most
desirable. The testing system should at least be able to
measure proficiency attainment and learning speed.
For example, language proficiency rating systems such
as OPI (US ACTFUL), ISLPR (Australia), and CEFR
(EU) can be explored in association with the principles
of Suggestopedia.

Summary

Cross-References

At the end of the chapter, learners are encouraged to try
a task in which they spontaneously use the chapter’s

▶ Authenticity in Learning Activities and Settings
▶ Cognitive and Affective Learning Strategies

Elaboration

Deutero-learning

▶ Cognitive Learning
▶ Collaborative Learning
▶ Developmental Cognitive Neuroscience and Learning
▶ Music Therapy
▶ Neuropsychology of Learning
▶ Semiotics and Learning
▶ Student-Centered Learning
▶ Superlearning

References
Hagiwara, K. (1993). An invitation to suggestopedia. The Language
Teacher, 17, 7–12.
Lozanov, G. (1978). Suggestology and outlines of suggestopedy.
New York: Gordon and Breach Science Publishers.
Lozanov, G. (2009). Suggestopedia/Reservopedia theory and practice of
the liberating-stimulating pedagogy on the level of the hidden
reserves of the human mind. Sofia: Sofia University Publishing
House.
Lozanov, G., & Gateva, E. (1988). The foreign language teacher’s
suggestopedic manual. New York: Gordon and Breach Science
Publishers.
Weitzenhoffer, A. M. (1989). The practice of hypnotism, Volume I and
II. New York: Wiley.
Further Reading
UNESCO (1980). Suggestopedia, Expert Working Group on Suggestology
and Suggestopedy. 1980. Rapport final. Vol. 2006, UNESCO.
http://unesdoc.unesco.org/ulis/cgi-bin/ulis.pl?database=unesbib&text_p=phrase+words&text=Suggestopedia&ti_p=inc&ti=&au=&
ca_p=inc&ca=&kw_p=inc&kw=&la=&me_p=inc&me=&da_p=
%3D&da=&se_p=inc&se_p=inc&se=&ib=&no=&dt=&tie=AND.
Accessed 06 May 2011.

Detour Problems
Problems in which the subject’s goal is to travel from
point A to point B and a direct route cannot be used.

Deutero-learning
WOLFRAM LUTTERER
Department of Education, University of Freiburg,
Freiburg, Germany

Synonyms
Context of learning; Learning II; Learning to learn;
Set learning; Transfer of learning

D

939

Definition
The word deutero comes from the Greek word deuteros,
which means second, next, or farther from. The term
deutero-learning was coined in 1942 by the angloamerican anthropologist Gregory Bateson (1904–
1980). Bateson distinguishes between two levels of
learning, proto- and deutero-learning. These levels of
learning are simultaneous. The term deutero-learning
describes the context in which (proto-)learning processes occur. You “learn” not only what you are supposed to learn (in a common sense understanding); so,
for example, riding a bike, learning a language, or
repairing a car – these processes are all proto-learning.
At the same time you are learning this, you are also
learning something about the world and something
about how things occur. You develop habits. This is,
at least partly, a result of deutero-learning.
In a later and more sophisticated version of
his learning theory, Bateson distinguishes five levels of
learning, Learning 0 to Learning IV. Deutero-learning
here is regarded as Learning II, and proto-learning as
Learning I. Learning II is formally defined as “change
in the process of Learning I, e.g., a corrective change in
the set of alternatives from which choice is made, or it is
a change in how the sequence of experience
is punctuated” (Bateson, Steps to an Ecology of Mind,
1972/2000, p. 293).
Bateson abstains from offering a concluding list
of all possible aspects of Learning II oder deuterolearning. He is content with describing several aspects
of this level: So adjectives describing character are an
example for Learning II. Behaviors as anxious, passive,
bold or careful are viewed as, at least partly, acquired
by learning. Learning II also describes the punctuation
of human interaction: Nobody is passive or anything
else in a vacuum. Behavior is part of social interaction
and it is an acquired way of punctuating the “stream of
events,” i.e., the interpretation of events as causes or
actions or whatever else. Finally, much of this Learning
II dates from early infancy and it is unconscious.
There are two common misinterpretations of this
theory:
1. Sometimes deutero-learning is viewed as somehow
higher in terms of better than proto-learning. But
this is a misunderstanding. Deutero-learning is in
context to proto-learning. They both are the complementarily parts of the human process of learning.

D

940

D

Deutero-learning

2. Sometimes it is assumed that deutero-learing is
chronologically following to proto-learning, as if
consecutive. But this is also not correct. Both
aspects of learning are synchroneous.
Maybe these misinterpretations are the result of the
fact that this theory is not strictly elaborated in all its
aspects.

Theoretical Background
Bateson’s learning theory was successively developed
over a period of almost three decades and a series of
several articles. The term deutero-learning was coined
in 1942, but it was abandoned in the further development of the theory because of the later differentiation
of five levels of learning instead of the initial two.
However, deutero-learning or Learning II remains
also in the later development of the theory as the core
of the concept (Visser 2003).
The explicit theoretical background of the original
article consists in an application of behavioristic
psychological theory, mixing this with gestalt theory
and the anthropological thinking of Margaret Mead.
Bateson builds his argument especially on Clark Hull’s
Mathematico-Deductive Theory of Rote Learning
(1940). It is shown with several learning curves that
the rate of learning increases in successive experiments.
This brings Bateson to a first step in his theory: the
learning to learn: In a series of similar learning experiments, the learning will get faster. Explicitly discussed
are classical Pavlovian contexts, contexts of instrumental reward or escape, of instrumental avoidance, and of
serial and rote learning. Bateson uses this all for
the exploration of the question, what is learned
about the world in a context created by repetitive
experiences – for example, in the Pavlovian way. So
a “pure” Pavlovian would probably live with a quite
fatalistic worldview. All events would be as if
preordained and he would probably assume not being
able to influence the course of events (cp. Bateson 2000,
p. 173). This is the context where he puts his concept
of deutero-learning in. Further evidence is shown by
the discussion of the apparently different typical
character of Balinese people. Shortly, “habit,” “behavior,” the individual “worldview” and the expectation
how things are done are all, at least partly, results of
deutero-learning.

The first draft of the theory was strongly motivated
by war time and questions of nation building. Bateson’s
question is if it really will be leading to the desired result
of supporting democratic ideals if people are only
manipulated to behave like they are supposed to
behave, instead of strengthening democratic ideals
also in learning contexts. Shortly, you cannot condition
a person to be a wholehearted supporter of democratic
(or other) ideals. The context of learning should be
governed by the ideas structuring society. In contrast
to that time behaviorism Bateson clarifies the significance of insight.
Bateson plays furthermore a certain role in early
constructivism and in cybernetics (Lipset 1982). The
later stage of his theory is largely influenced by the type
theory of Bertrand Russell (1903). With type theory,
Bateson creates the rigorous and formally elaborated
differentiation of the learning levels 0–IV.

Important Scientific Research and
Open Questions
The differentiation of proto- and deutero-learning is
a key concept for several other theoretical approaches.
In Bateson’s work itself, it is essential for his communication theory (1951) and the double-bind theory
(1956), but also for his later work on ecological thinking. Bateson’s theories are the core of the Pragmatics of
Human Communication by Paul Watzlawick (1967),
and also of Neurolinguistig Programming, coined by
John Grinder and Richard Bandler (The Structure of
Magic, 1975). Deutero-learning is also the key to the
Frame-analysis of Erving Goffmann (1974) and to the
concept of habitus by Pierre Bourdieu (La distinction,
1979). It is also used in the organizational theory of
Chris Argyris in his idea of a double-loop learning
(Argyris/Schön, Theory in Practice, 1974), and to psychotherapy (Gron 1993).
There are several open questions. Batesons learning
theory remains somehow isolated in the modern
research. This may be result of the fact that the
experimental examination of deutero-learning effects
is difficult. But, on the other hand, it is possible
that this lack in reception inhibits modern learning
theory from a further insight in the structure, development, and complexity of human worldviews, leaving
them on a formal, still quite simple, level of
exploration.

Developing Cross-cultural Competence

Furthermore, it remains as a problem, that deuterolearning or Learning II seems to be just a collection of
several aspects or contexts of learning. This is theoretically not satisfying. It is unclear if and how further
research can give more insight or if this is
a fundamental problem, which puts the whole theoretical approach in question.
Finally, Bateson abstains from an analysis of the
possible relation of his levels of learning to each other.
So it remains unclear how Learning 0 and Learning II
relate to each other.
Further development of some aspects of Bateson’s
learning theory is given by Lutterer (2011),
reformulating several aspects of this theory, especially
the relation of the several learning aspects to each
other and to the crucial significance of Learning
0 for the whole learning process. But all things
considered, there remain open questions and a task
for further research on Bateson’s learning theory
and its significance for a better understanding
on the development and the structure of human
thinking.

Cross-References
▶ Anthropology of Learning and Cognition
▶ Bateson, Gregory (1904–1980): Anthropology of
Learning
▶ Learning Set Formation and Conceptualization
▶ Social Construction of Learning

References
Bateson, G. (2000): Steps to an Ecology of Mind. Chicago, University of
Chicago Press. (therein esp.: “Social Planning and the Concept of
Deutero-Learning” and “The Logical Categories of Learning and
Communication”).
Gron, P. (1993). Freedom and determinism in Gregory Bateson’s theory
of logical levels of learning: An application to psychotherapy. Boston: Boston University Press.
Lipset, D. (1982). Gregory Bateson: The Legacy of a Scientist. Boston:
Beacon.
Lutterer, W. (2011). Der Prozess des Lernens: Eine Synthese der
Lerntheorien von Jean Piaget und Gregory Bateson. Velbrück Wissenschaft: Weilerswist.
Visser, M. (2003). Gregory Bateson on deutero-learning and double
bind: A brief conceptual history. Journal of the History of the
Behavioral Sciences, 39(3), 269–278.

D

941

Developing Cross-cultural
Competence
TATIANA STEFANENKO1, ALEKSANDRA KUPAVSKAYA2
1
Department of Social Psychology, Moscow State
University, Moscow, Russia
2
LITE College, London, UK

Synonyms
Cultural mentoring; Intercultural sensitivity; Multicultural education

Definition
Developing cross-cultural competence is a process which
involves methods and procedures dedicated to evolving
cross-cultural competence that leads to the adoption and
understanding of the features of one’s own culture,
develops a positive attitude to other cultural groups
and its participants, and increases the ability to understand and interact with them.
To a large extent, all theories and models of
cross-cultural competence rely extensively on four
core components: motivation (emotional attitude
toward another culture, the needs of the participants of
intercultural communication, social norms, self-images,
openness to new information, the ability to control emotions, etc.), knowledge (cultural self-awareness, deep
cultural knowledge, sociolinguistic awareness,
thoughtfulness expectations, perceptions of more
than one point of view, which could occur during
cross-cultural contact, knowledge of alternative interpretations and knowledge of cultural similarities and
differences, understanding situations and behavior
caused by specific rules, etc.), personality traits (cultural
empathy, emotional intelligence, etc.), and skills (tolerance to ambiguity and uncertainty, adaptability in
communication, flexibility in dealing with new cultural
situations and creating new categories, etc.) (Spitzberg
and Changnon 2009). A broader look at the phenomena and at the practical research shows that crosscultural effectiveness is also strongly influenced by
some external conditions (job/academic motivation,
family support and level of adjustment, cultural
distance).

D

942

D

Developing Cross-cultural Competence

Theoretical Background
High interactivity and technological advancement
of the modern world stimulates rapid and efficient
interaction between cultures. Moreover, multicultural
society is based on the idea that many people in today’s
world do not belong strictly to one ethnic group;
they are members of two or more communities, holders
of several cultures that are “overlapping” in various
combinations, which is fundamentally mobile.
Successful socialization in such a world must focus on
the child’s discovery of its complexity and involves
capability of independent choice (up to a change of
identity), increasing the variability of behavior in
different cultural environments and the level of
tolerance for “different” and “unlike” through the
development of cross-cultural competence. Crosscultural competence is the sum of knowledge about
one’s own and other cultures, which can be seen
through attitudes and behavior in order to ensure
effective and appropriate interaction in a variety of
cultural aspects.
Cross-cultural competence begins to develop from
2 to 4 years with the first flashes of identity with the
child’s ethnic group and differentiation from other
groups, i.e., from the moment when the child begins
to speak. The upper age limit of developing of crosscultural competence is hardly definable, because it can
be described as a self-editing process, which reflects an
awareness of one’s own culture, being capable to assess
the cultural position of others and interact efficiently in
different cultures.
The first and most famous attempt of conceptualizing cross-cultural competence as a developmental
phenomenon was made by M. Bennett (Bennett
1993). According to his concept, cross-cultural competence is directly determined and might be measured by
▶ intercultural sensitivity which demands attention to
the subjective experience of the learner. The key to such
sensitivity and related skills in intercultural communication is the way in which learners construe cultural
differences. There are six stages. Earlier stages of the
continuum define the denial of difference, the evaluative defense against difference, and the universalistic
position of minimization of difference. Later stages
define the acceptance of difference, adaptation to difference, and the integration of difference into one’s
world view. This model is highly influential in training
and research.

The other significant developmental model adapts
the concept of culture shock (Ward et al. 2001)
to a stage model of cultural adjustment. The model
suggests that people living abroad or spending substantial time in a different culture are going through
the following stages over time: honeymoon stage,
hostility stage, humorous stage, in-sync stage, ambivalence stage, reentry culture shock, and resocialization
stages. Initially the model was proposed as a U-curve
hypothesis by S. Lysgaard, later expanded by J. R.
Gullahorn and J. E. Gullahorn into a W-curve
model, including three final stages of repatriation.
Even though the model is still popular due to its
simple visualization, results of studies have proved
rather ambivalent.
In modeling the programs of developing crosscultural competence certain factors need to be taken
into consideration, that were accumulated in the model
of intensity factors in intercultural experiences by Paige
(Paige and Goode 2009): cultural differences (the
degree to which the differences between a person’s
own and another culture are directly proportional to
that person’s psychological stress), ethnocentrism (level
of it from both – the person and the new culture itself),
cultural immersion (the more immersed the person is
in another culture, the greater the amount of stress),
cultural isolation (an opportunity to have contact with
one’s own culture decreases stress), language (the more
essential language ability is for functioning in the particular culture, the greater will be stress of the experience), prior intercultural experience (gives an
opportunity to obtain an idea of intercultural communication and adaptation), expectations (awareness is
essential, as positive and unrealistic expectations
about the new culture and high expectation of individuals lead to psychological stress), visibility and invisibility (regarding physical or any other identity), status
(an ability to identify markers and gain a desirable level
in the new culture), and power and control (being
placed in a different culture an individual feels a loss
of power and control over events in comparison with
what they possessed at home).

Important Scientific Research and
Open Questions
Recent research shows that intercultural learning is
significantly enhanced when it is facilitated. The second
set of studies, however, indicates that this type of

Developing Cross-cultural Competence

cultural mentoring is uneven at best and sometimes
even nonexistent (Paige and Goode 2009). The lack of
a unique theory of cross-cultural competence and
a coherent methodological framework, and eclecticism
in the models and practical concepts of successful
intercultural communication show the importance
and need for conceptual development. There is certainly significant room for further study into methodological issues of researching and methods of
developing cross-cultural competence.
Among the questions of components, its structure
and relations, one of the key questions in conceptualization of cross-cultural competence development is the
relationship between communication competence and
cross-cultural competence itself. There are different
opinions on whether cultural differences and their influence on the competence need to be studied first or if
communication competence within a specific culture is
the core issue. According to B. Spitzberg, it is necessary
to establish a culture-independent model of competence, which then could be adapted to different cultures.
However, according to Y. Kim cross-cultural competence
is not simply a communicative competence in a cultural
context; it is above all a comprehensive personal ability
to cope with the effects of individuals belonging to
different cultural groups and the stress associated with
it, so intercultural competence cannot be defined as
competence adapted to different cultures (Stefanenko
and Kupavskaya 2010).
The lack of experimentally proven answers regarding the correlation between communication and crosscultural competence is well balanced by intensive
development of the practical approach which started
in the mid-1970s and continues up to the present day.
Most approaches to developing cross-cultural competence can be classified on the basis of three major
dimensions: (a) the degree to which the method is
experiential versus didactic (the didactic model based
on the assumption that cultural understanding comes
with knowledge of history, traditions, and customs; the
experiential model is based on the assumption that
most people derive knowledge from personal – direct
or simulated – experience); (b) the extent to which it is
culture-general (etic) or culture-specific (emic) (the
culture-specific approach allows people to understand
how to interact with representatives of a particular
culture; the culture-general approach leads to the
realization of the basic psychological phenomena

D

(negative stereotypes, prejudice, etc.) that interfere
with harmonious intercultural communication);
and (c) the field, in which key results need to be
achieved – cognitive, emotional or behavioral (the
cognitive approach focuses on getting information
about cultures and cultural differences; the emotional
approach focuses on the transformation of attitudes,
related to intercultural interaction, feelings toward
the “other”; behavioral approach is designed to
generate skills that will enhance the efficiency of
communication).
The main didactic types of programs for developing
cross-cultural competence are education (the acquisition of knowledge about a new culture mainly by
means of lectures and discussions, reading appropriate
literature and watching movies about general or
culture-specific aspects of the culture), orientation
(providing a broad look at potential problems or
focuses on particular aspects of adapting to a new
environment), and briefing (a quick introduction to
new cultural environments, the basic norms, values,
and beliefs of the other group). Despite its popularity
and common usage, there are strong disadvantages
of didactic methods in developing cross-cultural competence because they (a) assume a passive position of
the student; (b) suggest working with predefined
critical incidents and their solutions, whereas one of
the main skills of cross-cultural competence is an
ability to be able to identify the problem first and
a flexibility in solving it; and (c) orient to an impartial
investigation and information analysis, although in
practice the skills for charged emotional interactions
with people are far more important (Bhawuk and
Brislin 2000).
A simple transfer of knowledge about cultural and
ethnic diversity of the world is not enough, and it is
important for the person to empirically discover similarities and differences from others, which means that
experiential methods – such as cross-cultural training –
have higher potential in developing cross-cultural
competence. Four main stages could be identified in
the process of developing cross-cultural competence by
the means of cross-cultural training. Firstly, awareness
of cultural specificity of human behavior in general
(cultural awareness). Secondly, awareness of specific
features characteristic of the native culture (self –
awareness) as only then the gap between personal
values, beliefs, behavior, and those of others could be

943

D

944

D

Developing Team Schemas

explored. The main abstract categories, which describe
the variety of different cultures, are verbal and nonverbal behavior, communication, cognitive and learning
styles, values, interaction rituals, conflict styles, and
identity development. Thirdly, awareness of the importance of cultural factors in the process of intercultural
interaction (cross-cultural awareness). Fourthly, in
accordance with cross-cultural training practice, the
development of superordinate (the term and concept
was suggested and described in the realistic conflict
theory by M. Sherif) and personality development in
accordance to them.

Development and Learning
(Overview Article)
ANDREY I. PODOLSKIY
Department of Developmental Psychology, Moscow
State University, Moscow, Russia

Synonyms
Growth; Maturation and learning; Ontogenesis

Definition
Cross-References
▶ Competence-Based Learning
▶ Cross-cultural Factors in Learning and Motivation
▶ Cross-cultural Learning Styles
▶ Cross-cultural Studies on Learning and Motivation
▶ Cross-cultural Training
▶ Learning and Training: Activity Approach

References
Bennett, M. J. (1993). Towards ethnorelativism: A developmental
model of intercultural sensitivity. In R. M. Paige (Ed.), Education
for the intercultural experience (pp. 21–71). Yarmouth:
Intercultural.
Bhawuk, D. P. S., & Brislin, R. W. (2000). Cross-cultural training:
A review. Applied Psychology: An International Review, 49(1),
162–191.
Paige, R. M., & Goode, M. L. (2009). Intercultural competence in
international education administration. In D. K. Deardorff (Ed.),
The SAGE handbook of intercultural competence (pp. 333–349).
Thousand Oaks: Sage.
Spitzberg, B. H., & Changnon, G. (2009). Conceptualizing intercultural
competence. In D. K. Deardorff (Ed.), The SAGE handbook of
intercultural competence (pp. 2–52). Thousand Oaks: Sage.
Stefanenko, T. G., & Kupavskaya, A. S. (2010). Ethno-cultural competence as a component of competence in communication. In
Yu. P. Zinchenko & V. F. Petrenko (Eds.), Psychology in Russia:
State of the art. Scientific yearbook (pp. 550–564). Moscow:
Lomonosov Moscow State University/Russian Psychological
Society.
Ward, C., Bochner, S., & Furnham, A. (2001). The psychology of
culture shock. London: Routledge.

Developing Team Schemas
▶ Development of Team Schemas

Development and learning are considered as formally,
historically, and conceptually different terms which,
however, have a lot in common with regard to their
psychological content. Development describes the
growth of humans and other animals throughout the
lifespan. This includes all aspects of human growth,
including physical, emotional, intellectual, social, perceptual, and personality development. The scientific
study of human development seeks to understand and
explain how and why people change throughout life. In
contrast, learning is NOT the result of biological maturation and growth or of temporary effects of internal
and external factors. Often development and learning
are used to mutually exclude the each other.

Theoretical Background
One of the interesting tendencies in contemporary
psychology at the end of the twentieth century was
a reconsideration of the nature of interaction between
learning and developmental processes during childhood and along the broader life span. On the one
hand, it is hardly difficult to note the serious differences
between learning and developmental studies: They differ in their experimental and theoretical goals, methods
and techniques, preferences in the interpretations they
choose, lists of “key persons,” etc. On the other hand,
more and more scholars have been attracting the attention of the scientific community to a necessity to go
into depth while comparing the conceptual content of
the terms under consideration as well as general methodology and methodological practices in the study of
learning and development. One of the central questions
under discussion is the following: What do we mean
when we speak about development and learning? Are we
really speaking about different psychological realities or

Development and Learning (Overview Article)

are we only paying our dues to the old prejudgment?
Perret-Clermont expressed this dilemma in the following way: “Is there a difference between learning and
development? What is that which is learned? Also what
is that what cannot be learned but which, nevertheless,
structures (in an internal or external way) the organization of subjects’ responses at different ages? Can
intellectual development be considered as essentially
of the sum of acquired knowledge, or, is it, on the
contrary, partially (or even totally) independent of
that acquired knowledge?” (Perret-Clermont 1993,
p. 198). According to Siegler (2000), the background
of these claims lies in researchers’ natural reaction to
the longstanding dominance of Piaget’s assumption
that development and learning are fundamentally different processes. “Piaget frequently distinguished
between development, by which he meant the active
construction of knowledge, and learning, by which he
meant the passive formation of associations. Active
developmental processes were of interest; passive learning processes were not. This distinction was valuable in
focusing attention on children’s efforts to make sense of
the world and in exposing hidden assumptions that
had shaped previous research on children’s learning”
(Siegler 2000, p. 26). It is also important that the
process of transition from stage to stage, the crucial
point of the Piaget’s theory of child development, is
necessarily a slow one because it depends so strongly on
the child’s own activity. Piaget also assumed that
the process cannot easily be modified by external
interventions such as instruction.
According to Perret-Clermont, “psychologists have
often used the term development to refer to endogenous transformations that reveal biological maturation
or a supposed comparable psychological maturation
. . . Development has been understood to mean qualitative changes in thinking that would not result directly
from accumulation of information but would signal
the emergence of new stages or new structures – that
is, more powerful equilibration and self-regulation
processes. In contrast, learning in a strict sense or
acquisition of knowledge usually means those processes
that pertain to the reception of information” (PerretClermont 1993, p. 198). As a result of such an understanding, learning is considered to be a central part of
children’s lives. Nevertheless, the study of learning is
a rather peripheral part of the field of cognitive development (Siegler 2000).

D

Since the last third of the twentieth century, a brand
new trend has appeared in Western European and
North American developmental and learning psychology. Developmental psychologists “began to articulate
a view of conceptual development that was less monolithic than Piaget has proposed. Theorists began to
assert that children’ conceptual development was less
dependent on the emergence of general logical structures than Piaget has suggested and more dependent on
the acquisition of insights or skills that are domain,
task, and context specific. They also began to assert that
children’s thought is more responsive to external influence than Piaget has suggested and more dependent on
the sort of social interaction that has been described by
Vygotsky” (Case 1996, p. 2). These psychologists cited
the following findings in support of this new view of
conceptual development: (a) insignificant correlations
between developmental tests that Piaget had claimed
tapped the same underlying general structure, (b) substantial asynchrony in the rate of development of concepts that Piaget had claimed was dependent on the
same underlying structure, (c) the significant effect of
short-term training on logical tasks such as conservation, and (d) the transfer of such training to other tasks
with the same conceptual content but not to other tasks
whose conceptual content was different but which were
supposed to depend on the same underlying operational structure.
In turn, researchers involved in experimental learning studies began to pay more attention to the acquisition of concepts and skills that are important in
children’s lives. Despite differences in these researchers’
theoretical orientations (neo-Piagetian, cultural
contextualist, dynamic systems, and information
processing), modern investigations of children’s learning have yielded strikingly similar results. According to
Siegler, such similarities are especially encouraging
because they suggest that the regularities in children’s
learning are so strong that they remain evident despite
differences in investigators’ preconceptions and the
specifics of tasks, content domains, and populations.
One should not forget that the strict demands of concrete practice have spawned the development of a much
more enriched understanding of the nature of human
learning in applied areas of the science of learning like
instructional design and instructional technology (see
the entries in this volume). Modern cognitively based
instructional design models and theories utilize

945

D

946

D

Development and Learning (Overview Article)

concepts such as “self-directed learning,” “student selfregulation,” “meaningful content,” “cognitive and
metacognitive strategies,” “mental models,” etc. It is
not difficult to see how far removed a model of learning
conceptualized in this way is from a simplistic, nonrepresentational conception of learning which has little
relevance today but which formed the contrast to
a conception of development. “Modern research has
made it clear that learning processes share all of the
complexity, organization, structure, and internal
dynamics once attributed exclusively to development.
If the distinction has become blurred, it is not because
development has been reduced to ‘nothing but’ learning, but rather because we now recognize learning to be
more like development in many fundamental respects”
(Кuhn 1995, p. 138). According to Siegler, “the terms
‘learning’ and ‘development’ are used differently, with
development referring to changes that are more universal within the species, that occur over longer time
periods, and that occur in response to a broader variety
of experiences. At the level of process, however, the two
have a great deal in common” (Siegler 2000, p. 32).
One more issue which highlights the interrelation
between development and learning has to do with the
methodological paradigms researchers use to explore
these processes. Perret-Clermont believes that the difference between learning and development might then
appear to depend not so much on the subject as on the
relative position of the observer and the observed.
“When a researcher wishes to study learning, his or
her subject is an individual who is placed in
a situation that induces learning. Conversely, when he
or she wishes to observe supposed endogenous development, the investigator usually adopts a less interventionist position, where the subject can be considered
without being explicitly provided with particular cognitive objectives to be achieved. The subject, therefore,
does not feel induced to learn but feels observed
without having the observer’s intention clearly defined
for him or her” (Perret-Clermont 1993, pp. 199–200).
However in the reality of learning as well as development,
“we are concerned with processes having an omnipresent interindividual dimension” (Perret-Clermont
1993, p.199). Perret-Clermont believes that the differences in the paradigms used to collect data create the
differences between data collected as learning and data
pertaining to development. “But, in fact, from
a sociocognitive point of view, in both cases the

phenomena under study are the same: How does
a subject behave when facing tasks and expectations?
Of course, the tasks and the social agents differ,
depending on the context (e.g., school, laboratory,
family, street, playing field), but in each case the subject
has to solve problems, maintain relationships, save face,
send messages, understand questions, try to provide
answers, achieve goals, and manage emotive reactions
and desires” (Perret-Clermont 1993, p. 200).
One may distinguish two main sources of findings
and insights on this issue: first, the many thoughtprovoking publications by Western European and
North American neo-Piagetians and post-Piagetians
which appeared from the 1980s to the 2000s, several
of which were reviewed above, and second, theoretical
and empirical studies devoted explicitly to the issue of
the connection between learning and development
based on Vygotsky’s cultural-historical approach.
In formulating his own position, Vygotsky rejects
three established theoretical positions on the relationship between learning and development: (1) the maturational view, which suggests that a child’s individual
level of mental development constitutes a tightly
restrictive prerequisite or precondition for efficient
learning (e.g., Binet and Piaget’s advocacy of
a learning approach which feared “premature instruction”); (2) the view that learning and mental development are synonymous (as in James’ account of
education as the mere acquisition of habits of conduct
or behavioral tendencies); (3) the view which considered mind as a network of generalized rather than specific capabilities (as in Koffka’s Gestalt psychology
approach or the “Classics” training tradition, in
which the mind was assumed to be a homogeneous
muscle that when exercised in a given domain of
knowledge would produce imminently transferable
learning elsewhere) (Vygotsky 1978). It follows that
neither the full independence nor the partitioning of
these constructs brings clarification of their interrelation. The crucial step was the introduction of the concept of the ▶ zone of proximal development (see also
the entry in this volume). Vygotsky argued that it was
necessary to clarify two issues in order to reach a more
adequate view of the relation between learning and development: first, the general relation between learning and
development; and second, the specific features of this
relationship when children reach school age. The concept
of the zone of proximal development, that is., the

Development and Learning (Overview Article)

distance between a child’s actual developmental level as
determined by independent problem solving and the
level of its potential development as determined by
problem solving under adult guidance or in collaboration with more capable peers, is considered by Vygotsky
to be crucial for the consideration of both issues.

Important Scientific Research and
Open Questions
Many studies conducted by both Russian and Western
scholars in the Vygotskian tradition consider interaction between the child and adult as well as between
children in a collaborative learning activity which leads
to development. However, they do not touch upon the
specific mechanisms by means of which learning affects
cognitive development, nor do they answer the question of how such mechanisms might be taken into
account in designing teaching and instruction. “The
exact ways in which instruction participates in development remain to a large extent unclear. For example,
by what means learning within the zone of proximal
development propels children’s development and what
mechanisms underlie the child’s transition to
a qualitatively advanced cognitive functioning still
remain open questions” (Arievitch and Stetsenko
2000, p. 71). As Russian psychologist P. Galperin has
shown repeatedly in many publications, the main reason why it was not possible to operationalize Vygotsky’s
highly heuristic statements was a lack of knowledge
about a complete system of psychological conditions
that underlie the process of learning and enable mental
actions, images, and representations to form with the
desired and prescribed outcomes. According to
Galperin, mental action is a functional structure that
is formed continually throughout an individual’s lifetime. Using mental actions and mental images and
representations, a human being plans, regulates, and
controls his/her performances by means of socially
established patterns, standards, and evaluations. Mental action can and should be considered as the result of
a complex, multimodal transformation of initially
external processes performed by means of certain
tools (Galperin 1989). Galperin therefore defined
a system of conditions that guarantee the achievement
of prescribed, desired properties of action and image.
This system is termed the “system of planned, stage-bystage formation of mental actions” or the PSFMA system for short.

D

The PSFMA system includes four subsystems: (1) the
conditions that ensure adequate motivation for the
subject to master the action; (2) the conditions that
provide the formation of the necessary orientation base
of action; (3) the conditions that support the consecutive transformations of the intermediate forms of
action (materialized, verbal) and the final, endtransformation into the mental plan; and, (4) the conditions for cultivating, or “refining through practice,”
the desired properties of an action. Each subsystem
contains a detailed description of related psychological
conditions, which include the motivation and operational areas of human activity (see also the ▶ Internalization and ▶ Mental activities of learning entries in
this volume). Since the late 1950s, a significant number
of authors (both researchers and practitioners) have
tried to use Galperin’s theory to improve schooling
processes and results. Studies concerned the very different types of schools (primary, secondary, vocational,
special schools). The subjects (learners) were ordinary,
disabled, and gifted children of different ages (from 5 to
18). The specific domains were also very different: writing and arithmetic, native and foreign languages, math,
the sciences and the humanities, drawing, music, physical education. Finally, psychologically heterogeneous
structures were the objects of planned stage-by-stage
formation: separate domain-specific mental actions
with connected concepts and representations, groups
and systems of actions and concepts, and actions which
underlie cognitive as well as metacognitive strategies
and heuristics. As has been convincingly demonstrated
by hundreds of experimental and applied studies, it was
possible to fulfill the whole set of main objectives that
any schooling is aimed at: (a) the guaranteed acquisition
of the educational course by practically all of the
learners is ensured (provided, of course, that they
have the necessary level of preliminary knowledge and
skills) without prolonging (or sometimes – even reducing) the time allocated to it, and practically without any
additional costs; (b) the division between acquisition of
knowledge and its application is minimized or even
disappears; (c) the learners acquire the ability to transfer knowledge and skills to a new situation: not only
knowledge and skills themselves, but also the manner
of acquiring them; (d) the learners get more and more
interested in the very processes of acquiring knowledge
and in knowledge itself. Certainly, Galperin’s approach
to development has raised and in fact continues to raise

947

D

948

D

Development and Learning (Overview Article)

a lot of problems both of theoretical and practical origin;
however, this entry is not the place to analyze them.
In addition to the important results achieved in his
empirical research, Galperin distinguished three basic
types of learning which are characterized by various
levels of developmental potential. The main characteristics of the first type of learning consist in the subjects’
lack of orientation in the essential characteristics and
conditions of the problem situation to be solved.
Galperin believes that despite many differences in past
and contemporary teaching methods used worldwide,
they all share one basic similarity: incompleteness of
the child’s orientation. Arievitch and Stetsenko present
Galperin’s position as follows: “Galperin maintained
that the striking similarity of Vygotsky’s and Piaget’s
conclusion (despite all the differences in their views on
cognitive development) that children come to be able
to form and use genuine concepts only at the age of
10–12 years can be explained not by some inherent
regularities of the child’s mind and its development.
Rather, in Galperin’s view, the similar developmental
trajectory in the child’s conceptual development
described by both Piaget and Vygotsky can be attributed to the profound similarity of instruction that both
scientists indirectly dealt with when studying children’s
development in preschool and school settings and
modeled in their experimental work . . . What makes
the great many various methods of teaching-learning
just different versions of basically the same type of
instruction? Galperin argued that such a unifying
feature is that most instructional methods fail to provide the child with all the necessary tools and conditions for correct orientation in the task and therefore,
for correct performance” (Arievitch and Stetsenko
2000, p. 75).
In the second type of learning the child is provided
with all of the necessary conditions (see above) to
acquire, internalize, and appropriate an action. The
child is provided with the complete system of psychological conditions, a newly formed action undergoes
a set of transformations and finally gets internalized,
thus becoming a part of the child’s mind (Galperin
1982). The area of application of the newly formed
action is limited to a defined subject domain.
In the third type of learning students acquire
a general method of constructing a concrete orientation basis to solve any specific problem in a given
subject domain. A general method of this kind involves

a theoretical analysis of objects, phenomena, or events
in various subject domains. In such analyses, students
learn to distinguish essential characteristics of different
objects and phenomena, to form theoretical concepts
on this basis, and use them in subsequent problem
solving. More specifically, such analysis includes (a)
discriminating between different properties of the
object or phenomenon, (b) establishing the basic unit
of analysis of a particular property, and (c) revealing to
the child the general rules (common to all objects in the
studied area) of how those units are combined to form
concrete phenomena. Learning of this type always takes
the form of active exploration of the subject under the
guidance of a teacher. Instruction based on the third
type of learning was applied by Galperin and his collaborators in a number of experimental programs aimed
at teaching children a variety of different subjects, such as
mathematics, physics, native and foreign languages, history, etc. One of the most effective of these programs for
childhood cognitive development was that designed for
teaching 5 to 6-year-old children elementary mathematics, specifically basic mathematical concepts (Galperin
1982). Within this program children were taught fundamental concepts like those of numbers and units as
well as other related concepts and arithmetic operations. The general result of the program (as presented
by Arievitch and Stetsenko) was the formation of genuine mathematical concepts in children a whole age
period earlier than usual – in 6-year-olds rather than
in 10–12-year-olds. Even more importantly, however,
was that the children’s entire view of things had
changed: The children came to understand that things
cannot be judged by their visual properties alone. This
characteristic of the preschooler’s thinking underlies
the child’s spectacular display of nonconservation in
Piagetian tasks. In the case described, immediate judgment by visual characteristics was replaced by an analytical procedure in which children learned to
discriminate between different properties of objects
and transform a given property into quantities by
using certain measures. Consequently, the children
gained insight into the implicit structure of objects,
where each basic property of an object constitutes
a separate quantity and an object itself (as a whole) is
represented as a constellation of different quantities.
Thus, the children advanced from immediate (naiveegocentric) thinking to thinking mediated by measure
and measurement and thereby set themselves free from

Development and Learning (Overview Article)

the domination of perceptual impression. It came as no
surprise then that in follow-up experiments, the children who were initially identified as consistent
nonconservers in Piagetian tasks (they saw no problem
in immediately concluding that the whole object
changed, e.g., became “smaller,” when just one of its
properties, e.g., height, was transformed) refused to
give an immediate answer to a conservation task after
having mastered the idea of measurement and the
concept of number according to principles of systemic-theoretical instruction. Instead, they would say:
“First let’s measure!” (Galperin 1982). As a result, the
Piagetian phenomena of nonconservation completely
disappeared in those children and the concept of conservation emerged, although this concept had not been
taught to the children by direct means.
Arievitch and Stetsenko comment that from
Galperin’s analysis it follows that it is unproductive to
discuss the role of learning and instruction in childhood cognitive development without referring to the
specific type of instruction that is actually applied in
teaching children: Depending on the type of learning,
its role may be substantially different. Thus, within the
first type of learning and corresponding type of
instruction, the evidence of positive effects of learning
on development remains ambiguous and disputable
(Arievitch and Stetsenko 2000). Learning in traditional
instruction occurs, as a rule, through gradual and
mostly unsystematic (through trial-and-error) selection of successful versions of problem solving with little
transfer or generalization of knowledge and with
a heavy emphasis on rote memorization. The results
of learning then are largely a matter of individual effort
and luck. All this indeed leaves researchers and educators with little choice other than to conclude that the
role of instruction is very limited in cognitive development. However, this appears to be an overgeneralization based on the specifics of just one (traditional)
type of learning and instruction.
In the third type of learning, the character of knowledge itself (genuinely theoretical) and the way it is
presented to the child (in conceptually based analysis)
differs radically from the two other types of learning
distinguished by Galperin. This method provides children with qualitatively new tools (means of mathematical, linguistic, or other kinds of analysis) to deal
conceptually with a wide range of objects and phenomena extending far beyond the area under immediate

D

study. The central property that defines the developmental potential of certain types of instruction (and
therefore their specific role in development) is the
quality of the cognitive tools provided to the child in
the course of instruction to help it to orient itself with
regard to the conditions of the task and perform it
adequately. When the set of such tools is insufficient
for successful performance and based on empirical
concepts, the developmental potential of instruction
is severely limited, and cognitive development is then
contingent on vicissitudes of the child’s individual
experience within and outside of a given instructional
context. In contrast, when the set of cognitive tools
provided to the child is complete and based on theoretical concepts, instruction results in profound developmental progress. In this latter case instruction
generates cognitive development directly (Galperin
1982; Arievitch and Stetsenko 2000).

Conclusion
As the overview in this entry shows, contemporary
scholars understand development and learning as
closely connected and interrelated processes. It is
always necessary to bear in mind the social origin of
both fundamental processes and, accordingly, to take
into account the student’s attitudes and orientations in
learning environments, learning tasks, etc., in all their
complexity. The relationships between development
and learning are not uniform and can only be adequately understood when the particular character of
learning and instruction with a definite developmental
potential is taken into account.
When speaking about development in the current
context we mean “cognitive development,” which
should be quite understandable due to the obvious
priority of cognitive development among the various
areas of the developing mind. However, the interrelations between learning and development also can and
should be explored in such areas as moral, social, and
emotional development.
Another task of high priority is to continue the
search for a field in which the interactions between
development and learning processes are represented
in a most explicit form as well as the search for
a “common language” which might enable scholars to
synthesize facts, evidence, and regularities reached for
each of those two branches of studies separately. This
task may be characterized as necessary but exceedingly

949

D

950

D

Development of Expertise

challenging, as it requires for entirely new conceptual
and explanatory principles to be created. What could
serve as adequate content for such principles, and what
do we need to do to formulate them? – These questions
remain unanswered.

Cross-References
▶ Cultural-Historical Theory of Development
▶ History of the Sciences of Learning
▶ Internalization
▶ Learning Activity
▶ Mental Activities of Learning
▶ Zone of Proximal Development

References
Arievitch, I., & Stetsenko, A. (2000). The quality of cultural tools and
cognitive development: Gal’perin’s perspective and its implications. Human Development, 43, 69–92.
Case, R. (1996). Introduction: Reconceptualizing the nature of children’s conceptual structures and their development in middle
childhood. In R. Case & Y. Okamoto (Eds.), Monographs of the
society for research in child development: Vol. 61. The role of central
conceptual structures in the development of children’s thought
(Serial No. 246, 1–2, pp. 1–26). Chicago: University of Chicago
Press.
Galperin, P. (1982). Intellectual capabilities among older preschool
children: On the problem of training and mental development.
In W. W. Hartup (Ed.), Review of child development research (Vol.
6, pp. 526–546). Chicago: University of Chicago Press.
Galperin, P. (1989). Organization of mental activity and effectiveness
of learning. Journal of Soviet Psychology, 27(3), 65–82.
Кuhn, D. (1995). Microgenctic study of change: What has it told us?
Psychological Science, 6, 133–139.
Perret-Clermont, A.-N. (1993). What is it that develops? Cognition
and Instruction, 11(3/4), 197–205.
Siegler, R. S. (2000). The rebirth of children’s learning. Child Development, 71(1), 26–35.
Vygotsky, L. (1978). Mind in society. Cambridge, MA: Harvard
University Press.

Development of Expertise
FERNAND GOBET
Department of Psychology, Brunel University,
Uxbridge, Middlesex, UK

Synonyms
Acquisition of expertise

Definition
An ▶ expert is a person whose performance in a given
domain is superior to that of the large majority of the
population. Recursively, a super-expert is an expert
whose performance is superior to that of the large
majority of the expert population. The study of the
development of ▶ expertise has been the province of
psychology, but has also attracted interest in other
fields such as biology, education, sociology, and artificial intelligence. While theories based on ▶ talent
have emphasized fixed, innate traits, theories based on
learning and practice have shed light on the path that
developing experts have to travel. They have examined
the types of knowledge that must be acquired,
and the form that their acquisition takes over time
(e.g., Didierjean and Gobet 2008).

Theoretical Background
In 1905, Albert Einstein wrote five articles that revolutionized the world of physics. Arthur Rimbaud stopped
composing poetry at the age of 21. In just 5 years of
writing, he had revolutionized modern literature. In
sport, Martina Hingis dominated female tennis from
1997 to 2002; she was the youngest player to be number
one in the history of tennis. In chess, former world
champion Garry Kasparov obtained the highest ratings
of all times. His strength was such that he beat national
teams, consisting of professional players, in simultaneous games. These examples illustrate super-experts –
extreme cases of expertise. The label “expert” is also
used, more modestly, with individuals such as physicians, PhDs, national champions in sports, and so on.
The fact that the term “expert” applies to such diverse
kinds of people suggests that it might be problematic to
define what exactly an expert is. One could define an expert
as someone who attains performances at the level of an
experienced professional; but to what does “experienced”
refer? Does it refer to the amount of practice devoted to
a domain or even to the number of years spent in the
domain? However, time spent in a domain is a poor predictor of expertise. Think, for example, about those golf
amateurs who have practiced for years but have never
reached a high level of play. Similarly, the use of diplomas
is problematic, because diplomas are based on sociocultural criteria, which are rarely objective measures of relevant performance. For example, diplomas in medicine are
a reflection more of individuals’ ability to study rather
than an ability to diagnose and treat patients successfully.

Development of Expertise

The difficult task of classifying an expert is easier in
some domains where official ratings are available. The
best example is in the game of chess, which for decades
has had a rating (the Elo rating) that precisely and
quantitatively ranks players, from beginners to world
champions. The presence of such a rating system
(updated every few months) explains why so much
expertise research has been made on chess. Unfortunately, such rating systems are rare, even in games and
sports. To mitigate this rarity, researchers (e.g., in physics and medicine) have often used a simple dichotomy:
novice vs. expert. While this is a practical solution, it
loses much information. A further complication is that
“expert” is a label that is sometimes given more for
social reasons than for the skill level of an individual.
The criteria may vary between societies or even within
societies. This renders comparisons very difficult
indeed. In some cases, whether somebody is an expert
or not critically depends on the context: A fortuneteller is considered an expert in some societies but not
in others. In some domains at least, the status of an
expert is clear. A skeptic might doubt the expertise of
a Roger Federer, but playing a game of tennis against
him would swiftly dispel any doubts.
The study of expertise is interesting and important
for the sciences of learning, for several reasons. First,
individuals who are capable of extraordinary performances offer a unique window on human cognition.
Second, and related to the first point, these individuals
can shed light on strategies to push the previously
known limits of human cognition and rationality.
These strategies can be useful to other people, even
nonexperts. Third, studying experts can illuminate
which training methods are efficient and which ones
are not, again with applications for nonexperts and
even for education in general. Finally, a better understanding of the cognitive processes underpinning
expertise and its development may help the development of artificial intelligent systems capable of
performing at a level equal or even higher than the
best human experts.
In psychology, two traditions have dominated the
study of extraordinary performances: one based on the
notion of talent, and the other based on the notion of
expertise (see Table 1). The first tradition (e.g., Eysenck
1995) goes back to the nineteenth century, with Gall’s
phrenology in Germany and the works of Galton in
England. It aims to show that innate talent is necessary

D

951

Development of Expertise. Table 1 Comparison of the
two traditions that have dominated the study of exceptional performances: on the left, the talent approach; on
the right, the expertise approach
Extraordinary performance
Talent

Expertise

Correlational studies

Experimentation and
modeling

Psychology of intelligence

Cognitive psychology

Innate

Acquired

Differences between
novices and experts

Similarities between
novices and experts

Children/adults

Adults

Normal and pathological

Normal

for high levels of performance. For example,
researchers such as Eysenck in England and Jensen in
the USA have suggested that intelligence quotient (IQ)
correlates with the efficiency of elementary perceptual
processes. This approach is characterized by the use of
correlations, the use of data from neurobiology, and
a focus on interindividual differences. The second tradition (e.g., Ericsson et al. 2006) has focused on the
learning mechanisms and the environmental conditions that make possible the development of extraordinary performances. The emphasis is on adults and
“normal” individuals, as opposed to individuals suffering from pathologies (e.g., autistic calculators). Rather
than correlations, this approach uses laboratory experiments where experts are asked to carry out tasks representative of their skills. For example, given the
description of a case, a physician is asked to carry out
a diagnosis and venture a prognosis.
The experiments carried out by the expertise
approach tend to use standard experimental paradigms
in cognitive psychology. The focus is on the similarities
between the performances of individuals of the same
skill level, or even between the cognitive mechanisms
used by individuals at different levels of expertise. Differences between individuals of different skill levels
tend to be explained by differences in practice. Computational modeling is sometimes used to formalize
the processes thought to underpin the development
of expertise. In other words, a computer program
is designed which simulates expert performance.

D

952

D

Development of Expertise

The program is required to produce the same results
as an actual human expert, and therefore needs to
operate using the same cognitive processes as an actual
human. In some cases, the computer models are able to
reproduce the detail of the behavior at different
levels of skill.
Sadly, there have been only few interchanges
between these two traditions, and the rare examples
have tended to highlight the disagreements rather than
the possible commonalties (see, e.g., Howe et al. 1998,
article in Behavioral and Brain Sciences and the ensuing
replies). Some rare authors (e.g., Mackintosh 1998)
occupy an intermediate position on the continuum
ranging from a pure hereditarian position to a pure
environmentalist position.
Given its own research interests, the expertise tradition has had much more to say about the development of extraordinary performances than the talent
approach, which emphasizes the role of traits that are
essentially fixed. Starting with Binet in 1894, expertise
development has been seen as the acquisition of knowledge. De Groot in 1946 described in more details the
ways expert knowledge is organized, and also had the
critical insight that knowledge is closely linked to perception. In 1964, Fitts proposed three stages in the
acquisition of perceptual and motor skills. During the
cognitive phase, rules, procedures, and facts are learned
by instruction, trials and errors, and feedback. During
the associative phase, stimuli are associated with
responses, and chains of responses are built. During
the autonomous phase, behavior becomes self-sufficient
and independent of cognitive control. Practice and
feedback play an essential role in the last two phases.
Building on the work of De Groot, Simon and
Chase (1973) proposed that knowledge is encoded as
small ▶ chunks, fairly simple data structures. They also
suggested that the development of knowledge consists
of acquiring a large number (between 10,000 and
100,000) of these chunks. Some of the chunks are
linked to potentially useful actions. Gobet and Simon
in 1996 combined the idea of a chunk with that of
a schema, proposing that the chunks that are used
often by an expert become templates, more complex
data structures akin to a schema, which consist both of
fixed and variable information. Ericsson and colleagues
have pushed to the extreme the position of the expertise approach, arguing that ▶ deliberate practice is sufficient to attain high levels of expertise.

In line with Chase and Simon’s account, Newell
and Rosenbloom proposed in 1981 that expertise is
the product of the acquisition of a large number of
▶ productions – simple rules consisting of a condition
and an action. Combining computational and mathematical modeling, they showed that, following the
principle of diminishing returns, accretion of chunks
leads to a power law of learning in performance; that is,
there are rapid improvements at the beginning,
followed by increasingly slower improvements
thereafter.

Important Scientific Research and
Open Questions
In spite of considerable research in the last decades, the
field of expertise faces a number of open questions and
challenges. As noted above, defining expertise has
turned out to be tricky, and, in many domains,
a better operationalization of this concept is desirable.
The interaction between environment and the cognitive processes engaged by experts is rarely well understood. At the extreme, it could be argued that studying
experts will tell us little about the cognitive mechanisms underpinning expertise, but much about the
domain itself – the structure of the environment. For
example, will observing biologists reveal anything
about their thinking beyond what could be found in
a biology textbook? Obviously, students of expertise
believe this is the case, but stronger evidence than that
collected so far would be welcome.
While far from being resolved, the debate between
talent and practice, one the many variations of the
great debate between innate and acquired, has gained
in momentum with the rapid advances in neuroscience
and genomics in the last decades. The results seem
to bring support and comfort for both camps.
On the talent side, the developments in genomics
make it hard to doubt the role of genes in individual
differences, including differences between the best individuals in a domain. On the expertise/practice side, the
developments in neuroscience have highlighted
the plasticity of the brain and its remarkable ability to
learn. Given the complexity of these results, together
with the complexity of the results collected in expertise
research itself, it is likely that the only way forward is
to use some form of computational modeling. Finally,
one of the great unknowns in this field is whether
there are stages in the development of expertise,

Development of Expertise and High Performance in Content-Area Learning

as proposed for example by the early work of Bryan and
Harter in 1899 on telegraphy. The presence of power
laws seems to suggest that expertise development is
continuous, but stages keep appearing in theories of
expertise.

Cross-References
▶ Bounded Rationality and Learning
▶ Chunking Mechanisms and Learning
▶ Development of Expertise
▶ Individual Differences in Learning
▶ Learning in Practice
▶ Learning in the CHREST Cognitive Architecture
▶ Schema

D

Definition
Expertise denotes the reproducible superior performance of an individual in a particular professional
domain, e.g., medicine, music, and physics. The development of expertise is based on many years of dedicated and directed content-area learning. This kind of
learning is described as deliberate practice. Usually oldtimers guide newcomers in their deliberate practice,
e.g., teachers and trainers. Practice leads to substantial
adaptations to the task constraints of the domain. Most
prominent are cognitive adaptations like memory
improvement or knowledge acquisition, but physiological and perceptual-motor adaptations occur in many
domains as well. During the development of expertise,
subjects play an increasingly central role in professional
networks.

References
Didierjean, A., & Gobet, F. (2008). Sherlock Holmes – An expert’s
view of expertise. British Journal of Psychology, 99, 109–125.
Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman, R. R.
(2006). The Cambridge handbook of expertise and expert performance. New York: Cambridge University Press.
Eysenck, H. J. (1995). Genius: The natural history of creativity.
New York: Cambridge University Press.
Howe, M. J. A., Davidson, J. W., & Sloboda, J. A. (1998). Innate
talents: Reality or myth? The Behavioral and Brain Sciences, 21,
399–442.
Mackintosh, N. (1998). IQ and human intelligence. Oxford: Oxford
University Press.
Simon, H. A., & Chase, W. G. (1973). Skill in chess. American
Scientist, 61, 393–403.

Development of Expertise and
High Performance in ContentArea Learning
HANS GRUBER1, HELEN JOSSBERGER2
1
Institute of Educational Science, University of
Regensburg, Regensburg, Germany
2
Center for Learning Sciences and Technologies
(CELSTEC), Open Universiteit, Heerlen,
The Netherlands

Synonyms
Acquisition of expertise; Professional learning; Professional performance

953

Theoretical Background
Research about the development of expertise is
a significantly expanding area of interest in psychology,
educational science, sociology, economy, and many
other sciences (Ericsson et al. 2006). Understanding
the nature of high performance is of relevance in
many contexts in daily and professional life. Scientific
analyses focus on (1) characteristics of experts and
(2) learning processes and practice patterns during
the development of expertise.
Experts are persons who, by objective standards and
over time, show reproducible superior performance in
typical activities of a complex domain, i.e.,
a professional content area like medicine or physics.
Experts in any complex domain have intensively practiced for 10 years or more (10-year rule of necessary
preparation), and thus have extensive knowledge at
their disposal. Even the most “talented” individuals
need such a period of preparation to attain outstanding
performance and most experts practice considerably
longer.
Expertise is extremely domain specific and it cannot
be easily transferred to a different domain. It has been
shown that experts do not possess outstanding
domain-independent competences. For example,
excellence in memorizing new information is one of
the most striking characteristics of experts. However,
experts’ memory is not superior to novices’ in general.
De Groot’s (1965) seminal work on chess masters
pioneered the scientific orientation of research on
expertise. The most striking difference between

D

954

D

Development of Expertise and High Performance in Content-Area Learning

grandmasters and weaker players was revealed in
a memory task, in which subjects were presented
chess positions for a few seconds and asked to immediately reconstruct them. The experts’ superior recall
was explained with specific perceptual structures they
held in memory, that were closely related to their
domain-specific knowledge. De Groot’s interpretation
of the findings directed the future research on expertise
from the perspective of information processing theory
and cognitive psychology. The focus on the analysis of
cognitive adaptations during the development of
expertise (perception, memory, knowledge, problem
solving) has since been maintained and also influenced
the empirical research methods. Most studies on expertise use the contrastive approach, in which experts are
compared with novices; sometimes subjects with an
intermediate level of expertise (semi-experts) are investigated, too.

Important Scientific Research and
Open Questions
Memory and Knowledge
Early studies on expertise were characterized by a focus
on cognitive adaptations like knowledge and memory.
Based on the work by De Groot (1965), it was found
that experts remember domain-specific information
faster and more effectively as well as recall domainspecific information more accurately than novices.
This was the case in many different domains. The
experts’ superior recall is explained by specific perceptual structures they hold in memory, which are closely
related to their domain-specific knowledge. In addition, experts are able to store information faster in their
long-term memory and therefore can retrieve and
apply new information more quickly.
The concept of pattern recognition (Chase and
Simon 1973) explains experts’ ability to very quickly
recognize relevant patterns. It is based on the integration
of perception and knowledge through practice. Components of successful pattern recognition are the mechanisms of chunking and skilled memory. The mechanism
proposed in the chunking concept is the integration of
small knowledge units into larger chunks that are labeled
with indices that are subsequently used during recall
instead of separate knowledge units.
Experts’ superior memory performance is closely
related to their knowledge. Experts excel by a great

amount of knowledge and an advantageous knowledge
organization in order to make functional and efficient
use of their knowledge. However, expert knowledge
is much more than factual or declarative domain
knowledge. Though experts have much domain
knowledge available, they excel in flexibility using
this knowledge, modifying it or restraining from applying it within certain contextual circumstances.
Many different types and qualities of knowledge
can be differentiated, each of them with distinguished
functionality. Most important for the acquisition of
expertise is the distinction between declarative knowledge (knowing-what) and proceduralized knowledge
(knowing-how). Proceduralization of knowledge is
one of the basic mechanisms for explaining the acquisition of expertise.

Acting and Problem Solving
Many expert actions are highly automated. With growing expertise, knowledge increasingly becomes
proceduralized. In his ACT∗ model, Anderson (1982)
stated that skill acquisition mainly consists of changing
declarative knowledge through practice into
proceduralized knowledge. This is achieved through
knowledge compilation and rule tuning. Learners first
have to acquire much declarative knowledge, which is
later proceduralized and associated with action
sequences. Then the skill is automatized and tuned
through repeated practice. But experts are able to do
more: The above-mentioned concept of adaptive
expertise refers to an increasing flexibility of actions
based on a growing and refined knowledge base.
Obviously, transformations in the nature of knowledge during practice are at the core of the development
of expertise. The most illustrating example of such
transformations is the process of knowledge encapsulation, which was described by Boshuizen and Schmidt
(1992) in the medical domain. Through professional
activity and experience with patients, declarative
knowledge about diseases is transformed into illness
scripts. These are generalized knowledge structures,
which are based on episodic experiences with real
cases and therefore are most useful in daily work. The
close relation between declarative knowledge and case
information allows a quick reaction in diagnosing
future cases, because these tend to be rather similar to
each other. Only in “emergency” situations, retrieval of
declarative knowledge is effortful and explicit.

Development of Expertise and High Performance in Content-Area Learning

Knowledge transformation processes during the
development of expertise are neither arbitrary processes nor do they occur just as a function of time of
professional work. They require deliberate actions,
reflection in and about practice, and most probably
meta-cognitive awareness.

Deliberate Practice and Case-Based
Learning
Retrospective analyses of learning processes of high
performers in many different domains revealed the
same recurring pattern: Experts differed from other
individuals early in their career, because they practiced
more efficiently, had more committed teachers, and
showed higher achievement demands. Ericsson et al.
(1993) found that experts were more involved in effortful training activities over a long period of time that
solely had the purpose of improving performance.
They proposed to call such activities “deliberate practice.” Deliberate practice is effortful and not necessarily
enjoyable and therefore individuals must be extremely
motivated to persist training. Most people prefer activities that are motivated by inherent enjoyment (play) or
external rewards (work). Therefore, expert teachers are
important to support and stimulate deliberate practice
by offering explicit teaching goals, feedback and opportunities for gradual improvement through repetition
and correction of errors.
There is convincing evidence that deliberate practice is crucial to acquire the relevant knowledge and
tricks-of-the-trade in any complex content area. Deliberate practice focuses – usually with support of
a trainer or skilled teacher – on the identified areas of
“incompetence,” on designed step-by-step practice
units and on monitoring the degree of improvement.
Teachers or trainers play a crucial role as they function
not only as domain experts, but also as teaching
experts. As a domain expert, the teacher provides
knowledge about typical requirements of the domain.
As a teaching expert, the teacher functions as
a personified accumulation of knowledge about appropriate teaching methods for domain-specific contents
and for the development of skills.
There is converging evidence that educational support of deliberate practice in complex content areas is
closely related to case-based learning in (near-to)
authentic learning environments. The encapsulation
models linked the transformation of professional

D

knowledge with the experience gained while being
engaged in case-based reasoning. As learning by experience with cases transforms knowledge structures, the
acquisition of expertise can instructionally be
supported. A possibility to foster reflective application
of knowledge is the presentation of complex learning
environments in which real application situations
occur.
Case-based learning stresses the similarity between
learning situation and application situation and
implies a number of advantages: (a) By dealing with
complex initial case problems, learners get a notion of
the relevance of the learning matter. (b) Authenticity
and situativity of cases enable learners to make experiences in complex episodes of learning. (c) Multiple
perspectives on the same subject matter help to avoid
oversimplifications and to enhance the transferability
of the to-be-learned.

Open Questions
The outlined characteristics of the development of
expertise and high performance in content-area learning suggest three lines of research: (1) understanding
the nature of individual excellence and the learning
processes that lead to such excellence; (2) understanding the historical development of the respective
domain, in particular the development of the social
and cultural contexts in which expert performance is
embedded; and (3) understanding the longitudinal
development of individual expertise, including both
the experts’ cognitive development and their changing
role within professional communities.
So far, most research has focused on the first topic. In
particular, cognitive characteristics of individual excellence (knowledge, memory, problem solving, reflection,
knowledge transformation, practicing) have been studied intensively. Research on the roles of social and cultural contexts has mainly addressed the role of trainers,
coaches, teachers, etc., for the guidance of deliberate
practice; however, other social and cultural influences
have been hardly investigated. There are some interesting
developments in sociocultural approaches to the analysis
of expertise, but there are still many unresolved research
problems. As the acquisition of expertise is based on
deliberate practice and long-term training, socialization
(i.e., social influence through which a person acquires
the culture or subculture of his/her group) can be
expected to have a strong influence on expertise

955

D

956

D

Development of Judgment

development. Hakkarainen et al. (2004) studied
“networked expertise” activities from the perspective
of a community pursuing a certain activity. More
analyses of this kind have to be designed in the future
in order to make the processes of social negotiation
during the development of expertise more transparent.
The development of expertise requires a considerable
period of time. Observations of professional careers in
many different domains led to the conclusion that at
least 10 years of deliberate practice are required in order
to fully develop expertise. It is far from trivial to design
longitudinal research that reliably investigates the
development of expertise. It may be expected that
extensive longitudinal studies would reveal many new
components of the development of expertise in addition
to the cognitive structures and processes which have
been the main focus in cross-sectional research. The
motivational question, why some individuals are willing
to work and deliberately practice for many years,
whereas many others are not, is one of the big puzzles
to be resolved.

Cross-References
▶ Content-Area Learning
▶ Deliberate Practice and Its Role in Expertise
Development
▶ Development of Expertise
▶ Expertise
▶ Long-Term Expertise Development in Complex
Domains and Individual Differences

References
Anderson, J. R. (1982). Acquisition of cognitive skill. Psychological
Review, 89, 369–406.
Boshuizen, H. P. A., & Schmidt, H. G. (1992). On the role of biomedical knowledge in clinical reasoning by experts, intermediates and novices. Cognitive Science, 16, 153–184.
Chase, W. G., & Simon, H. A. (1973). Perception in chess. Cognitive
Psychology, 4, 55–81.
De Groot, A. D. (1965). Thought and choice and chess. The Hague:
Mouton.
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of
deliberate practice in the acquisition of expert performance.
Psychological Review, 100, 363–406.
Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman, R. R. (Eds.).
(2006). Handbook on expertise and expert performance.
Cambridge: Cambridge University Press.
Hakkarainen, K., Palonen, T., Paavola, S., & Lehtinen, E. (2004).
Communities of networked expertise: Educational and professional
perspectives. Amsterdam: Elsevier.

Development of Judgment
▶ Aesthetic Learning

Development of Musical
Experience
▶ Developmental Psychology of Music

Development of Selfconsciousness
ANNA STRASSER
Humboldt-Universität zu Berlin, Center of Integrative
Life Sciences, Berlin, Germany

Synonyms
Core self; Egological self; Minimal self; Narrative self;
Pre-reflexive self-consciousness; Reflexive self-consciousness; Self-knowledge; Self-recognition; Selfreference

Definition
Self-consciousness is a highly debated notion in
philosophical discussions. As such, there is no general
definition to be found. The same is consequently true
for the development of self-consciousness; positions
vary from claims of innateness up to positions
maintaining that the development of selfconsciousnesses is dependent on the development of
higher cognitive abilities.
Self-consciousness is a special type of consciousness
– whether there can be consciousness without selfconsciousness is naturally still under debate. Many
forms of self-consciousness can be differentiated:
A major distinction can be made between fully developed forms and antecedent, not fully developed forms
of self-consciousness. The former is referred to as
reflexive self-consciousness or narrative self. The latter
is often called pre-reflexive self-consciousness or minimal self. Such not fully developed forms can have the

Development of Self-consciousness

functional role to clarify developmental questions and,
more intensively, in philosophy to posit claims about
logical necessities.
Instead of a definition, an example shall give insight
about what is meant by the notion of selfconsciousness:
A person who is able to ascribe a mental state to
herself is self-conscious, e.g., that one is planning to
read this whole entry. This act of self-ascription implies
several abilities and special knowledge. First of all, one
has to have the ability to differentiate between oneself
and others (▶ self-other/world differentiation). Further, one needs the ability to refer to oneself as oneself
consciously (▶ conscious self-reference). This means
that the person has to know that this statement is
about the person she is and not about anyone else.
Further, ascribing mental states to oneself implies the
ability to ascribe mental states to others (▶ theory of
mind). If a person does not know what it means if
another person is making such a statement, it is hard
to explain how she would know what it means in her
case. Additionally, one needs the special knowledge that
one is just one and that one is continuously in time
(▶ synchronic and diachronic identity).
The question about the development of selfconsciousness can be answered by analyzing the necessary and sufficient conditions for the development of
those capacities.

Theoretical Background
Any theory concerning the development of selfconsciousness includes a clarification of what
self-consciousness is meant to be and what the object of
this type of consciousness is. The following overview will
constitute a snapshot of several important debates, and
will address topics like self-knowledge, the privileged firstperson perspective, the use of the first-person pronoun,
and the general ability of self-reference.
Historical debates focus on the idea of something
like a metaphysical substance of a self that was reflected
in the well-known debate about the mind-body
problem. Dualist positions – claiming there is
a nonmaterial, metaphysical self – have become, nowadays, more or less a minority and consequently, metaphysical ideas are not debated that intensely any more.
But still the negotiation of the existence of
a metaphysical self plays an important role concerning
skeptical accounts of self-consciousness. The existence

D

of the phenomenon of self-consciousness itself is
not debated rather what self-consciousness represents,
i.e., the self, can have various definitions: Marvin
Minsky (1990) claims that self consists in many different agents. Daniel Dennett (1991) describes selfconsciousness as a story made of many sketches of
self-interpretations, the center of this story is the “I,”
which is a fiction that can be even developed by computers. Most skeptical is Thomas Metzinger (2004),
according to him self-consciousness is illusory. His
theory about the phenomenal self-model, which
includes many representational features explains how
this illusionary, phenomenal feeling to be in direct
contact to oneself is developed. But, for Metzinger,
there is no self there to be represented. The historical
antecedent of those positions is Hume’s negotiation of
the notion of a substance claiming that there is just “a
together of several perceptions.” This means that the
perceiver himself cannot be perceived.
Many theories of self-consciousness focus on the
condition of conscious self-reference. This includes
the ability to refer to yourself as yourself as well as the
abilities to self-ascribe properties, parts of your body,
mental states, and actions. The main question is: How
is the ability of conscious self-reference possible?
One answer is given by the so-called “Heidelberger
Schule” (Manfred Frank 1991). Claiming that a prereflexive form of self-consciousness is logically necessary to make reflexive self-consciousness possible, the
statement is made that this pre-reflexive form is not
analyzable and seems to be innate.
In the actual debate, one can find the position that
primitive forms of self-consciousness make higher
forms possible; e.g., José L. Bermúdez argues that
nonconceptual representations are used to explain the
development of conceptual representations of oneself.
Another strategy to explain how one can be conscious of one’s own mental states is given by the socalled meta-representation theories (cp. Peter
Carruthers, Daniel Dennett, and David Rosenthal);
these theories introduce a level of higher-order
thoughts and this level is responsible for our ability to
recognize oneself as oneself. Developmental questions
are not treated by those theories.
Analytic theories about self-consciousness tend to
focus on the linguistic ability to use the first-person
pronoun in the right way (cp. Donald Davidson, Gareth Evans, John Perry, Ludwig Wittgenstein).

957

D

958

D

Development of Self-consciousness

Concerning developmental questions those theories
result in the claim that fully developed selfconsciousness can only be ascribed if linguistic abilities
are developed.

Important Scientific Research and
Open Questions
Findings of developmental psychology – just mentioning imitation behavior of newborn babies (cp. Andrew
F. Meltzoff), the so-called rouge test and results of
Theory of Minds studies – are used to illustrate the
development of self-consciousness.
Albert Newen and Kai Vogeley (2003) define,
e.g., five stages of self-consciousness: phenomenal
self-acquaintance, conceptual, sentential, metarepresentational, and iterative meta-representational
self-consciousness. They give a clear description at
what point of time in the human development those
stages can be found:
1. Phenomenal self-acquaintance understood as the
ability to recognize sensory states like pain is
claimed to be there even before birth.
2. The next stage – conceptual self-consciousness –
realized by conceptual representations and necessary for the ability to classify objects appears already
in 8–12 months-old babies.
3. Sentential self-consciousness understood as the
ability to categorize events or complex scenes can
be found in 1–3 years-old children.
4. Meta-representational self-consciousness understood as the ability to construct mental models of
other minds demands the ability to ascribe firstorder propositional attitudes like “Darja thinks that
p.” This ability can be found in 2–4 years-old
children.
5. So-called iterative meta-representational selfconsciousness does not develop before the 7th year;
this implies the ability to ascribe second-order attitudes like “Liz believes that Glenn thinks that p.”
Positions vary in many ways regarding the point of
time when the first ascription of self-consciousness in
the development of human beings can be made: some
hold a very basic notion of self-consciousness – nearby
the above-mentioned notion of pure self-aquaintance.
In this case, some will ascribe self-consciousness at
a very early stage, sometimes even before birth. For
example, it is claimed that self-consciousness in the

means of a minimal self is already there when newborn
babies show imitative behavior.
Other, more demanding notions of selfconsciousness maintain that higher cognitive abilities,
e.g., the above-described conceptual self-consciousness,
develop first. Whether higher cognitive abilities as recognizing something as something demand linguistic
abilities is still under discussion. Analytic theories
claim that a subject is self-conscious only if one uses
the first person pronoun correctly. This precludes the
ascription of self-consciousness to young children without linguistic abilities. In further debates, it is also
discussed whether self-consciousness in a basic form
can be found in animals (animal cognition) and whether
it will be found some days in AI systems.
Most important for developmental questions are
theories of self-consciousness that take intersubjectivity into account. Such theories focus on the mentioned
Theory of Mind abilities. One may even say that selfconsciousness is now being researched as a part of the
broader study of Social Cognition (cp. Shaun Gallagher,
Albert Newen & Gottfried Vosgerau, and Dan Zahavi).
Here, debates concerning social interaction lead to
debates about the development of self-consciousness.
George Herbert Mead (1968) gave the first philosophical theory about the functional role of the other in the
development of self-consciousness. Historically, ideas
about the role of the others can already be found in
Georg Wilhelm Friedrich Hegel, Edmund Husserl, and
Martin Heidegger. In modern times, the role of other
subjects plays an important role in the work of JeanPaul Sartre and Maurice Merleau-Ponty.
From a more philosophical point of view, research
about the development of concepts suggest the idea
that being able to ascribe psychological properties to
oneself demands the ability to ascribe those properties
to others as well (generality claim). This claim can be
seen as an analogy to the debate of Lynn Baker, Gareth
Evans, and Peter F. Strawson, which shows that the
ability to entertain a concept implies the ability to
apply it to more than one case. That means mere selfascription is not enough. What specific kinds of interaction are needed to ascribe psychological properties to
other mental agents still is an open question, which can
be resolved by examining the development of theory of
mind capacities.
Psychiatry with many descriptions of deficits
concerning self-consciousness opens up a new field

Development of Team Schemas

of interdisciplinary research. Cooperation with
psychiatry can enrich philosophical perspectives on
self-consciousness. Psychopathological descriptions of
self-consciousness as deficits of the ability of selfascription offer interesting possibilities to approach
questions like what can go wrong in the development
of self-consciousness. Especially, the debate about
schizophrenia gives several examples of how certain
aspects of self-ascription can go wrong. For example,
the symptom of thought insertion asks for an analysis
explaining why the self-thought thoughts are not selfascribed. Another fruitful field of cooperative research
concerns the deficits one can observe in autistic
patients; these may shed further light on the role of
social interaction for the development of full-fledged
self-consciousness.

Cross-References
▶ Imitation and Social Learning
▶ Mental Models
▶ Naturalistic Epistemology
▶ Philosophy of Learning
▶ Theory of Mind in Animals

D

Development of Team Schemas
SANDRA P. MARSHALL
Department of Psychology, San Diego State University,
San Diego, CA, USA

D
Synonyms
Developing team schemas

Definition
A team schema reflects common knowledge and procedural processes that are relevant to a given situation in
which the team is working. Well-functioning teams will
have a generally consistent understanding of the critical
aspects of the situation and the possible outcomes that
may derive from it. Team schemas are fluid and dynamic,
changing constantly as members of the team make
new contributions to the collective whole. The four
knowledge components – identification, elaboration,
planning, and execution – found in problem-solving
schemas and decision-making schemas are present in
team schemas as well.

References
Dennett, D. C. (1991). Consciousness explained. Boston: Little, Brown.
Frank, M. (1991). Die Unhintergehbarkeit von Individualität:
Reflexionen über Subjekt, Person u. Individuum aus Anlaß ihrer
“postmodernen” Toterklärung. Frankfurt: Suhrkamp.
Mead, G. H. (1968). Geist, Identität und Gesellschaft. Frankfurt a.
M: Suhrkamp.
Metzinger, T. (2004). Being noone. Cambridge: MIT Press.
Minsky, M. (1990). Mentopolis. Stuttgart: Klett-Cotta.
Newen, A., & Vogeley, K. (2003). Self-representation: searching for
a neural signature of self-consciousness. Consciousness and Cognition, 12, 429–543.

Development of Self-Control
▶ Volitional Learning

Development of SelfRegulation
▶ Volitional Learning

959

Theoretical Background
Not surprisingly, research about decision-making
schemas in military settings grew into research about
team decision making, with focus on how schemas are
developed within a team and how to assess the effectiveness of schema-based decision making. Just as decisionmaking schemas for an individual are related to situation
awareness, team schemas are linked to what is known as
“shared situation awareness,” a topic of great interest in
the military (Endsley 2000). This area of research is in its
infancy, with few published studies thus far available.
A basic question about a team schema centers on
how information necessary for the schema is shared.
Each individual in a team has unique experiences,
which contribute to his or her own schemas. But,
a team needs to have common experiences, so that
each team member has in memory the relevant features
that contribute to schema development. Simply having
common experiences, however, is insufficient evidence
of shared situation awareness or of the development of
a team schema that can be used to guide future experiences. Stronger evidence comes from communications among team members.

960

D

Development, Emergence, and Maturation of Memory

When teams work together to solve a problem
or make a decision, they invariably must communicate
with each other to convey changing features of
the situation. Analysis of these communications provides evidence of strong or weak schema development
among the team member. For example, Marshall
(2007) analyzed the communications of teams by
coding each comment according to the type of schema
knowledge reflected (i.e., identification, elaboration,
planning, and execution). Teams who performed
better required fewer communications than teams
who performed poorly, suggesting that these teams
did indeed share a common view whose details did
not need to be discussed. Better performing teams
made more comments reflecting planning and execution knowledge than did poor performing teams.
Moreover, the poorly performing teams made many
more irrelevant comments that were not pertinent
to the task at hand. All teams made a large number of
elaboration comments, indicating the importance of
building and maintaining the shared mental model
that teams hold.
Further examination of elaboration knowledge via
communication analysis shows at least five different
types of statements that strengthen or modify this
schema component (Marshall 2008). Conjectures
represent hypotheses about situation elements and are
indicators of well-developed schemas. They help other
team members understand better what happens
next. Updates add to the collective body of knowledge, keeping everyone informed about changes
that occur in the situation. Acknowledgments are
the most frequent of comments among team members
and are essential to communicate receipt of information. They also imply agreement with the received
information which further strengthens the team
view. Queries and corrections are the final two
types of elaboration knowledge and both indicate
weak or poorly formed schemas. These two forms
of communication are disruptive to team schema
development. A query usually indicates that the
speaker is uncertain or has missed part of the information he or she should already have. The normal flow of
conversation is interrupted by the query. Similarly,
corrections are also disruptive because they require
the other team members to change information
that may already have been incorporated into the
group schema.

Important Scientific Research and
Open Questions
Research on team schemas is an area of some
importance in applied areas such as military decision
making because these decisions are made in realworld settings by groups of individuals working
together. Many basic questions remain to be answered,
such as how schemas can be modified, how schemas
can be developed quickly and efficiently, and how
new team members acquire an existing team schema
when they are integrated into a team.

Cross-References
▶ Schema-Based Problem Solving
▶ Schemas and Decision Making
▶ Shared Cognition

References
Endsley, M. (2000). Theoretical underpinnings of situation awareness: A critical review. In M. Endsley & D. Garland (Eds.),
Situation awareness analysis and measurement (pp. 3–28).
Mahwah, NJ: LEA.
Marshall, S. P. (2007). Measures of attention and cognitive
effort in tactical decision making. In M. Cook, J. Noyes, &
V. Masakowski (Eds.), Decision making in complex
environments (pp. 321–332). Aldershot, Hampshire UK: Ashgate
Publishing.
Marshall, S. (2008). Cognitive models of tactical decision
making. In Proceedings of the Second International Conference
of Applied Human Factors & Ergonomics Las Vegas, July 2008.

Development, Emergence, and
Maturation of Memory
▶ Ontogeny of Memory and Learning

Developmental and Practice
Histories
▶ Expert Perceptual and Decision-Making Skills:
Effects of Structured Activities and Play

Developmental Cognitive Neuroscience and Learning

Developmental Cognitive
Neuroscience and Learning
ANNA MATEJKO, DANIEL ANSARI
Department of Psychology, The University of Western
Ontario, London, Canada

Synonyms
Developmental science; Educational neuroscience;
Neurocognitive development

Definition
Developmental cognitive neuroscience is a multidimensional and interdisciplinary field that attempts
to explain how cognitive development is supported
by changes in underlying brain structure and function, and how brain organization changes over developmental time (Johnson 2011). Developmental
cognitive neuroscience lies at the intersection of multiple
fields including brain imaging, electrophysiology,
neurogenetics, computational modeling of development,
and comparative research with nonhuman primates.
Neuroscience provides a means by which to constrain
our understanding of cognitive development and learning to biologically plausible mechanisms. Developmental
cognitive neuroscience will help determine the neurobiological processes of learning and development, and the
mechanisms that support changes (neuronal plasticity)
in brain function and structure over time.

Theoretical Background
Historical Background of
Developmental Cognitive
Neuroscience
Developmental cognitive neuroscience has emerged
from an intersection between developmental psychology and neuroscience. Though both developmental
psychology and neuroscience have had a protracted
history, developmental cognitive neuroscience did not
become a unified field until the first decade of the
twenty-first century. The field of cognitive development became firmly established after Jean Piaget examined how children think and learn. Even though Hans
Berger first measured “brain waves” in children in 1932
using electroencephalography (Niedermeyer 2005),

D

neural systems were not systematically studied until
much later. In the 1980s, magnetic resonance imaging
(MRI) became a useful tool to examine brain structure.
Noninvasive imaging of the functioning brain became
possible with MRI when the blood oxygen level dependent (BOLD) response was discovered. The BOLD
signal reflects changes of blood oxygenation in the
brain. Functional brain activity in a given brain region
can then be inferred from differences in BOLD signal
between experimental conditions. The first application
of this method to study ontogenetic changes in brain
function was carried out by Casey and colleagues
(1995). These authors conducted the first controlled
functional magnetic resonance imaging (fMRI) study
with children and examined brain regions involved in
verbal working memory. Since that time, research in
developmental cognitive neuroscience has grown exponentially and encompasses many aspects of cognitive
development including perception, language, cognitive
control, number, memory, and social cognition.

Developmental Cognitive
Neuroscience: A Multidimensional and
Interdisciplinary Field
Methods
Some of the primary methods used to understand
learning in the developing brain are electroencephalography (EEG), magnetoencephalography (MEG), positron emission tomography (PET), near-infrared
spectroscopy (NIRS), functional magnetic resonance
imaging (fMRI), diffusion tensor imaging (DTI),
transcranial magnetic stimulation (TMS), computational models of development, and comparative
research with nonhuman primates (see Table 1 for
explanations of imaging and electrophysiological measures). These methods complement each other and
provide additional means by which to better understand child development.

Populations
Developmental cognitive neuroscience examines
developmental populations from infancy to young adulthood and encompasses both typically and atypically
developing populations. Both learning disabilities with
and without a known genetic cause help explain the
biological mechanisms underlying learning and reveal
atypical trajectories of cognitive and brain development.

961

D

962

D

Developmental Cognitive Neuroscience and Learning

Developmental Cognitive Neuroscience and Learning. Table 1 Methods in developmental cognitive neuroscience
Method

What it measures

How it works

Advantages

Disadvantages

Electroencephalography
(EEG) and event-related
potentials (ERPs)

Measures electrical
activity of the brain
that originates from
neural firing

Multiple electrodes
on the scalp
measure small
voltage changes.
EEG waves can be
combined to
determine average
task-dependent
activity across trials

● High temporal
resolution

● Difficult to
determine the
source of the activity

Magnetoencephalography Measures magnetic
(MEG)
fields elicited by
electrical currents
from neural firing

Magnetic fields are
measured on the
scalp using sensors
in a helmet-shaped
apparatus

● Better spatial
● Requires a lot of
resolution than EEG equipment and has
an elaborate set up
● Head and brain
tissue interferes little
with the signal

Positron emission
tomography (PET)

Detects how much of
a particular
compound is being
metabolized. When
the compound breaks
down it emits photons
of light. Brain regions
that are more active
will metabolize more
of the compound

A radioactive
compound is
injected into the
bloodstream.
Regions that are
more active will emit
more photons. The
scanner detects the
emitted photons
and a map of the
brain can then be
reproduced

● Flexible in the
types of tasks that
can be administered
in PET scans

● Uses radioactive
chemical tracers
● Images averaged
over times longer
than processes likely
require

Near-infrared spectroscopy A form of optical
(NIRS)
imaging, NIRS
measures small
changes in the
absorption, scattering,
and bending of
emitted light

Weak light beams
are emitted at the
scalp. The distortion
of the light beams is
then measured. Just
like in fMRI, changes
in blood
oxygenation can be
measured

● Not very sensitive
to motion
● No cumbersome
equipment
● Good alternative
to fMRI with small
children

● Limited to regions
close to the skull
● Limited spatial
resolution

Functional magnetic
resonance imaging (fMRI)

The blood oxygen
level dependent
(BOLD) response is
measured using MRI,
assessing blood
oxygen levels in
different parts of the
brain

● High spatial
resolution
● One subject can be
scanned repeatedly

● Low temporal
resolution
● Tasks limited by
the space in the
scanner

Oxygen is transported
to tissues through
networks of blood
vessels on oxygenated
hemoglobin. When
brain regions are
active they call for
more oxygen,
increasing
oxygenated
hemoglobin and
decreasing
deoxygenated
hemoglobin

Developmental Cognitive Neuroscience and Learning

D

963

Developmental Cognitive Neuroscience and Learning. Table 1 (continued)

Method

What it measures

How it works

Advantages

Disadvantages

Diffusion tensor imaging
(DTI)

Diffusion in white
matter is more
unidirectional than in
gray matter due to
barriers to diffusion
(myelin sheath, cell
membrane, etc.)

An MRI can be
sensitized to the
movement of water.
The more directional
the movement of
water, the higher the
white matter
integrity

● Determines
strength of white
matter connections
● Can detect subtle
abnormalities in
white matter

● Still a new
technique and little
consensus on image
processing

Magnetic fields induce
electrical fields, which
change the
membrane potential
that temporarily
blocks neuronal
activity

A coil is placed over
the scalp and emits
magnetic pulses.
The cortex below
will be rendered
inactive. This is akin
to a temporary
lesion

● Can determine
whether deficits are
due to the
dysfunction of
a region
● Can confirm lesion
studies

● Poor control of the
location of
stimulation
● Only stimulates
regions close to the
surface

Transcranial magnetic
stimulation (TMS)

For example, children with Williams Syndrome have
a genetic abnormality and exhibit atypical trajectories of
cognitive abilities. They have been examined to determine both genetic and environmental factors in learning,
brain function, and cognition (Karmiloff-Smith 1998).
Developmental disabilities such as dyslexia (reading disability) or dyscalculia (disorder of arithmetic learning)
have also been studied to determine if functional
or structural brain deficits are associated with learning
difficulties. The study of developmental disorders is not
only important for clinical and societal reasons, but
may also elucidate the cognitive and neurological processes in typical development (Karmiloff-Smith 1998).

Theories and Frameworks
Several broad frameworks have been proposed to
explain learning and development from a cognitive
neuroscience perspective. Johnson (2011) outlines
three major frameworks for the study of developmental
cognitive neuroscience.
1. Maturational approach: According to this perspective, the emerging ability to perform a cognitive operation is associated with a brain region “coming
online.” The maturational approach is closely related
to nativist perspective, according to which cognitive
functions are thought to be “hard wired” in the brain.
Preexisting skills simply unfold over time rather than
being constructed over the course of learning and
development.

2. Skill learning: Skill learning of cognitive, motor, and
perceptual skills in childhood is hypothesized to be
identical or parallel to skill acquisition in adults. Brain
regions become specialized as a result of experience;
patterns of brain activation for basic skill acquisition
in children would be highly similar to patterns of
complex skill acquisition in adults. This suggests
that there is continuity of learning and skill acquisition from childhood to adulthood (Johnson 2011).
3. Interactive specialization: This so-called neuroconstructivist framework posits that the brain becomes
increasingly specialized over time due to complex
interactions between genes and the environment.
Critically, functional brain specialization occurs
through a process of activity-dependent interactions
between brain regions rather than the shaping of
single brain regions. This perspective emphasizes
developmental shifts in networks of brain regions
underlying developing cognitive functions (Johnson
2011).

Important Scientific Research and
Open Questions
Since the emergence of the field of developmental cognitive neuroscience, several key findings have changed
the way we understand learning and development.
These findings illustrate how neuroscience can constrain and further our understanding of cognitive
development and learning.

D

964

D

Developmental Cognitive Neuroscience and Learning

Brain Development
The brain is highly plastic and amenable to change
throughout development. Animal and human
researches converge to suggest that the brain is a
dynamic structure that changes over time in response
to environmental inputs, such as learning. MRI made it
possible to assess structural brain changes in vivo,
which was previously only measurable through postmortem studies of the animal brain. Further advances
in imaging technology such as DTI have also been
useful in measuring how the brain becomes rewired
and how the strength of connections change over developmental time. The brain is especially plastic in early
childhood. Synaptogenesis refers to the process by
which synaptic density increases from birth until 2–12
months of age (depending on the brain region) at
which point the process of synaptic pruning begins
where frequently used connections are maintained
while
unused
connections
are
eliminated
(Huttenlocher 2002). Synaptic density gradually
decreases to adult levels by 2–3 years of age in visual
processing regions and later in multisensory regions
associated with higher cognitive functions such as language and cognitive control (Huttenlocher 2002).
More generally, there tend to be reductions in gray
matter density and simultaneous increases in cortical
white matter from childhood to adulthood (Giedd
et al. 1999). Indeed, the brain has a protracted period
of structural development and continues to change
until young adulthood. Brain areas develop at different
times; visual and auditory regions develop first,
whereas the frontal lobe has a more extended period
of development. Impulsive behavior in childhood and
adolescence is thought to be a result of an immature
prefrontal cortex. Once these regions become more
developed, impulsive behavior and other aspects of
executive functioning improve (Johnson 2011).
Changes in brain structure, such as cortical thickness,
are also related to measures of intelligence; there is
a negative correlation between cortical thickness and
intelligence in early childhood, but a positive correlation later in development (Shaw et al. 2006).

Specialization of Brain Function
Neural systems tend to become increasingly specialized
over time, suggesting that functional specificity of
brain regions for particular cognitive functions is the
outcome of a developmental trajectory. Younger

children tend to show more diffuse brain activation
for a particular task than both older children and adults
(Casey et al. 1997). Specialized regions begin to appear
early in development. Both number and face processing
are examples of neural functional specificity that continue to become increasingly specialized. Converging
evidence suggests that there are distinct parietal circuits
that become increasingly specialized for number magnitude processing and calculation (Ansari 2008).
A frontoparietal shift is evident from childhood to
adulthood for number processing; activation in the
frontal cortex decreases with age while the parietal
cortex becomes more engaged (Ansari 2008). Face
processing also becomes highly localized over the
course of development. The fusiform face area (FFA)
in the occipital and temporal lobes is activated for face
processing, but adults tend to have a larger area of
activation in FFA than children (Grill-Spector et al.
2008). By mid-childhood, the same regions of the cortex are reliably activated to faces, however, the FFA
continues to become increasingly selective for faces
over time (Grill-Spector et al. 2008). Grill-Spector
et al. (2008) suggest that the specialization of the FFA
may involve sharpening neural tuning to faces, a larger
magnitude of response to faces, or a greater number of
face-selective neurons in the FFA. These streams of
research demonstrate localization of function and
increasing specialization with development and
experience.

Effect of Learning and Training on
Brain Structure and Function
There appear to be sensitive periods of brain development where environmental stimulation is particularly
important for skills to develop normally. Sensitive
periods have been clearly demonstrated in the development of the visual system and for language acquisition. Determining how environmental factors and
learning affect brain structure and function has important educational implications. With the availability of
noninvasive imaging techniques, it is now possible to
determine whether remediation programs or interventions have lasting effects on behavior and on brain
structure and function. There has been some success
in determining changes in brain function in atypically
developing children following intervention. Children
with dyslexia are characterized by reading difficulties
and poor phonological awareness. They tend to show

Developmental Cognitive Neuroscience and Learning

reduced or absent brain activity for phonological
processing in left temporo-parietal and frontal regions
(Gabrieli 2009). Remedial programs demonstrate that
improved readers show more typical patterns of brain
activation and that they compensate for weak performance of reading systems in left hemisphere with
increases in activation in the right hemisphere (Gabrieli
2009). The effects of remediation can be observed for
up to a year, this highlights how the learning brain is
highly plastic and can compensate for weaknesses.

Neural Networks and Development
Brain areas do not function in isolation, but need to
communicate and interact with other regions. These
interactions change over the course of learning and
development (see interactive specialization hypothesis
above). Thus, determining how neural networks
develop is essential to understanding how the brain
works and the mechanisms behind cognitive change.
In a recent study, Fair et al. (2007) demonstrated how
brain networks thought to subserve cognitive control
functions such as inhibition and task switching develop
through segregation (decreased short-range connections) and integration (increased long-range connections). This was observed through age-related changes
in functional connectivity between brain regions in
children, adolescents, and adults. This study illustrates
how cognitive development may not only depend on
changes in brain activity, but also on changes in the
connectivity between regions.

Future Directions
Developmental cognitive neuroscience is a relatively new
field, and there are still many unanswered questions.
First, it is unclear whether structural and functional
changes in the brain are a result of maturation, learning,
or a product of their interactions. Neurocognitive development is a complex process where innumerable factors
play a role. Parsing out the causes of cognitive development will be a formidable task. Second, little research has
focused on training and its effect on behavioral and
brain-related changes in both typically and atypically
developing children. Few educational programs are solidly founded in knowledge of cognitive and brain development and determining the mechanisms of behavioral
and cognitive changes will help direct the establishment
and evaluation of educational programs (Johnson
2011). Third, very little is known about how brain

D

structure and function are related to each other. Recent
advances in imaging technology will make it easier to
determine the relationship between white matter, cortical thickness, gray matter density, and other measures
of brain structure with corresponding functional activity. Fourth, the role of genetics in learning is still
unclear. Research on children with genetic abnormalities and with twins has provided some insight into how
genetics may affect learning, cognition, and brain
development. However, advances in neurogenetics
will likely clarify the mechanisms behind neurological
changes. Fifth, the development of neural networks at
both structural and functional levels of analyses is still
poorly understood. Furthermore, how neural networks
change as a result of learning is largely unknown.
Future research will need to uncover the processes
involved in the development of neural networks and
how learning shapes them. Finally, developmental cognitive neuroscience is increasingly being applied to
educational settings, yet there is still a substantial gap
and a lack of knowledge transfer between labs and
classrooms. There needs to be a better integration
between educational practices and research. We now
know that neural factors are important in learning and
cognitive development, this knowledge will be crucial
in developing targeted programs for children.

Cross-References
▶ Dyscalculia in Young Children: Cognitive and Neurological Bases
▶ Neural Network Assistants for Learning
▶ Neuroeducational Approaches on Learning
▶ Neuropsychology of Learning
▶ Vulnerability for Learning Disorders

References
Ansari, D. (2008). Effects of development and enculturation on
number representation in the brain. Nature Reviews Neuroscience, 9, 278–291.
Casey, B. J., Cohen, J. D., Jezzard, P., Turner, R., Noll, D. C., Trainor,
R. J., Giedd, J., Kaysen, D., Hertz-Pannier, L., & Rapoport, J. L.
(1995). Activation of prefrontal cortex in children during
a nonspatial working memory task with functional MRI.
Neuroimage, 2, 221–229.
Casey, B. J., Trainor, R. J., Orendi, J. L., Schubert, A. B., Nystrom, L. E.,
Giedd, J. N., Castellanos, X., Haxby, J. V., Noll, D. C., Cohen,
J. D., Forman, S. D., Dahl, R. E., & Rapoport, J. L. (1997).
A developmental functional MRI study of prefrontal activation
during performance of a go-no-go task. Journal of Cognitive
Neuroscience, 9, 835–847.

965

D

966

D

Developmental Dyscalculia

Fair, D. A., Dosenbach, N. U. F., Church, J. A., Cohen, A. L.,
Brahmbhatt, S., Miezin, F. M., Barch, D. M., Raichle, M. E.,
Petersen, S. E., & Schlagger, B. E. (2007). Development of distinct
control networks through segregation and integration. PNAS, 104,
13507–13512.
Gabrieli, J. D. E. (2009). Dyslexia: a new synergy between education
and cognitive neuroscience. Science, 325, 280–283.
Giedd, J. N., Blumenthal, J., Jeffries, N. O., Castellanos, F. X., Liu, H.,
Zikdenbos, A., Paus, T., Evans, A. C., & Rapoport, J. L. (1999).
Brain development during childhood and adolescence:
A longitudinal MRI study. Nature, 2, 861–863.
Grill-Spector, K., Golarai, G., & Gabrieli, J. (2008). Developmental
neuroimaging of the human ventral visual cortex. Trends in
Cognitive Neuroscience, 12, 152–162.
Huttenlocher, P. R. (2002). Neural plasticity: the effects of environment
on the development of the cerebral cortex. Cambridge, MA: Harvard University Press.
Johnson, M. H. (2011). Developmental cognitive neuroscience. Malden, MA: Wiley-Blackwell.
Karmiloff-Smith, A. (1998). Development itself is the key to understanding developmental disorders. Trends in Cognitive Neuroscience, 2, 389–398.
Niedermeyer, E. (2005). Maturation of the EEG: development of
waking and sleep patterns. In E. Niedermeyer & F. Lopes Da
Silva (Eds.), Electroencephalography: Basic principles, clinical
applications, and related fields (pp. 209–234). Philadelphia:
Lippincott Williams & Wilkins.
Shaw, P., Greenstain, D., Lerch, J., Clasen, L., Lenroot, R., Gogtay, N.,
Evans, A., Rapoport, J., & Giedd, J. (2006). Intellectual ability
and cortical development in children and adolescents. Nature,
440, 676–679.

Developmental Dyscalculia
▶ Dyscalculia in Young Children: Cognitive and
Neurological Bases

Developmental Language
Disorders
▶ Language-Based Learning Disabilities

Developmental Learning
▶ Learning to Sing Like a Bird: Computational Developmental Mimicry

Developmental Psychology of
Music
CLINT RANDLES1, JULIE DERGES KASTNER2
1
Center for Music Education Research, School of
Music, University of South Florida, Tampa, FL, USA
2
College of Music, Michigan State University, East
Lansing, MI, USA

Synonyms
Development of musical experience; Music learning
over time; Musical maturation

Definition
Developmental psychology of music refers to the study
of the cognitive and generative processes associated
with making music, as these processes occur over time.

Theoretical Background
Theories of Musical Development
Generative Processes. One of the theories in the developmental psychology of music, proposed by Mary Louise
Serafine, based loosely on the work of Piaget, is termed
Generative Processes. For Serafine, what the profession
would call music theory – analyses of the structures of
music – was merely thinking about music, rather than
thinking in music. She found the latter to be a more
valuable path of inquiry when examining musical development. Similar to Gordon’s music learning theory
(described later in this paper), Serafine views musical
development as best articulated in whole-part-whole
progressions. Unlike Gordon, however, Serafine’s theory
has not been as systematically tested as some of the other
theories of the developmental psychology of music.

Symbol System Theories
Another of the theoretical areas in the developmental
psychology of music is based on symbol systems. This
category includes Harvard’s Project Zero and
Bamberger’s Theory of Developmental Cumulation.
Project Zero. The crux of the work of Harvard’s
Project Zero is Howard Gardner’s theory of multiple
intelligences, which posits that intelligence can be
manifested in a number of different ways, including
musically. Another important aspect of this work is the

Developmental Psychology of Music

exploration of the presumably inherent relationship
between musical patterns and the affective life of the
individual. Gardner’s work echoed that of Suzanne
Langer, whose philosophical writings proposed the
idea of multiple symbolic forms. Gardner stressed the
importance that music intelligence should be considered in terms of sound, not visual representations of
sound. However, some researchers who have followed
this line of inquiry have focused more on music intelligence as expressed in terms of musical notation.
Bamberger’s Theory of Developmental Cumulation.
The focus of Bamberger’s theory is that music is “multiple” and “cumulative.” Musical development in this
theory rejects the idea that learning occurs in
a unidirectional path, where older mental structures
are replaced by newer mental structures. Rather, this
theory suggests that multiple dimensions of musical
understanding cumulatively build on each other.

Music Learning Theory
Another theory in the developmental psychology of
music is Edwin Gordon’s Music Learning Theory.
Music Learning Theory is perhaps the most logical of
the theories of musical learning, and perhaps the most
controversial. Within this theory, aptitudes for all of
the dimensions of music making, most notably rhythm
and melody, are possessed by all humans from before
birth. The goal of musical instruction within this theory is to from birth, sequentially build musical skills in
the attempt to raise student ▶ aptitude before children
reach age 9, where their cognitive development
becomes essentially wired for the rest of their lives.
Much research has been conducted to verify Gordon’s
theory in numerous teaching and learning settings.
What makes the theory controversial is the rigidity in
the methods used to build aptitude, the idea that the
focus of music instruction should be to build aptitude,
and the centrality of Western music sound structures in
the practical application of the skill building portion of
the theory.

The Developmental Spiral
The final theory regarding the development of music
learning is the Swanwick and Tillman Developmental
Spiral. Central to this theory is the idea that there
should be hierarchical educational objectives for
music. These categories are: (1) skill acquisition,
(2) recognizing and producing expressive gesture,

D

(3) identifying and displaying the operation of the
norms and deviations (form), and (4) aesthetic
response. An important component of this theory,
articulated by Swanwick, is that all artistic engagements, including music, include the following psychological elements: mastery, imitation, and imaginative
play. This theory is often articulated as a spiral model,
with each objective being revisited over the course of
development.

Relationships Among Theories
There are a number of conceivable connections
between the various theories of the developmental psychology of music learning. Gordon’s Music Learning
Theory could be viewed as a more refined and tested
version of Serafine’s Generative Processes. The developmental spiral seems to account for all of the theories
in scope. Seen in this light, Gordon’s Music Learning
Theory would develop the first two categories of the
Developmental Spiral. The symbol system theories
seem to address the last two categories of the developmental spiral.

Musical Development Through Life
Prenatal. Musical learning begins before birth, as the
fetus interacts with the sounds of her mother’s (1) physical self: heartbeat, breathing, voice, and movements,
and (2) emotional self: marked by the exchange of
biochemicals associated with emotions. Through this
exchange, the fetus begins to process sound at around
16–20 weeks from the moment of conception. At this
point in time, the child’s brain, hearing, balance, orientation, movement, and heart rate develop at a (what
kind of rate?) rate. Just as a child’s language exposure
prepares her for a lifetime of using language, a child’s
musical exposure in utero prepares her for a lifetime of
interaction with music.
Research on prenatal development can be categorized into two categories: scientific-conservative, and
romantic-progressive. Work in the scientificconservative area focuses on being objective, impersonal, and detached with regard to research methods
utilized and the reporting of data. Examples of this sort
of work would be controlled experiments conducted to
measure numerically biological processes in as nonevasive of ways as possible. Conversely, work in the
romantic-progressive area tends to focus on the subjective, the personal, power of associations, and

967

D

968

D

Developmental Psychology of Music

spirituality. Work in this area tends to exaggerate fetal
abilities and project adult qualities onto the fetus as
a way of explaining complex processes.

Development from Birth to Age 5
Infants. Although infants begin life without an ability
to comprehend musical structures, they are equipped
to take in music from the environment through their
excellent abilities to hear and discriminate between
many musical aspects, regardless initially of the music
of their culture. Infants can perceive differences
between pitches of a semitone or less and are especially
sensitive to differences in melodic contour. Melodic
contour is believed to be connected to the pitch
changes in mothers’ infant-directed speech. Rhythmically, infants under 1 year of age may not have a fixed
sense of meter, and thus, they may be able to hear
differences in simple and complex meters found in
many non-western cultures. Initial musical interactions, typically through infant-specific musical genres
like lullabies and play songs, provide a way for parents
to communicate to their nonverbal child, and infants’
ability to aurally discriminate minute musical differences provides a blank slate for them to become acculturated to the music of their culture.
Toddlers and Preschoolers. Young children’s musical
vocalizations become more diverse during toddlerhood, especially when they have opportunities to hear
musical modeling from caregivers. Between 1 and 2years, toddlers begin to explore their voices, create
spontaneous songs, and perform musical phrases with
repeated rhythmic or melodic ideas. Three-year-old
children create longer spontaneous songs, typically
weaving in elements from their musical culture. By
age 4 and 5, young children can describe intended
emotions in their created songs. Overall, this period is
marked with the use of vocalizations during times of
play as young children create their own songs and
chants, adapt existing songs, and create vocal sound
effects.

School Age Development
The musical development of school-aged children
encompasses two main periods of growth. First, musical
development from the age of 5 until 8 to 10 is concerned
with becoming enculturated into the music of their culture and develop basic skills like singing voice and beat
competency. Musical development in children from 10 to

18 is marked by greater sense of personal choice and
expression of identity.
Developments from age 5 to 10. Children from
around age 5 to 10 develop musical perceptions and
skills consciously and unconsciously as they perceive in
the music of their culture. At around age 7, children’s
perception of musical elements becomes more complex,
and it is believed that they can begin to process multiple
elements, like melody and harmony, simultaneously. In
their skill development, much of the research has
focused on singing. Singing voice development progresses from a speech-like chant, to speech-like singing,
to singing with an expanded pitch range and consistent
accuracy. Research is inconclusive as to whether children
sing more accurately while performing songs with lyrics
or on a neutral syllable. By age 6 or 7, children’s vocal
range extends to around an octave. By age 11, the majority of children exhibit singing competency, although
there appears to be a greater number of competent girls
than boys. This may be due to cultural influences that
perceive singing as a “female” practice.
Developments from age 10 to 20. As children age, they
continue to develop more sophisticated musical skills.
Some will begin to play musical instruments through the
public schools, private lessons, or through informal
means. Vocally, boys’ and girls’ voices both change during puberty, although not at consistent rates. Also during
puberty, and even beginning earlier, musical preference
becomes an important part of musical development.
Adolescents begin to look to their peers when making
musical decisions, significantly increase their musical
listening, and start to view music as a component of
their personal and musical identities.

Adult Development
Musical development in adulthood is marked by individuals’ identifying and coming to terms with their
interest in music and perception of musical abilities
within their musical culture. For some adults who do
not perceive of themselves as having much musical skill,
their musical participation may diminish or cease. For
other adults, musical participation involves making
music professionally or as a hobby. These adults call
upon the musical skills and knowledge developed in
earlier years to decide their amount and type of musical
participation. As adults age, changes in physiology and
cognition may make it difficult for them to perform at
the same level as they did earlier in their adulthood.

Developmental Robotics

Important Scientific Research and
Open Questions
A number of interesting questions remain. Wherever
there are theories of learning, there should be empirical
tests of those theories. In exploring musical development, more research needs to be developed to understand the acquisition of aesthetic expression. When does
it first occur? What causes it to occur? How does learning
musical patterns and fundamental elements affect aesthetic expression? In terms of what is known about
musical development across the lifespan, more research
needs to focus on the development of rhythm and how
movement affects this development. Finally, much of the
research here has focused on the development of the
individual. However, musical development occurs
within specific contexts and through interactions with
others. More research needs to be conducted to explore
how musical interaction, specifically through actively
making music with others, affects an individual’s musical development.

Cross-References
▶ Shared Cognition
▶ Situated Cognition
▶ Social Construction of Learning
▶ Social Influence and the Emergence of Cultural
Norms
▶ Social Interaction Dynamics in Supporting Learning
▶ Social Interaction Learning Styles
▶ Social Interactions and Effects on Learning
▶ Social Interactions and Learning
▶ Social Learning
▶ Social Learning Theory

Further Reading
Gembris, H. (2002). The development of musical abilities.
In R. Colwell & C. Richardson (Eds.), The new handbook of
research on music teaching and learning (pp. 3487–508).
New York: Oxford University Press.
Parncutt, R. (2006). Prenatal development. In G. McPherson (Ed.),
Child as musician: A handbook of music development (pp. 1–31).
New York: Oxford University Press.
Runfola, M., & Swanwick, K. (2002). Developmental characteristics
of music learners. In R. Colwell & C. Richardson (Eds.), The new
handbook of research on music teaching and learning
(pp. 373–397). New York: Oxford University Press.
Trehub, S. E. (2006). Infants as musical connoisseurs. In G. E.
McPherson (Ed.), Child as musician: A handbook of

D

969

musical development (pp. 33–49). New York: Oxford University
Press.
Welch, G. F. (2006). Singing and vocal development. In G. E.
McPherson (Ed.), Child as musician: A handbook of musical
development (pp. 311–329). New York: Oxford University Press.

D

Developmental Robotics
PIERRE-YVES OUDEYER
INRIA, Talence, France

Synonyms
Autonomous mental development; Cognitive developmental robotics; Epigenetic robotics; Ontogenetic
robotics

Definition
Developmental robotics is a scientific field which aims at
studying the developmental mechanisms, architectures,
and constraints that allow lifelong and open-ended learning of new skills and new knowledge in embodied
machines. As in human children, learning is expected to
be cumulative and of progressively increasing complexity,
and to result from self-exploration of the world in combination with social interaction. The typical methodological approach consists in starting from theories of human
and animal development elaborated in fields such as
developmental psychology, neuroscience, developmental
and evolutionary biology, and linguistics, then to formalize and implement them in robots, sometimes exploring
extensions or variants of them. The experimentation of
those models in robots allows researchers to confront
them with reality, and as a consequence developmental
robotics also provides feedback and novel hypothesis on
theories of human and animal development.

Theoretical Background
Can a robot learn like a child? Can it learn a variety of
new skills and new knowledge unspecified at design
time and in a partially unknown and changing environment? How can it discover its body and its relationships with the physical and social environment? How
can its cognitive capacities continuously develop without the intervention of an engineer once it is “out of the
factory”? What can it learn through natural social

970

D

Developmental Robotics

interactions with humans? These are the questions at
the center of developmental robotics. Alan Turing, as
well as a number of other pioneers of cybernetics,
already formulated those questions and the general
approach in 1950 (Turing 1950), but it is only since
the end of the twentieth century that they began to be
investigated systematically (Lungarella et al. 2003;
Weng et al. 2001; Asada et al. 2009; Oudeyer 2010).
Because the concept of adaptive intelligent machine is
central to developmental robotics, is has relationships
with fields such as artificial intelligence, machine learning, cognitive robotics, or computational neuroscience.
Yet, while it may reuse some of the techniques elaborated
in these fields, it differs from them from many perspectives. It differs from classical artificial intelligence because
it does not assume the capability of advanced symbolic
reasoning and focuses on embodied and situated sensorimotor and social skills rather than on abstract symbolic
problems. It differs from traditional machine learning
because it targets task-independent self-determined
learning rather than task-specific inference over “spoon
fed human-edited sensori data” (Weng et al. 2001). It
differs from cognitive robotics because it focuses on the
processes that allow the formation of cognitive capabilities rather than these capabilities themselves. It differs
from computational neuroscience because it focuses on
functional modeling of integrated architectures of
development and learning. More generally, developmental robotics is uniquely characterized by the following three features:
1. It targets task-independant architectures and learning mechanisms, i.e., the machine/robot has to be
able to learn new tasks that are unknown to the
engineer.
2. It emphasizes open-ended development and lifelong
learning, i.e., the capacity of an organism to acquire
continuously novel skills. This should not be understood as a capacity for learning “anything” or even
“everything,” but just that the set of skills that is
acquired can be infinitely extended at least in some
(not all) directions.
3. The complexity of acquired knowledge and skills
shall increase (and the increase be controlled)
progressively.
Developmental robotics emerged at the crossroads
of several research communities including embodied
artificial intelligence, enactive and dynamical systems

cognitive science, and connectionism. Starting from
the essential idea that learning and development happen as the self-organized result of the dynamical interactions among brains, bodies and their physical and
social environment, and trying to understand how this
self-organization can be harnessed to provide taskindependant lifelong learning of skills of increasing
complexity, developmental robotics strongly interacts
with fields such as developmental psychology, developmental and cognitive neuroscience, developmental
biology (embryology), evolutionary biology, and cognitive linguistics. As many of the theories coming from
these sciences are verbal and/or descriptive, this implies
a crucial formalization and computational modeling
activity in developmental robotics. These computational models are then not only used as ways to explore
how to build more versatile and adaptive machines, but
also as a way to evaluate their coherence and possibly
explore alternative explanations for understanding biological development (Oudeyer 2010).

Important Scientific Research and
Open Questions
Main Research Directions
Research in developmental robotics can be described as
organized along three main axes: the domains of skills
that shall be learnt by developmental robots, the mechanisms and constraints that allow for developmental
learning, and the degree to which these mechanisms
and constraints are made bio-mimetic or only functionally inspired.

Skill Domains
Due to the general approach and methodology, developmental robotics projects typically focus on having
robots develop the same types of skills as human
infants. A first category that is importantly being investigated is the acquisition of sensorimotor skills. These
include the discovery of one’s own body, including its
structure and dynamics such as hand–eye coordination, locomotion, and interaction with objects as well
as tool use, with a particular focus on the discovery and
learning of affordances. A second category of skills
targeted by developmental robots are social and linguistic skills: the acquisition of simple social behavioral
games such as turn-taking, coordinated interaction,
lexicons, syntax and grammar, and the grounding of

Developmental Robotics

these linguistic skills into sensorimotor skills (sometimes referred as symbol grounding). In parallel, the
acquisition of associated cognitive skills are being
investigated such as the emergence of the self/non-self
distinction, the development of attentional capabilities,
of categorization systems and higher-level representations of affordances or social constructs, of the emergence of values, empathy, or theories of mind.

Mechanisms and Constraints
The sensorimotor and social spaces in which humans
and robot live are so large and complex that only
a small part of potentially learnable skills can actually
be explored and learnt within a lifetime. Thus, mechanisms and constraints are necessary to guide developmental organisms in their development and control of
the growth of complexity. There are several important
families of these guiding mechanisms and constraints
which are studied in developmental robotics, all
inspired by human development:
1. Motivational systems, generating internal reward
signals that drive exploration and learning, which
can be of two main types:
(a) Extrinsic motivations push robots/organisms
to maintain basic specific internal properties
such as food and water level, physical integrity,
or light (for e.g., in phototropic systems);
(b) Intrinsic motivations push robot to search for
novelty, challenge, compression or learning
progress per se, thus generating what is sometimes called curiosity-driven learning and
exploration, or alternatively active learning
and exploration;
2. Social guidance: as humans learn a lot by interacting
with their peers, developmental robotics investigates mechanisms which can allow robots to participate in human-like social interaction. By
perceiving and interpreting social cues, this may
allow robots both to learn from humans (through
diverse means such as imitation, emulation, stimulus enhancement, demonstration, etc. . . .) and to
trigger natural human pedagogy. Thus, social
acceptance of developmental robots is also
investigated;
3. Statistical inference biases and cumulative knowledge/skill reuse: biases characterizing both representations/encodings and inference mechanisms can

D

typically allow considerable improvement of the
efficiency of learning and are thus studied. Related
to this, mechanisms allowing to infer new knowledge and acquire new skills by reusing previously
learnt structures is also an essential field of study;
4. The properties of embodiment, including geometry,
materials, or innate motor primitives/synergies
often encoded as dynamical systems, can considerably simplify the acquisition of sensorimotor or
social skills, and is sometimes referred as morphological computation. The interaction of these constraints with other constraints is an important axis
of investigation;
5. Maturational constraints: In human infants, both
the body and the neural system grow progressively,
rather than being full-fledged already at birth. This
implies, for example, that new degrees of freedom,
as well as increases of the volume and resolution of
available sensorimotor signals, may appear as learning and development unfold. Transposing these
mechanisms in developmental robots, and understanding how it may hinder or on the contrary ease
the acquisition of novel complex skills is a central
question in developmental robotics.

From Bio-mimetic Development to
Functional Inspiration
While most developmental robotics projects strongly
interact with theories of animal and human development, the degrees of similarities and inspiration
between identified biological mechanisms and their
counterpart in robots, as well as the abstraction levels
of modeling, may vary a lot. While some projects aim at
modeling precisely both the function and biological
implementation (neural or morphological models),
such as in neurorobotics, some other projects only
focus on functional modeling of the mechanisms and
constraints described above, and might for example
reuse in their architectures techniques coming from
applied mathematics or engineering fields.

Open Questions
As developmental robotics is a relatively novel research
field and at the same time very ambitious, many fundamental open challenges remain to be solved.
First of all, existing techniques are far from allowing
real-world high-dimensional robots to learn an openended repertoire of increasingly complex skills over

971

D

972

D

Developmental Science

a lifetime period. High-dimensional continuous sensorimotor spaces are a major obstacle to be solved. Lifelong cumulative learning is another one. Actually, no
experiments lasting more than a few days have been set
up so far, which contrasts severely with the time period
needed by human infants to learn basic sensorimotor
skills while equipped with brains and morphologies
which are tremendously more powerful than existing
computational mechanisms.
Among the strategies to explore in order to progress
toward this target, the interaction between the mechanisms and constraints described in the previous section
shall be investigated more systematically. Indeed, they
have so far mainly been studied in isolation. For example, the interaction of intrinsically motivated learning
and socially guided learning, possibly constrained by
maturation, is an essential issue to be investigated.
Another important challenge is to allow robots to
perceive, interpret, and leverage the diversity of multimodal social cues provided by non-engineer humans
during human-robot interaction. These capacities are
so far mostly too limited to allow efficient general
purpose teaching from humans.
A fundamental scientific issue to be understood and
resolved, which applies equally to human development,
is how compositionality, functional hierarchies, primitives, and modularity, at all levels of sensorimotor and
social structures, can be formed and leveraged during
development. This is deeply linked with the problem of
the emergence of symbols, sometimes referred as the
“symbol grounding problem” when it comes to language acquisition. Actually, the very existence and need
for symbols in the brain is actively questioned, and
alternative concepts, still allowing for compositionality
and functional hierarchies are being investigated.
During biological epigenesis, morphology is not
fixed but rather develops in constant interaction with
the development of sensorimotor and social skills. The
development of morphology poses obvious practical
problems with robots, but it may be a crucial mechanism that should be further explored, at least in simulation, such as in morphogenetic robotics.
Similarly, in biology, developmental mechanisms
(operating at the ontogenetic time scale) strongly interact with evolutionary mechanisms (operating at the
phylogenetic timescale) as shown in the flourishing
“evo-devo” scientific literature (Müller 2007). However,
the interaction of those mechanisms in artificial

organisms, developmental robots in particular, is still
vastly understudied. The interaction of evolutionary
mechanisms, unfolding morphologies, and developing
sensorimotor and social skills will thus be a highly stimulating topic for the future of developmental robotics.

Cross-References
▶ Active Learning
▶ Affordances
▶ Artificial Learning and Machine Learning
▶ Cognitive Artifacts and Developmental Learning in
a Humanoid Robot
▶ Cognitive Robotics
▶ Curiosity and Exploration
▶ Development and Learning
▶ Human–Robot Interaction
▶ Imitation Learning of Robot
▶ Learning Algorithms
▶ Motor Schemas in Robot Learning
▶ Play, Exploration, and Learning
▶ Robot Learning
▶ Robot Learning from Demonstration
▶ Robot Learning via Human–Robot Interaction
▶ Robot Learning Via Human–Robot Interaction:
The Future of Computer Programming

References
Asada, M., Hosoda, K., Kuniyoshi, Y., Ishiguro, H., Inui, T.,
Yoshikawa, Y., Ogino, M., & Yoshida, C. (2009). Cognitive developmental robotics: A survey. IEEE Transactions on Autonomous
Mental Development, 1(1), 12–34.
Lungarella, M., Metta, G., Pfeifer, R., & Sandini, G. (2003). Developmental robotics: A survey. Connection Science, 15, 151–190.
Müller, G. B. (2007). Evo-devo: Extending the evolutionary synthesis.
Nature Reviews Genetics, 8, 943–949.
Oudeyer, P.-Y. (2010). On the impact of robotics in behavioral and
cognitive sciences: From insect navigation to human cognitive
development. IEEE Transactions on Autonomous Mental Development, 2(1), 2–16.
Turing, A. M. (1950). Computing machinery and intelligence. Mind,
LIX(236), 433–460.
Weng, J., McClelland, J., Pentland, A., Sporns, O., Stockman, I., Sur,
M., & Thelen, E. (2001). Autonomous mental development by
robots and animals. Science, 291, 599–600.

Developmental Science
▶ Developmental
Learning

Cognitive

Neuroscience

and

Dewey, John (1858–1952)

Developmental Teaching
A school education approach based on the ideas
of cultural-historical psychology and, in particular,
on the theory of learning activity (D. Elkonin, V.
Davydov). The very name recalls the famous statement
by L. Vygotsky about the leading role of teaching in
mental development.

Developments
▶ Trajectories of Participation - Temporality and
Learning

Dewey, John (1858–1952)
MICHAEL JACKSON
Department of Government and International
Relations, University of Sydney, Sydney, NSW,
Australia

Life Dates
John Dewey (October 20, 1859–June 1, 1952) was born
in Burlington Vermont to a family of modest means.
He graduated from the University of Vermont. He
taught in a high school for 3 years before attending
Johns Hopkins University where he obtained a Ph.D.,
with a study of Immanuel Kant. He joined the University of Chicago in its earliest days, and later served at
the University of Michigan. In 1904, he went to Columbia University and its Teachers College, where he
remained. His bibliography runs to more than twenty
monographs. His collected works total 37 volumes. He
was also actively engaged with many causes of his times
and wrote in aid of them, too. As befits one of the
founders of pragmatism, he had the reputation of
a hard bargainer in money matters.

Theoretical Background
While his published works range over topics as diverse
as art, democracy, logic, ethics, nature, they always
come back to education. Inspired by Georg Hegel, he

D

sought synthesis. He is often linked to William James
and Charles Sanders Pierce as a founder of pragmatism,
though Dewey used the term “instrumentalism” to
describe his epistemology. More than half a century
before post modernism, Dewey set out a philosophy
without a privileged foundation.
He is likewise credited with being one the progenitors of functional psychology and served as president
of the American Psychological Association in its early
days. One of the tenets of functional psychology was
the principle that the environment affects the mind.
The social context matters in understanding a mind,
and that is the path to understanding the mind.
Some of his most widely read books were My Pedagogic Creed (1897), The Child and Curriculum (1902),
Democracy and Education (1916), and Experience and
Education (1938). He contributed articles to all manner
of popular and scholarly magazines and journals and
had a high public profile throughout his life, making
him a public intellectual in today’s terms. He always
considered himself a philosopher more than
a psychologist or an educator. The Journal of Philosophy
was founded for the purpose of discussing his work. It
remains one of the most important of journals for
professional philosophers.
All through his work, popular and professional, the
major concern was practical, how should we live? To
answer that question, he studied children, quickly concluding that children shape their environment as readily as it shapes them. They were not passive pieces of
white paper upon which to write. No doubt, observation of his own children paved the way for this conclusion. He always stressed the activity of the learner in
grasping, in making knowledge, rejecting what he
called the spectator approach to teaching which
makes the learner into an audience for the possessor
of knowledge – the teacher. Were he to observe current
practice in university teaching, he would see many
presentations both in the class room and on Web sites
that make students into an audience.
The fundament of his philosophy was that everything had to be tested against reality. Human history
and social institution embodied the knowledge gained
from that process of evaluation. Mistakes will be made,
yes, but over time, they will be revealed by further
testing. This empirical, experimental attitude is what
he saw in democracy. Democracy was a mentality, not
a set of rules about voting, to Dewey. In contemporary

973

D

974

D

Diagnosis of Asperger’s Syndrome

terms his focus was on civil society rather than government. He attributed an important role to the democratic citizenry in directing government and this
brought him into conflict with Walter Lippmann and
others who embraced a plebiscitary model of democracy. Elections passed judgment on what had been done
in the name of electorate; whereas Dewey saw the
democratic process as the means to select which programs, practices, and policies would be pursued.
Dewey played a vital role in the creation of the
Laboratory School at the University of Chicago in
1896, which school continues to this day. In it the effect
of teaching methods was observed, evaluated, and
reported. His book School and Society (1900) is one
such report. He was also one of the founders of the
New School of Social Research in New York City in
1919, which school had enormous influence on the
development of the social sciences after World War II.
His influence, gained both through his books and articles and his extensive service on public committees, led
to a role in the foundation of Bennington College and,
thereafter, he served on its Board of Trustees. But perhaps his greatest and most lasting achievement was
during his years at Columbia, where he helped make
teaching a profession with its own discipline and body
of knowledge. He advocated experiential learning,
guided by an educated and trained teacher. Such contemporary practices as problem-based learning and its
kin are sometimes traced back to Dewey.
His willingness to embrace good causes saw him
direct the so-called Dewey Commission in Mexico City
in 1937, which exonerated Leon Trotsky of Stalin’s
many charges. He championed the rights of women
in both word and deed, writing many pieces on the
subject and marching in parades.
Melville Dewey created the Dewey Decimal System
for library shelving and he was no relation to John
Dewey, though they were contemporaries.

Contribution(s) to the Field of
Learning
While a major figure in his own times with considerable influence in succeeding generations, his books
have not aged well. The intelligence and good will of
Dewey shines through the pages of these books, but
they are discursive and obscure. They are seldom read
outside the history of education. Despite his role in the
institutionalization of psychology in the United States,

his intellectual impact on psychology today is not
apparent. On the other hand, the institutions he
founded like the Columbia Teachers College, the American Psychological Association, the Journal of Philosophy, and the Chicago Laboratory School have endured,
thrived, and changed.

Cross-References
▶ Functional Learning
▶ Learning Environment(s)
▶ Pragmatic Reasoning Schemas
▶ Psychology of Learning

Further Reading
Hickman, L., & Alexander, T. (Eds.). (1998). The essential Dewey (Vol.
1 and 2). Bloomington: University of Indiana Press.
McDermott, J. J. (Ed.). (1981). The philosophy of John Dewey.
Chicago: University of Chicago Press.
The John Dewey Society. (2010). Retrieved 3 April 2011, from http://
doe.concordia.ca/jds/
University of Chicago Laboratory School. (2010). Retrieved 3 April
2011, from http://www.ucls.uchicago.edu/

Diagnosis of Asperger’s
Syndrome
MICHAEL FITZGERALD
Department of Psychiatry, Trinity College Dublin
(TCD), Dublin 2, Ireland

Synonyms
Asperger’s disorder; Asperger’s syndrome; Autistic psychopathy; Criminal autistic psychopathy; High functioning autism

Definition
Asperger’s syndrome (Autistic Psychopathy) was first
described in 1938 by Hans Asperger (Asperger 1938,
1944). He gave a fuller account in 1944 and called the
condition Autistic Psychopathy. The name Asperger’s
syndrome was given to the condition in 1981 by Lorna
Wing (Wing 1981). She noted the:
1. Lack of normal interest and pleasure in people
around them
2. A significant reduction in shared interest

Diagnosis of Asperger’s Syndrome

3. A significant reduction in the wish to communicate
verbally and non-verbally
4. A delay in speech acquisition and impoverishment
of content
5. No imaginative play, or imaginative play confined
to one or two rigid patterns
Van Krevelen and Kuipers (1962, p. 22) pointed out
that persons with this disorder showed “personal unapproachability with problems distinguishing between
dream and reality.” They also noted that “the eye
roams, evades, and is turned inwards.” The speech is
stilted, it is not addressed to the person but into empty
space. . . it sounds false owing to exaggerated
inflection.”
In schools, persons with Asperger’s syndrome are
loners, often bullied, but can show islets of originality
with narrow interest (Fitzgerald 2004). Occasionally,
they can show great ability in narrow areas, for example, science, mathematics, etc. (Fitzgerald and James
2007). Examples would be Isaac Newton (Fitzgerald
1999), Albert Einstein (Harpur et al. 2004), and Andy
Warhol (Harpur et al. 2004).

Theoretical Background
In terms of differential diagnosis, Asperger’s syndrome
is a developmental disorder and not an illness. In the
past it was confused with schizophrenia, Obsessive
Compulsive Personality Disorder (which can cooccur), Attention Deficit Hyperactivity Disorder
(which can co-occur), Avoidant Personality Disorder,
social phobia, and schizoid personality in childhood –
a condition in which there is some overlap with
Asperger’s syndrome.
In terms of persons with Asperger’s syndrome who
commit serious crime, the diagnosis of Criminal Autistic
Psychopathy (Fitzgerald 2010) has been outlined. This
takes up Asperger’s original diagnosis of Autistic Psychopathy and his awareness of the callous unemotional
traits that some persons with this condition showed.
In terms of these disorders there is a lesser type called
Pervasive Developmental Disorder Not Otherwise Specified (DSM-IV-TR, APA 2000) which is less than the full
criteria for Asperger’s syndrome. This requires impairment in reciprocal social interaction associated with
impairment in communication skills or with stereotyped behavior, interests, or activities. Asperger’s syndrome is on the very wide spectrum of autism.

D

975

The DSM-IV-TR (APA 2000) criteria for Asperger’s
syndrome remains controversial. Twachtman-Cullen
(2001, p. 17) identifies the following difficulties with
DSM-IV criteria for Asperger’s syndrome:
1. The DSM-IV definition uses the term “clinically
significant general delay in language”, which is
open to different interpretations.
2. The milestone of single words at age 2 years, used as
an example of normal language development, actually represents a significant expressive language
delay.
3. Use of communicative phrases at age 3 years
involves not just saying a sequence of words but
also communication, meaning the appropriate use
of language for social purposes, which is frequently
not normal in youngsters with Asperger’s syndrome, even if they speak in phrases or sentences.
ICD-10 criteria state that Asperger’s syndrome is
“a disorder of uncertain nosological validity” and “differs from autism primarily in that there is no general
delay or retardation in language or cognitive development” (Rausch et al. 2008, p. 23).
There has been a great deal of research in separating
Asperger’s syndrome from autism. It has never been
possible to categorically and clinically separate these
two conditions (Mayes et al. 2001). It is correct at this
point of time that the most accurate diagnosis is
Autism Spectrum Disorder which should include
Asperger’s syndrome.
Nevertheless, throughout the world huge numbers
of persons have been given the diagnosis of Asperger’s
syndrome which they are happy with and which
describes their disability. There are current discussions
about deleting Asperger’s syndrome from the American
Psychiatric Association DSM-V classification which is
due for publication in the next few years. This would
not be acceptable because it would cause enormous
distress to persons who are satisfied with their diagnosis of Asperger’s syndrome. The change would not
assist these people in any way. The use of Asperger’s
syndrome as one of the synonyms for Autism Spectrum
Disorder would be a satisfactory solution.
Hippler and Klicpera (2003, p. 291) point out that
a study of “74 clinical case records of children with
Autistic Psychopathy (Asperger’s syndrome) diagnosed
by Asperger... revealed (at follow-up) that 68% of the
sample did meet ICD-10 (International Statistical

D

976

D

Diagnosis of Asperger’s Syndrome

Classification of Diseases) criteria for Asperger’s syndrome, although they construed that 25% fulfilled the
diagnostic criteria for autism.”
While it has not been possible scientifically to separate Asperger’s syndrome and High Functioning
Autism from a clinical perspective, Luke Tsai (2001, p.
5) has identified the following features as being more
pronounced in Asperger’s syndrome than High Functioning Autism:
1. Preoccupation with one or more stereotyped and
restricted patterns of interest
2. Talking, reading about violence, and death
3. Condescension in behavior
4. Pedantic speech
5. Moody and easily frustrated with tantrums
This would support Hans Asperger (1979, p. 45)
statement that “it has become obvious that the condition described by myself and Leo Kanner (autism)
concerned basically different types, yet in some respects
there is complete agreement.”

Important Scientific Research and
Open Questions
Asperger’s syndrome is a developmental disorder
which begins with abnormal cell migration in utero in
the developing brain. Heritability is 93%. It shows
theory of mind and empathy difficulties and abnormal
neural connections with excess local connections and
reduced long range connectivity.
Future research needs to continue efforts to
delineate Asperger’s syndrome and the wider ASD
spectrum and to subtype the condition. In the long
term, genotyping may be able to help in this regard.
Rausch et al. (2008, p. 29) note that “future diagnostic
language may better distinguish the markedly impaired
social interaction of autism from that of Asperger’s
cases, where such is not present. In this case the
term “marked” could be elaborated to differentiate
from the more subtle speech stereotype, idiosyncrasy,
or difficulties with conversation maintenance prior to
three years described in Asperger’s.” They also point
out that “advances in endophenotypy (will) come with
advances in diagnostic distinction, but there is also
great promise from the potential for advances with
psychological and biological phenotyping e.g., performance on theory of mind tasks and facial emotion

recognition, coupled with studies of brain activation
studied during such tasks” (p. 54).

Cross-References
▶ Intact Implicit Learning in Autism
▶ Learning and Consolidation in Autism

References
American Psychiatric Association. (2000). DSM-IV-TR. Washington,
DC: American Psychiatric Association.
Asperger, H. (1938). Das psychisch abnormale Kind. Wiener Klinische
Wochenschrift, 1314–1317.
Asperger, H. (1944). Die autistischen Psychopathen im Kindesalter.
Archiv für Psychiatrie und Nervenkrankheiten, 117, 76–136.
Asperger, H. (1979). Problems of infantile Autism. Communication,
13, 45–52.
Fitzgerald, M. (1999). Did Isaac Newton have Asperger’s disorder?
European Child and Adolescent Psychiatry Journal, 8, 244.
Fitzgerald, M. (2004). Autism and creativity: Is there a link between
autism in men and exceptional ability? New York: Brunner
Routledge.
Fitzgerald, M. (2010). Young, violent & dangerous to know. New York:
Nova Science.
Fitzgerald, M., & James, I. (2007). The mind of the mathematician.
Baltimore: Johns Hopkins University Press.
Harpur, J., Lawlor, M., & Fitzgerald, M. (2004). Succeeding in college
with Asperger’s syndrome. London: Jessica Kingsley.
Hippler, K., & Klicpera, C. (2003). A retrospective analysis of the
clinical case records of “Autistic Psychopaths” diagnosed by Hans
Asperger and his team at the University Children’s Hospital
Vienna. Philosophical Transactions of the Royal Society of London.
Series B, Biological Sciences, 358(1430), 291–301.
Mayes, S., Calhoun, S., & Crits, D. (2001). Does DSM-IV
Asperger’s disorder exist? Journal of Abnormal Child Psychology,
29(3), 63.
Mesibov, G., Shea, V., & Adams, L. (2001). Understanding Asperger’s
syndrome and high functioning autism. Lancaster: Kluwer/
Plenary.
Rausch, J., Johnson, M., & Casanova, M. (2008). Asperger’s disorder.
New York: Informa.
Tsai, L. (2001). From Autism to Asperger’s disorder. In American
Academy of Child and Adolescent Psychiatry meeting, Hawaii.
Washington DC: American Academy of Child and Adolescent
Psychiatry.
Twachtman-Cullen, D. (2001). In G. Mesibov, V. Shea, & L. Adams.
Understanding Asperger’s syndrome and High Functioning Autism.
Lancaster: Kluwer/Plenary.
Van Krenelen, A., & Knipers, C. (1962). The Psychopathology of
Autistic Psychopathy Acta Paedopsychiatrica, 29, 22–31.
WHO. (2004). International classification of diseases (Vol. 10).
Geneva: World Health Organisation.
Wing, L. (1981). Asperger’s syndrome – A clinical account. Psychological Medicine, 11, 115–129.

Diagnosis of Learning

Diagnosis of Learning
NORBERT M. SEEL
University of Freiburg, Freiburg, Germany

Synonyms
Assessment; Evaluation of learning

Definition
The word diagnosis is derived through Latin from the
Greek word diάgignω̃skein and means to discern or
distinguish. Accordingly, a central feature of diagnosis
is to distinguish between objects of interest and their
features. Basically, diagnosis refers to both a critical
analysis of the nature of something and a conclusion
reached by the analysis.
In human sciences, such as medicine or psychology,
the term diagnosis refers to collecting and interpreting
information with the aim of determining which nonobservable states can be considered as the “true state of
nature.” Often it is not possible to influence or determine what state of nature will occur; but what one can
do is collecting and processing data with the aim to
arrive at a probabilistic estimation of the true state.
This also holds true for learning which can be considered as a change in discrete mental states that are
basically not observable. Actually, the true mental states
of learners will never be directly accessible, but must be
inferred from data gained with the help of particular
techniques of measurement and on the basis of theoretical assumptions and hypotheses about the students’
mental states.

D

example of medical diagnosis, instances of D would
be blood pressure, appetite, special aches as well as
data from interviews, tests, and specific examinations.
The more data the physician obtains the more likely
a correct diagnosis will result. Basically, the same holds
true for the field of psychology and education where
diagnosis of learning consists in proceeding from
a relatively diffuse prior probability distribution
po(Hi) to a more informative posterior probability
distribution pm(Hi). Clearly, this transformation from
po(Hi) to pm(Hi) is not unconstrained due to learner
characteristics (abilities and skills), organizational conditions, and the curriculum.
Human sciences have at their disposal a remarkable
wide and varied pool of methods to assess learning and
cognition, ranging from naturalistic observation in
clinical situations to computer simulation, from traditional experimental methods to performing linguistic
analyses, from recording electrical impulses of the
brain to measuring reaction times or verbal protocols
(Simon and Kaplan 1989). In psychology and education, usually functional approaches of diagnosis are the
golden standard. They include (a) experimental
methods based on the systematic observation of the
learners’ overt behavior when operating with given
tasks, (b) protocol analyses including verbal reports,
“think-aloud” data, and content analysis of linguistic
expressions, and (c) computer-based modeling and
simulation.
In accordance with psychometric conventions,
a basic distinction can be made between (a) responsive
(reactive) diagnosis and (b) constructive (nonreactive)
agent-based diagnosis.

Responsive (Reactive) Diagnosis
Theoretical Background
Speaking about learning, discussed here in terms of the
acquisition, storage, and retrieval of information, is
speaking about a theoretical construct, i.e., something
which cannot be observed but can be assessed on the
basis of observable behavior or verbal statements made
in the course of solving particular tasks. The basic
assumption of diagnosis is that there are sources of
information which provide data D that might be used
to modify the initial hypotheses about possible states of
nature (e.g., of learning). The processing of D succeeds
in transforming the prior probabilities po(Hi) into posterior probabilities pm(Hi). Remaining within the

Since the nineteenth century, psychologists have
thought that human reaction time provides clues
about mental processes and the organization of the
mind. Accordingly, reaction times and their changes
under different experimental manipulations have been
used as evidence for testing hypotheses about processes
and structures of the human mind. Measurements of
reaction time have been used, e.g., to distinguish
between serial and parallel processing of information.
Furthermore, this methodology has been applied to
research on attention control, information flow, and
the acquisition of skills (Marinelli et al. 2010; Robertson 2007). In accordance with the basic assumption

977

D

978

D

Diagnosis of Learning

that every mental state corresponds with a specific
physical state of the brain, cognitive science operates
also with psychophysical methods (such EEG or as
Positron Emission Tomography, PET) to assess mental
states. Since the 1990s, numerous studies have been
conducted by using PET-methodologies (e.g., Blaxton
et al. 1996; Halsband et al. 1998).
Another tradition of responsive diagnosis consists
in the use of standardized tests and questionnaires,
most often within the realm of experimental research.
A standardized test or questionnaire is an assessment
which is administered under standardized or controlled conditions that specify where, when, how, and
for how long individuals may respond to the questions
or “prompts” of the test. The advantages of standardized tests and questionnaires are seen in their objectivity and reliability whereas the most understated risk for
a valid interpretation is the error produced by the
respondent in accomplishing the test items. Even
when the respondent subject is well intentioned and
cooperative, several errors, such as awareness of being
tested and response sets, may reduce the reliability and
validity of the measure. However, what is considered
a risk for responsive measurements can be considered
an advantage for the nonreactive procedures of knowledge diagnosis.

explanations, inferences, hypotheses, speculations,
and justifications are considered effective means to
assess knowledge. In spite of their indisputable ecological validity, verbal data and protocols have been
criticized by some authors (e.g., Nisbett and Wilson
1977) for their deficiencies with regard to psychometric
standards of reliability and validity. However, Chi
(1997) and Ericsson and Simon (1993) have developed
practical guides to quantify verbal data and to control
their reliability.
Another class of constructive methods of diagnosis
operates with a mixture of verbal and graphical representations of knowledge. Concept maps and semantic
networks are probably the best representatives for this
kind of knowledge assessment. Semantic networks
allow to map the relevant structure of knowledge at
one time and to suppress irrelevant details. They make
the relevant objects and relations explicit and expose
natural constraints with regard to “causality,” i.e., how
an object or relation may influence another one. In
recent years, semantic networks and conceptual graphs
have emerged as important and computable tools for
diagnosing knowledge about the world (see for an
overview Ifenthaler et al. 2010).

Constructive Methods of Diagnosis

Comparable with the problems of measuring theoretical constructs in general, the assessment of learning is
associated with problems of reliability and validity of
the applicable psychometric methods.

Virtually all learning takes place by talk and text.
Accordingly, discourse is an important mediator of
learning, and language can be considered as one of
the most important windows to the mind. Verbal communication is what subjects regularly use to mediate
their ideas, knowledge, thoughts, and feelings. Accordingly, various methods of verbalization play a central
role in the diagnosis of individual knowledge. So for
example, introspection and introspective data are used
as a means of self-reflection in everyday life and in
science since the times of Wundt, Titchener, and others
pioneers of modern psychology. Gillespie (2007) points
out that the idea that thought is a self-reflective internal
dialogue goes back, at least, to Plato and can be
found until today as a method of self-reflection that
can be used for assessing mental states. Closely
related with this idea is the use of verbalizations and
think-aloud protocols to assess mental states. Besides
this “direct communication” of thoughts and ideas by
means of verbalizations, more extensive verbal

Important Scientific Research and
Open Questions

Problems of Reliability
In test theory, reliability is defined as the degree with
which measures do not contain errors: Measurement
errors impair the reliability, and thus, the generalizability of measures resulting from a single measurement of
an individual. Corresponding with the generalizability
theory of Cronbach, errors of measurement are indicative for the control of the situational dependency of
measures as well as for the control of temporal stability
of measures.
The error variance of any psychometric method
ðO; ; P; fXi ji 2 I g; QÞ with a distribution Bl from
2
þ
2
Q is a function

 2slE : fXi ji 2 I g ! < with slE ðXi Þ ¼
2
s ðEl Xi Þ ¼ E Ei Xi . The reliability is then a function
rlxx : fXi ji 2 I g ! ½0;1, which determinates the true
score ΤlXi through rlxx ðXi Þ ¼ s2 ðTl Xi Þ=s2 ðXi Þ. Thus,

Diagnosis of Learning

the reliability is defined through the proportion of the
variance of the true scores and the variance of test data.
Actually, this definition of reliability refers to
a particular measurement with a test, and under the
assumption of consistency, various measurements with
the same test should not change the score of the quotient. This presupposed invariance grounds on the
individual specific approach which assigns a constant
true score ip and an individual error variable eip (with
a fixed distribution) to every person for every measurement Xip. A corresponding application of this reliability
concept basically presupposes parallel measurements
or repetitions of measurement aiming at the comparison of individuals across different situations.
Psychometric methods assume basically that a test
person demonstrates an individual pattern of stable
behaviors in similar situations whereas the person’s
behavioral pattern varies across different situations.
However, in measuring learning processes and outcomes substantial changes of behavior or knowledge
serve as indicator of learning when tests are administered to the same individuals in different situations to
different times. Therefore, we need a measurement of
change that demands for appropriate test models. Tack
(1980) has developed the probabilistic foundations of
a test theory for the measurement of change by differentiating between a person-related p-reliability and
a situation-related s-change reliability with the aim to
provide evidence for effects of individual specific
conditions.
1. The n-change value regarding m is a consistent
family of Βmn-measurable random variables:
Dn=m Xi ¼ Tmn Xi  Tm Xi .
2. The n-change variance regarding m is a
þ
function s2n=mD

 : fXi ji 2 I g!2 < , for which
2
2
sn=mD ðXi Þ ¼ s Dn=m Xi ¼ e Dn=m Xi .
3. The n-change reliability (change sensitivity) regarding m is a function rn=mDD : fXi ji 2 I g ! ½0; 1, for
which rn=mDD ðXi Þ ¼ s2n=mD ðXi Þ=s2 ðXi Þ.
When we now define m = p and n = s, so that
a system Bp of individuals and a system Bs of situations
result (with Bp, Bs 2 Q), then the s-change reliability
regarding p is exactly the difference resulting from p  sreliability
minus
p-reliability:
rs=pDD ðXi Þ ¼
rspXX ðXi Þ  rpXX ðXi Þ. Furthermore, the s-change
reliability regarding p can be maximally one minus
the p-reliability: rs=pDD ðXi Þ  1  rpXX ðXi Þ. That is,

D

the usual (person-related) p-reliability determines the
upper limit of the possible situation-dependent change
reliability.
Finally, the covariance between s-change scores
regarding p for different variables of test scores corresponds to the difference of the covariances
of p  s-true



2
scores
and the p-true
scores:
s Ds=p Xj ; Ds=p Xk ¼





s2 Tsp Xj ; Tsp Xk  s2 Tp Xj ; Tp Xk . Discussed in
terms of variance analyses, the s-error variance corresponds to the variance within the examined situations
that are characterized through the realization of
a possible variation of conditions, and the variance of
the s-true scores corresponds to the variance between
the situations. Thus, the s-change reliability regarding p
depends upon Bs, i.e., it exists only for a particular kind
of the situation-dependent variation. On the whole,
this is a practicable procedure to determine the reliability in the context of change measurement as it is necessarily involved in research on learning.

Problems of Validity
Undoubtedly, the question whether a psychometric
method really assess the interesting person variable
concerns one of the most central issues of diagnostics.
In the frame of differential diagnostics, it constitutes
the problem of validity asking for an adequate theoretical interpretation of measured test scores considered
representative for a certain population of subjects.
However, the diagnosis of learning depends additionally on variations of situations and times.
Remaining within the classical test theory, the question for adequate theoretical interpretations of test
scores had led to the conception of construct validity
whereby construct is usually defined as a theoretical
concept to describe the individual’s disposition to be
measured. The construct is considered embedded into
a network of hypothetical relationships between other
concepts, and thus, a test is valid if the test variables
indicate covariances with other variables close to the
construct in question. Accordingly, structure equations
models´ and causal models, respectively, have been proposed (e.g., Eid 1995) to validate causal relations
through an analysis of complex linear relations between
variables considered partly either as latent or manifest.
Clearly, every interpretation of test scores goes
beyond the test-related variables: It essentially involves
attempts to make accessible variations of latent variables on the basis of the frequency of correct answers in

979

D

980

D

Diagnosis of Learning

test items. Of course, research interest focuses squarely
on true states of nature – a commodity that, if measurement error is large, may differ considerably from
observed data. Similarly, when members of a group are
changing over time on some important attribute, it is
not the fallible observed changes that are of critical
interest but the underlying true changes. However,
this presupposes (a) a precise clarification of those
cognitive processes that occur in a special learning
situation, (b) a specification of those variations of
personal, situational, and task-related characteristics
that determine these cognitive processes, and
(c) a specification of the resulting observable data to
be interpreted.
This immediately leads to several important questions to be answered with regard to the validation of
theoretical constructs such as learning: Which cognitive operations and processes can be theoretically justified in order to map the structure of latent variables
adequately? Which test situations and items are suitable or can be constructed in order to initiate the aimed
cognitive processes? Does the theoretically postulated
information processing constitute a homogenous class
of test items, i.e., can the test item solutions be
explained on the basis of the same theoretical assumptions? Is an empirically reasonable comparability of all
answers possible that allows inferences with regard to
the internal processes? Are person-specific variations of
operational sequences identifiable with regard to single
items, and is it possible to assign distinctive characteristics of the person and the given answers to these
interindividually different sequences?
To give reasonable answers to these questions is the
central goal of the approach of cognitively diagnostic
assessment (cf. Nichols 1994) according to which the
fundamental basis of the validation of theoretical constructs consists of a comprehensive cognitive task analysis. Thus, cognitive task analysis creates a rational
basis for the selection of the contents of test items as
well as for the expectancy of consistencies of the
answers and the prognosis of specific results. The probability of a correct answer is considered then a function
of both the faculty of a person and the task
characteristics.

Measurement of Change
Why is measurement of change over time so important
in research on learning? The answer is straightforward.

When students learn something new, acquire new
knowledge, when they modify their preconceptions,
they are changing in fundamental and interesting
ways that indicate the constructive and cumulative
characteristics of learning. Thus, only by measuring
individual change is it possible to document each person’s progress in learning.
Contrary to a simple misconception according to
which individual change has been considered as an
increment (i.e., the difference between before and
after), individual change in learning takes place in
a sequence of discrete steps over time. Therefore, each
subject should be measured repetitively over extended
periods of time in order to understand the progression
of learning. Taking a snapshot of a learner’s observed
status before and after an intervention is certainly not
the best way to reveal the intricacies of their learning
progress. Changes might be occurring smoothly over
time with some complex and substantively interesting
trajectory. Crude pre-/post-measurement can never
reveal the details of that trajectory. Therefore, a truly
longitudinal perspective must be adopted. Willett
(1988) suggests to assemble and inspect observed
growth records for every subject in the dataset (i.e.,
a graph of observed data displayed against time) in
order to provide evidence of the subject’s changing
over time.
Corresponding to the exploratory inspection of
observed growth records and empirical growth trajectories, a formal multi-wave analysis of changes requires
that a statistical model is available to represent individual change over time. This model consists of two
parts: a structural part that represents the dependence
of true states on time, and a stochastic part that represents the random effects of measurement error.
Perhaps, the investigator’s most important task is to
select the appropriate individual growth model that
can be used. Usually, this decision is made via extensive
preliminary data-analyses in which the observed
growth records are systematically and carefully
explored. Then, between-person differences in change
can be tested.
Willett (1988) has suggested applying ordinary leastsquares regression analysis in order to estimate the
within-person growth parameters, and thus, he argues
that growth rates provide precise measurements of individual change. Since the 1990s, there are several dedicated computer programs available for simultaneously

DICK Continuum in Organizational Learning Framework

estimating all of the parameters of growth models, and
for providing appropriate standard errors and goodnessof-fit statistics (see, e.g., Kreft et al. 1990).

Cross-References
▶ Assessment in Learning
▶ Automated Learning Assessment and Feedback
▶ Dynamic Assessment
▶ Formative Assessment and Improving Learning
▶ Learning Criteria and Assessment Criteria
▶ Measurement of Change in Learning
▶ Models of Measurement of Persons in Situations

References
Blaxton, T. A., Zeffiro, T. A., Gabrieli, J. D. E., Bookheimer, S. Y.,
Carrillo, M. C., Theodore, W. H., & Disterhoft, J. F. (1996).
Functional mapping of human learning: A positron emission
tomography activation study of eyeblink conditioning. Journal
of Neuroscience, 16(12), 4032–4040.
Chi, M. T. H. (1997). Quantifying qualitative analyses of verbal data: A
practical guide. The Journal of the Learning Sciences, 6(3),
271–315.
Eid, M. (1995). Modelle der Messung von Personen in Situationen.
Psychologie Verlags Union: Weinheim.
Ericsson, K. A., & Simon, H. A. (1993). Protocol analysis: Verbal
reports as data. Cambridge: The MIT Press.
Gillespie, A. (2007). The social basis of self-reflection. In J. Valsiner &
A. Rosa (Eds.), The Cambridge handbook of socio-cultural psychology (pp. 678–691). Cambridge: Cambridge University Press.
Halsband, U., Krause, B. J., Schmidt, D., Herzog, H., Tellmann, L., &
Müller-Gärtner, H. W. (1998). Encoding and retrieval in declarative learning: a positron emission tomography study.
Behavioural Brain Research, 97(1–2), 69–78.
Ifenthaler, D., Pirnay-Dummer, P., & Seel, N. M. (Eds.). (2010).
Computer-based diagnostics and systematic analysis of knowledge.
New York: Springer.
Kreft, I. G. G., de Leeuw, J., & Kim, K. S. (1990). Comparing four
different statistical packages for hierarchical linear regression:
Gemmohlm2, andvarc. Los Angeles: Center for the Study of
Evaluation, University of California at Los Angeles.
Marinelli, L., Perfetti, B., Moisello, C., Di Rocco, A., Eidelberg, D.,
Abruzzese, G., & Ghilardi, M. F. (2010). Increased reaction time
predicts visual learning deficits in Parkinson’s disease. Movement
Disorders, 25(10), 1498–1501.
Nichols, P. D. (1994). A framework for developing cognitively diagnostic assessments. Review of Educational Research, 64(4), 575–603.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can
know: Verbal reports on mental processes. Psychological Review,
84, 231–259.
Robertson, E. M. (2007). The serial reaction time task: Implicit motor
skill learning? Journal of Neuroscience, 27(38), 10073–10075.
Simon, H. A., & Kaplan, C. A. (1989). Foundations of cognitive
science. In M. I. Posner (Ed.), Foundations of cognitive science
(pp. 1–47). Cambridge: The MIT Press.

D

981

Tack, W. H. (1980). Zur Theorie psychometrischer Verfahren:
Formalisierung der Erfassung von Situationsabhängigkeit und
Veränderung. Zeitschrift für Differentielle und Diagnostische
Psychologie, 1(2), 87–106.
Willett, J. B. (1988). Questions and answers in the measurement of
change. Review of Research in Education, 15, 345–422.

D

Dialogical Mapping
Dialogic mapping is a development of conceptmapping technology that allows the connections
between ideas to be described in a less controlled way
than previously. It encourages dialogic discussion of
maps and the description of them in a wide variety of
communicative forms that accommodates the modes
of thought and expression found in contrasting academic disciplines.

Dialogue
▶ Argumentation and Learning

DICK Continuum in
Organizational Learning
Framework
MOHAMAD HISYAM SELAMAT, MOHD AMIR MAT SAMSUDIN
College of Business, Universiti Utara Malaysia, Sintok,
Kedah, Malaysia

Synonyms
Communication; Direction; Internality; Knowledge;
Organizational learning

Definition
Klimecki and Lassleben (1998) define organizational
learning (OL) as a communication-based process
where the organization overcomes its previous boundaries of knowledge and ability by allowing its members
to share knowledge, interact, influence each other, and
cope with difficult situations. On the other hand,
Nonaka and Takeuchi (1995) viewed OL as involving

982

D

DICK Continuum in Organizational Learning Framework

the generation, absorption, and sharing of tacit knowledge, and emphasized the importance of interaction
among people toward the development of OL capabilities. In short, OL is the process of continuous innovation through the creation of new knowledge. It is an
ongoing process that takes place as staff members
engage in knowledge work (Davenport et al. 1998).
These views illustrate the importance of having continuous information systems (IS) updates using the
medium of communication among the employees.
However, to enable the communication process,
employees have to be self-confident and to be encouraged to talk to others in the workplace. The lack of
confidence and anxiety will demotivate an individual
from communicating with others and consequently
reduce the effectiveness of the OL framework (Harvey
and Butcher 1998). Thus it is declared that individual
development is the starting point for OL-based IS.

Theoretical Background
Tacit knowledge is an individual’s intuition, beliefs,
assumptions, and values, formed as a result of experience (Saint-Onge 1996). Augier and Vendelo (1999)
argued that due to its transparent and subjective nature
tacit knowledge is not easily externalized. Many
researchers in the organizational development area
(Butcher et al. 1997) feel that there are two major
stumbling blocks in involving individuals in the organizational development process:
1. The difficulty of individuals to externalize and share
their tacit knowledge (Harvey and Butcher 1998)
2. The difficulty of individuals to obtain information
from their colleagues (Butcher et al. 1997; Harvey
and Butcher 1998)
3. The difficulty to self-document the externalized
and shared tacit knowledge (Selamat and Choudrie
2007)
In summary, the capability to externalize, share, and
document tacit knowledge is of paramount importance
for OL framework. This in turn illustrates that staff
members should be instilled with that capability. Therefore, at this point it is declared that individual development should become the starting point in an OL-based IS
developmental framework. This entry intends to use this
reasoning to illustrate the role of direction (D), internality (I), communication (C), and knowledge (K) in OL. In
this entry, it is coined as DICK continuum.

This entry argues that staff members need to first
and foremost understand their direction within the
organization. Being equipped with this value enables
staff members to know what should be done when
working in the organizations. To assist in this process,
three fundamental aspects need to be understood:
1. Task completion – staff members need to realize
that behind every single penny that they earn from
the company is the responsibility to complete the
given tasks (Selamat and Choudrie 2007). When
recruited, an individual must be grateful to his/her
organization. This acknowledgment should be
followed by an inspiration to work hard and
smartly for the sake of the company.
2. Task scope – staff members must be able to prioritize organizational tasks (Butcher et al. 1997). To
ease the process of understanding task priority,
staff members can divide tasks into major or
minor categories (Selamat and Choudrie 2007).
Major tasks include the superior’s instructions
and operational tasks. Minor tasks are ceremonial
committee and union affairs. In this case, staff
members must be clear about their job specifications so that they will not waste time on tasks that
do not contribute to performance recognition.
Major tasks must be given a higher priority than
minor tasks.
3. Personal targets – staff members need to be clear
about their organizational aims and targets in
a period of time. Working without aims or targets
is like a blind person touching things in a dark
room. Aims and targets should be developed to
shed light on how to monitor organizational activities and determine future directions (Butcher et al.
1997; Selamat and Choudrie 2007).
After determining direction, staff members must
have strength to achieve the intended direction. This
research proposes eight internal strengths that should
be instilled to staff members which are as follows:
1. Personal confidence – it is a self-belief in undertaking and accomplishing organizational tasks
(Butcher et al. 1997). As one of the elements that
prevent staff members from externalizing and sharing
their tacit knowledge is lack of confidence (Harvey
and Butcher 1998), this element should be emphasized when understanding an OL framework.

DICK Continuum in Organizational Learning Framework

2. Observing accepted organizational approaches –
this value allows individuals to create and adapt
specific competencies for specific situations
(Selamat and Choudrie 2007). By observing
accepted organizational approaches, staff members
can undertake tasks based upon the right
approach for the right situation (Selamat and
Choudrie 2007).
3. Undertaking tasks with commitment and selfdiscipline – this value requires every staff member
to bear in mind in the workplace that we must do
the job with commitment and self-discipline
(Selamat and Choudrie 2007). This internal
strength is the backbone of enabling knowledge
and skills utilization among staff members.
4. Self-awareness – it is an ability to determine the tasks
that need to be accomplished at the current time and
accomplish the determined tasks according to an
accepted organizational approach (Selamat and
Choudrie 2007). In other words, it is related to the
phrase “do the right things at the right time.”
5. Self-remembrance – the value that requires staff
members to mind their actions when undertaking
a task so that it can be accomplished effectively and
to remember that through their effective actions the
company can achieve a good profit and consequently give them a good salary and bonus (Selamat
and Choudrie 2007).
6. Compassion – it is defined as having a feeling that
the whole organization is like a family (Selamat and
Choudrie 2007). Each staff member should appreciate the other members’ efforts because all of them
have the same aim and objective in terms of job
security.
7. Sincerity – every staff member must have a feeling
that he/she works for the sake of the company and
for fulfilling his/her responsibility to the company.
8. Willingness to change – this value leads to a continuous improvement in an organization so that its
competitiveness does not deteriorate. This is due to
rapid changes in the organizational life and business environment.
From the aforementioned discussion, eight internal
strengths were considered to be relevant in establishing
learning environment. These internal strengths can
become the elements that can be referred to in understanding situations and taking actions. Additionally,

D

these values are related closely to OL; therefore, they
should be instilled in staff members in order to establish a learning environment.
Another element that is proposed by this research
when promoting OL is the ability to communicate
within the organization. This is because staff members
are not alone in the organization and thus need to
communicate when undertaking task. This research
divides communication into two categories which are:
1. Formal and informal discussions – staff members
face various tasks in daily activities (routine, nonroutine, official, and unofficial). To cope with this
variety, the integration of formal and informal discussion in handling tasks becomes necessary (Earl
and Hopwood 1980; Selamat and Choudrie 2007).
Formal approaches are procedures such as meetings, progress reports, and performance evaluation
reports. Informal approaches include dialog, faceto-face interaction, corridor meeting, lunch table
chats, and coffee/tea table chats.
2. Rational discourse – whenever an IS is applied, it
serves some human interests; therefore, the design
choices are made to serve some interests at the
expense of others and involve moral value judgments. This means that practical advice concerning
the design of a learning-based IS must not be limited
only to technical aspects, but also address moral
issues, such as what is good or bad, or right or
wrong in any particular application. Therefore,
there is a need to establish a platform to approach
such value judgments in a rational way. A rational
discourse can legitimize the selection of a design ideal
because it ensures that the arguments of all interested
parties are heard, that the choice results in an
informed consensus about the design ideal, and the
formal value choice is only made by the force of the
better argument (Klein and Hirschheim 1996).
The last element is knowledge (K). Knowledge is
defined as a generic knowledge and not as a specific
expertise. In addition, it represents the ability to utilize,
externalize, and share tacit knowledge in an effective
manner. From the aforementioned discussions, it is
declared that this entry intends to propose DICK continuum in the process of establishing learning environment. This is something that prior research has not
undertaken. This strategy makes this entry unique to
that of others. In short, for the first element of this

983

D

984

D

DICK Continuum in Organizational Learning Framework

entry’s conceptual framework an understanding of
DICK continuum is required. The DICK continuum
can assist in building a confident and responsible individual (Selamat and Choudrie 2007). These values, in
turn, create three important competencies as follows:
(1) influencing skills; (2) sharing attitudes; and
(3) inquisitive tendencies (Selamat and Choudrie
2007). In other words, influencing skills, sharing attitudes and inquisitive tendencies are the second element
of this entry’s conceptual framework.
It was also found above that there are problems when
developing OL-based IS, which is the need to develop an
individual’s ability when externalizing and sharing tacit
knowledge. In such an instance, DICK continuum,
influencing skills, sharing attitudes and inquisitive tendencies are the humanistic elements that should be considered when considering means of overcoming the
difficulties in externalizing and sharing tacit knowledge.
This is because by practicing the above influencing skills,
sharing attitudes and inquisitive tendencies, individuals
can generate creative ideas (I), actions (A), reactions (R),
and reflections (R) (Selamat and Choudrie 2007). The
terms IARR represent forms of activities within an
organization. These activities then allow the externalizing and sharing of tacit knowledge that can provide
synergistic inputs for a continuous development of IS.
Therefore, the IARR continuum is the third element of
this entry’s conceptual framework.
However, for this the tacit knowledge must be initially documented. This can be achieved by the value of
self-documentation, which is also developed by DICK
continuum (Selamat and Choudrie 2007). Due to the
development of the elements of DICK, the willingness
to question implicit assumptions, explore new possibilities, and directing energies toward higher standards
enables a staff member to be well prepared and to use
good quality documented progress reports or working
papers. In the longer term, this then ensures that there
is a tangible means of verifying and validating tacit
knowledge. Therefore, tacit knowledge documentation
is the fourth element of this entry’s conceptual
framework.
Reflecting on the above discussion, it can be determined that individual development is the starting
point of an OL framework. Additionally, it can be
learnt from the previous discussion that DICK continuum should become the starting point for the individual development.

As mentioned above, externalized and shared tacit
knowledge must be documented. This knowledge is
then transformed into explicit knowledge (e.g.,
through business reports, written descriptions, or
instructions). All these self-documentation is then
given to the systems officers. At this stage, the systems
officers study the documented inputs provided by staff
members and codify them. By the time the inputs are
transformed into codified domains within the systems,
they become information that assist staff members in
fulfilling their responsibility. This is the fifth element of
this entry’s conceptual framework.
To understand the relationship between the earlier
mentioned five elements a diagrammatic representation has been developed, which is illustrated in Fig. 1.
As shown in the diagram, individual development is
initially fostered by the elements of direction, internality, communication, and knowledge (DICK continuum). In this case, the first element of the framework
is represented by Stage A in the diagram. As the elements of DICK enable the use of knowledge and skills
in an effective manner, they are pertinent for the development of influencing skills (Stage B), sharing attitudes
(Stage C), and inquisitive tendencies (Stage D)
(Selamat and Choudrie 2007). These three stages represent the second element of the OL framework. When
undertaking influencing, sharing, and inquiring activities, an individual implicitly expresses tacit knowledge.
This expression is either in physical form (actions and
reactions) or verbal form (ideas and reflection)
(Selamat and Choudrie 2007) (Stage E). The continuum of ideas, actions, reactions, and reflections provides externalized tacit knowledge for OL-based IS
development (Selamat and Choudrie 2007). Stage
E of the diagram represents the third element of the
framework (as noted above). However, the externalized
ideas, actions, reactions, and reflections must initially
be documented. This process is undertaken at Stage F,
and it represents the fourth element of the above theoretical framework. The documented inputs provided
by staff members can be transformed into codified
domains within the systems (system database) or compiled in files (Stage G). In turn these databases or files
can be utilized to refine decision and develop strategies
for future development. Through this process, an individual’s understanding of the organization’s activities
(tacit knowledge) is also enriched. This new understanding in turn becomes a platform for a continuous

Didactics

Knowledge (K)

Stage B:
Influencing skills

Stage A: D-I-C
• Direction
• Internality
• Communication

Stage C:
Sharing attitudes

D

Organization
Problematic
situations

Stage D:
Inquisitive tendencies

Stage E: I-A-R-R Continuum
• Ideas
• Actions
• Reactions
• Reflections
Internalization
Stage F: Knowledge documentation

Stage G: System database
Information dissemination
Stage H: Organizational
strategies/operation/approach
improvement

DICK Continuum in Organizational Learning Framework. Fig. 1 Framework for a continuous organizational learning

learning process. In the diagram, this process is
represented by Stage H. Stages G and H represent the
fifth element of the aforementioned theoretical
framework.

Cross-References
▶ Knowledge Management
▶ Organizational Change and Learning
▶ Shared Cognition

References
Augier, M., & Vendelo, M. T. (1999). Networks, cognition and management of tacit knowledge. Journal of Knowledge Management,
3(4), 252–261.
Butcher, D., Harvey, P., & Atkinson, S. (1997). Developing business
through developing individuals. Cranfield: Cranfield University.
Davenport, T. H., De Long, D. W., & Beers, M. C. (1998). Successful
knowledge management projects. Sloan Management Review,
39(2), 43–57.
Earl, M. J., & Hopwood, A. G. (1980). From management information to information management. In L. Lucas & S. Lincoln (Eds.),
The information systems environment. Amsterdam: NorthHolland.

985

Harvey, P., & Butcher, D. (1998). Those who make a difference:
Developing businesses through developing individuals. Industrial and Commercial Training, 30(1), 12–15.
Klein, H. K., & Hirschheim, R. (1996). The rationality of value
choices in information systems development. Foundations of
Information
Systems.
http://www.cba.uh.eduparks/fis/
kantpap.htm. Accessed 15 Sept 2002
Klimecki, R., & Lassleben, H. (1998). Modes of organizational learning: Indications from an empirical study. Management Learning,
29(4), 405–430.
Nonaka, I., & Takeuchi, H. (1995). The knowledge creating company.
New York: Oxford University Press.
Saint-Onge, H. (1996). Tacit knowledge: The key to the strategic
alignment of intellectual capital. Strategy and Leadership Journal,
24(2), 10–14.
Selamat, M. H., & Choudrie, J. (2007). Using meta-abilities and tacit
knowledge for developing learning based systems: A case study
approach. The Learning Organization, 14(4), 321–344.

Didactics
▶ Choreographies of School Learning

D

986

D

Didactics, Didactic Models and Learning

Didactics, Didactic Models and
Learning
KARL-HEINZ ARNOLD
Department of Applied Educational Science,
Institute of Education, University of
Hildesheim, Hildesheim, Germany

Synonyms
Classroom teaching and learning; Curriculum development; Designing lessons; General education; Lesson
planning

analysis of single acts of teaching and learning is not
intended. Therefore, empirical research on teaching
and learning can been seen as complementary micro
theories modeling the processes of teachers’ delivery of
didactically planned lessons and students’ performance
of the intended (and unintended) learning tasks.
From the perspective of methodology, general didactics serves as a scientific means of decision making in
educational planning at different levels. It is regarded as
a value-laden, prescriptive theory with a normative
background, whereas instructional science preserves
descriptive theories on classroom teaching and
learning.

Theoretical Background
Definition
The word didactics comes from the Greek word
“didάskein” (didáskein), which means teaching. The
scientific term didactics (sometimes also spelled
“Didaktik” as in German) stems from the German
tradition of theorizing classroom learning and teaching. It is a singular noun spelled in the plural form,
indicating that connotations to the somewhat pejorative English word “didactic” (text overburdened with
instructive matter or oversimplifying way of teaching)
are not intended. Didactics serves as a major theory in
teacher education and syllabus development, especially
in the German-speaking and Scandinavian countries,
as well as in Finland (didaktiikka) and in Russia
(didaktika). With a slightly different meaning, it is
also employed in France (didactique comparée) and
Spain (didáctica general) as well as in the Dutch
(algemene didactiek) and Africaans literature.
General didactics represents the overarching theory
of both decision making on and processes of teaching
and learning in societal institutions (especially in
schools and universities devoted to general and
domain-specific education), whereas subject-matter
didactics covers the theories of teaching and learning
a particular school subject. The two disciplines are
considered complementary; it also can be argued that
general didactics provides the common framework for
the particular purposes of teaching school subjects.
Regarding the conceptual level of operationalization,
general didactics as a macro theory of classroom teaching and learning is concerned with decision making on
general goals as well as specific objectives on content
and methods of instruction; the description and

The foundational period of didactics lies in the seventeenth century and is associated with the work of
Johann Amos Comenius and Wolfgang Ratke. Comenius worked out a comprehensive concept for teaching
in public schools: Didactica Magna (1657) intended “to
teach everybody everything completely, quickly, pleasantly, and thoroughly.” Comenius also penned the
famous Orbis Sensualium Pictus (1658). This textbook
with short chapters was written in German; Latin
translations of the key terms were added, and figures
illustrated the topics.
At the end of the eighteenth century, educators
from the Philanthropist movement (e.g., Basedow,
Campe) made significant contributions to subjectmatter education, especially in the natural sciences,
and to teaching methods that were practiced in newly
founded, innovative schools. It was Johann F. Herbart
(1800), an academic disciple of the German philosopher Immanuel Kant, who not only outlined education
as a scientific discipline but also provided a sequencing
model of teaching lessons (= Artikulation) that still
receives a lot of attention in the European and USAmerican literature. Instructional science also refers
to this model since it is based on a more psychological
(i.e., cognitive) approach to learning. Herbart’s successors (i.e., the Herbartians, e.g., Tuiskon Ziller, Wilhelm
Rein) watered down these steps to form a rigid
sequencing of lessons. Herbart’s notion of schooling
as “general education” (= [Allgemein-]Bildung) also
provides a basis for the comprehensive concept of
“general didactics.”
Otto Willmann (1889) made the fruitful distinction
between “content of education” (= Bildungsinhalt) and

Didactics, Didactic Models and Learning

“substance of content” (= Bildungsgehalt). He
recommended including only topics that are educationally valuable (= content of education) in the syllabus, because learning about those topics draws on their
inherent “substance of content.”
Erich Weniger, an academic disciple of Herman
Nohl whose name is associated with the Human Science
Theory of Education in the first part of the twentieth
century, outlined a comprehensive theory of syllabus
development as a societal task. Weniger (1930) defined
didactics as the theory of educational content and
syllabus.
Referring to Willmann and Weniger, Wolfgang
Klafki (1959) framed his concept of Categorical Education (= Kategoriale Bildung), which he set up against
the restrictions of both “formal education” (i.e.,
enhancing the methods of learning and thinking) and
“material education” (i.e., covering a comprehensive
range of topics). The Exemplar Approach (Klafki
1991) referred to Bruner’s concept of “learning by
discovery” and thereby also applied a more psychological concept of transfer. It can be argued that Klafki’s
(1994) notion of “Bildung” as the potential for selfdetermination, codetermination (= participation),
and solidarity has some strong links to generalized
learning results as it is captured in both the process of
transfer and competency constructs (Arnold 2007).
Weniger and his academic disciple Klafki are
regarded as the founders of the Bildung-Centered
Approach to Didactics. In his famous paper “Didactic
analysis as the core of preparation of instruction,”
Klafki (1958/1995) succeeded in transposing Weniger’s
basic ideas of syllabus decision making to the level of
lesson preparation. Teachers are considered as partly
autonomous educators who are responsible for the
children taught in their classrooms and therefore
must reflect critically on the syllabus. The core questions of the didactic analysis refer to the (2) current and
the (3) future significance of the chosen content for the
students, which means that teachers should reflect on
the educational substance of the mandatory ▶ curriculum. Question 1 focuses on the aforementioned exemplarity of the chosen topic, question 4 on its
structuredness, and question 5 on the accessibility of
the topic (= methodical aspects of teaching).
In the 1950s and 1960s, Paul Heimann, professor at
the teacher education college of West Berlin, devised
the Learning-Centered Approach to General Didactics,

D

putting it against the emerging tradition of the
Bildung-Centered Approach. He argued that university
teacher education and its newly introduced internships
should be based on scientific analysis and therefore
should rely on empirical learning theories. This
approach was further developed by Heimann’s student
Wolfgang Schulz, who established the more comprehensive Berlin Model of Lesson Planning (1965), which
has – as an alternative to Klafki’s model – found wide
use in German teacher education. Schulz discerned
four fields of decision making: (1) intentions; (2) themes,
topics; (3) methods; and (4) media. The usage of this
fourfold scheme should be guided by three principles of
lesson planning: (a) interdependency of the fields, (b)
variability of the lesson plan, and (c) controllability
(analysis of differences between the lesson plan and
the lesson delivered). Two fields of conditions are to be
considered: (a) sociocultural and (b) individual conditions of the students, which involve both adaptivity to
their prerequisite knowledge and a look at the societal
demands of education.
After moving to the University of Hamburg, Schulz
(1980) revised his model and called it the Hamburg
Model of Lesson Planning. Partly adopting Klafki’s
work, Schulz focused more on teachers’ critical reasoning on the societal dimension of schooling. Schulz also
emphasized the interactive nature of classroom teaching by introducing Cohn’s Theme-Centered Interaction (TCI) into his model. In accordance with
curriculum theory, he distinguished between three
levels of educational objectives (Möller 1973):
(1) far-reaching educational goals, (2) broadly defined
educational objectives, and (3) narrowly defined
educational objectives. The latter introduced Robert
Mager’s (1962) method of operationalization, which has
been criticized as narrowing the educative (= Bildungcentered) purposes of schooling.
Around the beginning of the 1970s, a controversial
debate on the two major models was going on in
Germany, although Klakfi as well as Schulz never saw
them as being opposed to each other. Under the
impression of the 1968 student protest movement,
Klafki further developed his model to the Perspective
Schema of Lesson Planning (1980, 1994) and formulated
its background in the Critical Constructive Approach to
Didactics (Klafki 1985, 1994). In doing so, he renewed
the traditional concept of general education by relating
it to the Critical Theory of Society of the Frankfurt

987

D

988

D

Didactics, Didactic Models and Learning

School of social research (e.g., Horkheimer,
Habermas). The model established emancipation as
the central goal of education, encompassing the three
facets of “Bildung.” He also incorporated some central
features of Schulz’s model (accessibility and ways of
presentation, e.g., choice of media; evidence of understanding and learning outcomes, e.g., assessment of
learning outcomes).
General didactics and its notion of both general
education and interdisciplinary instruction received
a general thematic framework in Klafki’s
(1994) Epoch-Making Key Problems of the Modern
World (e.g., the question of peace; the environment;
the unequal distribution of wealth, employment, and
unemployment; freedom and participation; the relationship between the generations and between men
and women; interaction with those who have special
needs and with ethnic minorities). It has been argued
(Arnold and Koch-Priewe 2011) that teaching subject
matter with regard to key problems also contributes to
a worldwide curriculum and provides a concept of
global education.
A more psychological approach to general didactics
has been developed by the Swiss researcher Hans Aebli
(1983/1998), a student of the developmental psychologist Jean Piaget. One main feature of his work was the
identification of 12 teaching methods (e.g., narration
and reporting, establishment of a concept, flexible
work-through, practice and rehearsal). Aebli did not
refer to the more content-focused approaches of his
German colleagues Klafki and Schulz, who likewise
ignored Aebli’s work. Rudolf Messner and Kurt Reusser
(2006), academic disciples of Aebli, broadened his
approach. Aebli’s constructivist understanding of
learning has been further developed by Oser and
Baeriswyl (2001), who initially distinguished 12 (and
later even more) basis models of learning scripts that are
meant to explain why aspects of the sight structure of
teaching are effective.
Since the 1990s, constructivist conceptions have
received a lot of attention in subject-matter didactics,
especially in mathematics and science education. Reich
(2006) framed a constructivist approach to general
didactics that expanded on Dewey’s notion of learning
experience, which also provides one of the two basic
concepts of a new approach to general didactics developed by Meinert Meyer (2007) and colleagues. They
placed the notion of developmental tasks in psychology

in relation to the aims of general education, which call
for lesson planning to be individualized, i.e., adapted to
the needs and experiences of each student.

Important Scientific Research and
Open Questions
There is not much empirical research on the concepts
of general didactics; their explanatory power in
research on teaching and curriculum development has
scarcely been analyzed at all. It could be shown, however, that all-day lesson planning by teachers is at least
implicitly in line with the basic concepts of Klafki’s
didactic analysis (Koch-Priewe 2000) and that classroom learning tasks can be categorized by Bildungcentered concepts (Bloemeke et al. 2006).
A lot of textbooks on general didactics have been
published in the German language; almost all of them
tackle the two major models described above. It is
worth noting that neither Schulz nor Klafki wrote
monographs on their approaches. Some of Klafki’s
articles have been translated into English and other
languages; his work is well known in Scandinavia, Finland, and Japan. The work of Schulz remains accessible
only to German-speaking scholars and teachers.
Yet some remarkable writings on the German tradition of general didactics are available in English. Oser
and Baeriswyl (2001) wrote a chapter in the third
edition of the famous Handbook of Research on Teaching. Westbury and Hopmann (2000) provided an
anthology on the “German Didaktik Tradition”
containing translations of some of Weniger’s and
Klafki’s central articles. In the area of curriculum theory, some US (e.g., Deng and Luke 2008) and Finnish
authors (e.g., Uljens 1997) have highlighted the profound significance general didactics attaches to lesson
planning. A subdivision of the European Educational
Research Association (EERA), “Didactics – Learning
and Teaching,” was founded in 2005. Arnold and
Koch-Priewe (2011) suggested a merger and integration of the didactics tradition with empirical instructional research.
Some crucial problems still remain under debate:
(a) How do the concepts of syllabus and curriculum
relate to education standards that rely on competency
descriptions? (b) What concepts can be shared by general and subject-matter didactics? (c) How much can
instructional effectiveness be improved by promoting
teachers’ planning abilities?

Didactics, Didactic Models and Learning

Cross-References
▶ Adaptive Instruction Systems and Learning
▶ Aligning the Curriculum to Promote Learning
▶ Content-Area Learning
▶ Curriculum and Learning
▶ Generative Teaching: Improvement of Generative
Learning

References
Aebli, H. (1998). Zwölf Grundformen des Lehrens. Eine allgemeine
Didaktik auf kognitionspsychologischer Grundlage [Twelve basic
forms of teaching. An approach to General Didactics founded on
Cognitive Psychology; 1st ed.: 1983] (10th ed.). Stuttgart: KlettCotta.
Arnold, K.-H. (2007). Generalisierungsstrukturen der kategorialen
Bildung aus der Perspektive der Lehr-Lernforschung [Generalizing structures of Categorical Education: The view of empirical
research on learning and instruction]. In B. Koch-Priewe, F.
Stübig, & K.-H. Arnold (Eds.), Das Potenzial der Allgemeinen
Didaktik (pp. 28–42). Weinheim: Beltz.
Arnold, K. H., Koch-Priewe, B. (2011). The merging and the future of
the classical German traditions in General Didactics: A comprehensive framework for lesson planning. In B. Hudson, M. A.
Meyer (Eds.), Beyond fragmentation: Didactics, learning and
teaching in Europe (pp. 252–264). Opladen: Budrich.
Deng, Z., & Luke, A. (2008). Subject matter: Defining and theorizing
school subjects. In F. M. Connelly, M. F. He, & J. Phillion (Eds.),
The Sage handbook of curriculum and instruction (pp. 66–87).
Los Angeles, CA: Sage.
Klafki, W. (1959). Das pädagogische Problem des Elementaren und die
Theorie der kategorialen Bildung [The pedagogical problem of the
elementary and the theory of categorical eduction]. Weinheim:
Beltz.
Klafki, W. (1985). Grundlinien kritisch-konstruktiver Didaktik [Basic
aspects of the Critical-Constructive Approach to General Didactics]. In W. Klafki (Ed.), Neue Studien zur Bildungstheorie und
Didaktik (pp. 31–86). Weinheim: Beltz.
Klafki, W. (1991). Exemplar approach. In A. Lewy (Ed.), The international encyclopedia of curriculum (pp. 181–182). Oxford:
Pergamon.
Klafki, W. (1994). Grundzüge eines neuen Allgemeinbildungskonzepts. Im Zentrum: Epochaltypische Schlüsselprobleme
[Essential features of a new concept of general education. In
focus: epoch-making key problems]. In W. Klafki (Ed.), Neue
Studien zur Bildungstheorie und Didaktik. Zeitgemäße
Allgemeinbildung und kritisch-konstruktive Didaktik (4th rev.
ed., pp. 43–82). Weinheim: Beltz.
Klafki, W. (1980/1994). Zur Unterrichtsplanung im Sinne kritischkonstruktiver Didaktik [Lesson planning in the critical-constructive approach to general didactics; first published: 1980].
In W. Klafki (Ed.), Neue Studien zur Bildungstheorie und Didaktik
(4th ed., pp. 251–284). Weinheim: Beltz.
Koch-Priewe, B. (2000). Zur Aktualität und Relevanz der
Allgemeinen Didaktik in der LehrerInnenausbildung [The

D

relevance of General Didactics to teacher education]. In M.
Bayer, F. Bohnsack, B. Koch-Priewe, & J. Wildt (Eds.), Lehrerin
und Lehrer werden ohne Kompetenz? Professionalisierung durch
eine andere Lehrerbildung (pp. 148–170). Bad Heilbrunn:
Klinkhardt.
Mager, R. F. (1962). Preparing objectives for programmed instruction.
San Francisco: Fearon.
Messner, R., & Reusser, K. (2006). Aeblis Didaktik auf psychologischer
Grundlage im Kontext der zeitgenössischen Didaktik [Aebli’s psychologically based approach to General Didactics in the context
of contemporary theories on Didactics]. In M. Baer, M. Fuchs,
P. Füglister, K. Reusser, & H. Wyss (Eds.), Didaktik auf
psychologischer Grundlage (pp. 52–73). Bern: h.e.p. Verlag.
Meyer, M. A. (2007). Didactics, sense making, and educational
experience. European Educational Research Journal, 6(2),
161–173.
Möller, C. (1973). Technik der Lernplanung. Methoden und Prinzipien
der Lernzielerstellung [Techniques of planning learning. Methods
and principles of preparing learning objectives] (4th ed.).
Weinheim: Beltz.
Oser, F. K., & Baeriswyl, F. J. (2001). Choreographies of teaching:
Bridging instruction to learning. In V. Richardson (Ed.), Handbook of research on teaching (4th ed., pp. 1031–1065). Washington, DC: American Educational Research Association.
Schulz, W. (1965). Unterricht – Analyse und Planung [Analysis and
planning of lessons]. In P. Heimann, G. Otto, & W. Schulz (Eds.),
Unterricht – Analyse und Planung (pp. 13–47). Hannover:
Schroedel.
Schulz, W. (1980). Ein Hamburger Modell der Unterrichtsplanung.
Seine Funktion in der Alltagspraxis [The Hamburg model of
lesson planning: Its functioning in everyday teaching]. In B.
Adl-Amini (Ed.), Didaktische Modelle und Unterrichtsplanung
[Models of general didactics and lesson planning] (pp. 49–87).
Weinheim: Juventa.
Weniger, E. (1930/2000). Didaktik as a theory of education [1st ed.:
Theorie der Bildungsinhalte und des Lehrplans. Weinheim: Beltz,
1930]. In I. Westbury, S. Hopmann, & K. Riquarts (Eds.), Teaching as a reflective practice. The German Didaktik tradition (pp.
111–126). Mahwah: Erlbaum.
Westbury, I., Hopmann, S., & Riquarts, K. (Eds.). (2000). Teaching as
a reflective practice. The German Didaktik tradition. Mahwah, NJ:
Erlbaum.
Willmann, O. (1889/1967). Didaktik als Bildungslehre nach ihren
Beziehungen zur Socialforschung und zur Geschichte der Bildung
[Didactics as a theory of education according to its relations to
social research and to the history of education; 1st ed.: Braunschweig: Vieweg, 1889]. (7th ed.). Freiburg: Herder.
Uljens, M. (1997). School didactics and learning. A school didactic
model framing an analysis of pedagogical implications of learning
theory. Hove: Psychology Press.
Further Reading
Hopmann, S., & Keitel, C. (1995). Editorial: The German Didaktik
tradition. Journal of Curriculum Studies, 27(1), 1–2.
Hudson, B., Buchberger, F., Kansanen, P., & Seel, H. (Eds.). (1999).
Didaktik/Fachdidaktik as science(s) of the teaching profession?
(TNTEE Publications Vol. 2, No. 1; available under http://tntee.

989

D

990

D

Difference

umu.se/publications/publication2_1.html, 2010-12_14, accessed
14 Dec 2010). Umeå: TNTEE, University of Umeå.
Kansanen, P. (2002). Didactics and its relation to educational psychology: Problems in translating a key concept across research
communities. International Review of Education, 48(6), 427–441.
Klafki, W. (1995). Didactic analysis as the core of preparation of
instruction. Journal of Curriculum Studies, 27(1), 13–30.

criminology has had impact on how people reflect on
crime as has differential association. In the words of
Cressey (1952), differential association is the most outstanding sociological formulation of a general theory of
crime causation.

Theoretical Background

Difference
▶ Simultaneous Discrimination Learning in Animals

Differential Access to Learning
Skills
▶ Differential Association Theory

Differential Association Theory
THOMAS ANTWI BOSIAKOH
Department of Sociology, University of Ghana, Legon
Accra, Ghana

Synonyms
Differential access to learning skills; Differential association to learning skills

Definition
Differential association is a crime predictive theory. It
can be defined as a process by which individuals come
to have differential access to criminal values through
interaction with other people. The theory holds that,
criminal behavior is learned in the same way that lawabiding values are learned, and that, this learning activity is accomplished, in interactions with others, and the
situational definitions we place on the values. The
theory can be reduced to the notion that, people
become criminals because they associated with, and
absorbed pro-criminal definitions. Differential association has attracted more attention, over a longer period
of time, than any other criminological theory.
According to some scholars, no single idea in modern

Edwin H. Sutherland was one of the sociologists from
the famous Chicago school. In the 1920s and 1930s, the
study of crime was almost like a tale of one city, Chicago. At the University of Chicago, sociologists devoted
enormous attention to the study of crime. A general
conclusion easily gleaned from these studies is their
attribution of crime causation to external factors.
Shaw and McKay for example attributed crime to social
disorganization. Sutherlands criticized Shaw and
McKay’s concept of social disorganization as
a variable for crime causation, and replaced it with his
own term, the differential social organization. It was
through this concept of differential social organization
that Sutherland developed his differential association
theory. It is also true from the work of Sutherland that,
differential association theory was developed in an
attempt to explain career criminal behavior.
Sutherland first presented differential association
theory in 1939, and in 1947, revised it. The theory
consists of nine principles, as outlined below:
1. Criminal behavior is learned; it is not inherited.
With this principle, Sutherland rebuffed the argument that, crime was the outcome of social disorganization. He also rejected the view that, criminals
were biologically different from noncriminals.
Sutherland sought to explain with this principle
that, a person who has not been trained in criminal
acts does not invent such acts, just as a child does
not make courteous remarks unless they have been
socialized as such.
2. Criminal behavior is learned in interaction with
others through communication. Sutherland
suggested with this principle that, criminal behavior is acquired through association with others
which also includes communication. The use of
communication here refers to the sum total of
interactions. This communication is verbal in
many respects but includes also the communication
of gestures often described as nonverbal
communication.

Differential Association Theory

3. The principal part of learning criminal behavior
occurs in intimate groups. Sutherland argued that,
only small, face-to-face groups influence behavior.
For this reason, he placed enormous emphasis on
peer and family groups as the most likely sources of
initiation into delinquent values and activities.
From this analysis, Sutherland discounted impersonal agencies of communication such as picture
shows and newspapers as part of the process for the
initiation into criminality.
4. When criminal behavior is learned, the learning
includes (a) techniques, which are sometimes complicated, and sometimes very simple; and (b) the
specific direction of motives and drives,
rationalizations and attitude. By this principle,
Sutherland explained the specific items that go
into learning criminal behavior including the skill
(procedure/method), motivation (the driving
force), justification (reasoning), and the general
behavioral outlook that supports criminality.
5. The specific direction of motives and drives are
learned from definitions of legal codes as favorable
or unfavorable. Again, the learning here is
influenced by other people (mainly those with
whom one has intimate relations). If such people
define the law as deserving/nondeserving to be
observed, corresponding attitudes toward the law
are learnt.
6. A person becomes criminal because of excess of
definitions favorable to the violation of law over
definitions unfavorable to the violation of law.
This principle can be expressed as a ratio between
favorable definitions and unfavorable definitions to
the violation of criminal law. If the ratio is toward
favorable definitions, violation of the law will be
impeded, and if it is toward unfavorable definitions,
the person will violate the law.
7. Differential association (tendency toward criminality) varies in frequency, duration, priority, and
intensity. This means that, the earlier in one’s life,
the longer the time, the more intensely and more
frequently people are exposed to a set of attitudes
about criminality, the more likely they will become
criminals themselves.
8. The process of learning criminal behavior involves
the same (all the) mechanisms involved in any other
learning. This means that, the mechanisms for
learning criminal behaviors are the same as those

D

for noncriminal behaviors/any law-abiding behavior and even social skills. However, the content and
motive of what is learnt are entirely different in the
two situations.
9. Both criminal and noncriminal behaviors are
expressions of the same needs and values. Put differently, the goals of criminals and noncriminals are
usually the same. What is different is the means they
adopt to pursue this goal. For instance, thieves
generally steal in order to secure money but honest
citizens also work for it.

Important Scientific Research and
Open Questions
Since 1947 when differential association theory was
reformulated, a number of attempts have been made
to empirically test and or apply the theory to different
criminal behaviors (see Antwi Bosiakoh and Andoh
2010, pp. 200–201). Perhaps the earliest of these was
Donald Cressey’s application and verification attempt
in 1952. Since then, several studies have been undertaken for empirical clarification (Tittle et al. 1986),
to test the theory as a hypothesis (Short 1960) and to
verify the applicability of the theory in relation
to specific criminal behaviors (Glaser 1960; Voss 1964;
Antwi Bosiakoh and Andoh 2010). While many of these
studies found empirical application for the main argument of differential association theory, that criminality
is learned, and in most cases found the theory to be
superior to alternative theories on crime causation and
crime prediction, some of these studies have suggested
modifications in the principles and even the term “differential association.” Burgess and Akers (1966) have
provided a reformulated version of differential association theory with the view to incorporate reinforcement theory. In the process, they reduced the nine
principles of differential association theory into seven
and renamed the theory as differential associationreinforcement theory of criminal behavior (Burgess
and Akers 1966, cited in Antwi Bosiakoh and Andoh
2010). While arguing that there have been renewed
interest in differential association, some scholars have
also
suggested
differential
identification
as
a modification to differential association to
reconceptualize Sutherland’s theory in role construction and role reconstruction imageries (Matthews
1968; Glaser 1956, cited in Antwi Bosiakoh and
Andoh 2010). In 1960, Daniel Glaser acknowledged

991

D

992

D

Differential Association to Learning Skills

that, differential association is superior to alternative
theories of criminological prediction, but quickly
suggested that differential anticipation theory would be
more appropriate and adequate than differential association (Glaser 1960, p. 13).

Cross-References
▶ Social Learning Theory
▶ Sutherland, Edwin H. (1883–1950)
▶ Value Learning

References
Antwi Bosiakoh, T., & Andoh, P. (2010). Differential association
theory and juvenile delinquency in Ghana’s capital city – Accra:
The case of Ghana Borstal Institute. International Journal of
Sociology and Anthropology, 2(9), 198–205.
Burgess, R. L. & Akers, R. L. (1966). Differential association-reinforcement theory of criminal behavior. Social Problems, 14(2),
128–147.
Cressey, D. R. (1952). Application and verification of the differential
association theory. The Journal of Criminal Law, Criminology,
and Police Science, 43(1), 43–52.
Glaser, D. (1956). Criminal Theories and Behavioral Images. American Journal of Sociology, 61, 433–444.
Glaser, D. (1960). Differential association and criminological prediction. Social Problems, 8(1), 6–14.
Matthews, Victor M. (1968). Differential identification: An empirical
note. Social Problems, 15(3), 376–383.
Short, J. F., Jr. (1960). Differential association as a hypothesis: Problems of empirical testing. Social Problems, 8(1), 14–25.
Tittle, C. R., Burke, M. J., & Jackson, E. F. (1986). Modeling
Sutherland’s theory of differential association: Toward an empirical clarification. Social Forces, 65(2), 405–432.
Voss, H. L. (1964). Differential association and reported delinquent
behavior: A replication. Social Problems, 12(1), 78–85.

Differentiation
The process by which a schema can be split into two
new schemas. This happens when a schema attempts to
assimilate a situation which requires considerable
adjustment of the schema. The result is the generation
of a new schema which is more appropriate to the new
situation, while the old schema is refined (made more
specific) so that it no longer assimilates the new
situation.

Cross-References
▶ Generalization Versus Discrimination

Difficulty Level of Authentic
Listening Input
▶ Effects of Task
Comprehension

Difficulty

in

Digit
▶ Learning Numerical Symbols

Digital Learning
Differential Association to
Learning Skills

▶ Interactive Learning Environments
▶ Neural Network Assistants for Learning

▶ Differential Association Theory

Digital Literacy
▶ General Literacy in a Digital World

Differential Conditioning
Classical or operant conditioning in which different
stimuli are paired with different outcomes. A form of
discrimination learning since subjects learn to respond
differently to the different stimuli.

Dilatory Behavior
▶ Procrastination and Learning

Listening

Directed Forgetting

Dimension of Movement
▶ Impaired
Learning

Multidimensional

Motor

Sequence

Direct Memory Test
▶ Recall and the Effect of Repetition on Recall

Directed Forgetting
COLIN M. MACLEOD
Department of Psychology, University of Waterloo,
Waterloo, ON, Canada

Synonyms
Instructions to forget; Intentional forgetting; Motivated forgetting; Voluntary forgetting

Definition
Directed forgetting is an experimental procedure developed in the late 1960s as an analog to the normal
updating of memory. Essentially, individuals are told
that they can forget some of the information being
presented to them. This is done in one of two ways.
In the item method, an instruction to remember or to
forget is given immediately after each presented item.
In the list method, a single instruction is given half way
through the list of items either to forget or to continue
remembering the first half of the list. Contrary to
instruction, under both methods, memory for both
to-be-remembered items (R items) and to-beforgotten items (F items) is ultimately assessed. The
standard finding is poorer memory for the F items than
for the R items – the directed forgetting effect.

Theoretical Background
In the beginning, four quite intuitive ideas were proposed to explain how people intentionally forget specified information. The most obvious was the erasure
hypothesis – that we essentially delete information that
we are told to forget (see the first directed forgetting

D

study by Muther 1965). Although this idea made some
sense when applied to very small sets of information
that could be held in or dropped from working (shortterm) memory, it did not make sense when applied to
larger sets in long-term memory. Moreover, empirical
work quickly showed that F items were not gone from
memory. Consequently, the erasure hypothesis was
quickly discarded.
Muther also considered what he called the
partitioning hypothesis, which soon became better
known as the set differentiation or selective search
hypothesis. Here, the idea was that, at the time of
study, the items were sorted into two bins corresponding
to the instructions (i.e., an F bin and an R bin). Then, at
the time of test, when an individual tried to retrieve the
studied items, priority was given to searching in the
R bin, which favored remembering the R items. In the
early days of research on intentional forgetting, this
retrieval-based account became one of the two leading
explanations for the directed forgetting effect.
The third account that was put forth early on was
the inhibition or repression hypothesis. Here, the core
idea was that F items were suppressed at the time of
study – essentially, their activation in memory was
reduced. Then, at the time of test, when the individual
was trying to remember, the F items were less likely to
be retrieved. The inhibition hypothesis clearly is related
to the selective search hypothesis in that both involve
operations carried out at the time of study that are then
influential at the time of test. Yet the inhibition view fell
quickly into disfavor and disappeared from the theoretical landscape for quite some time, with selective
search becoming the dominant retrieval-based account.
The fourth account focused on encoding – what
happened at the time of study – and grew largely out
of experiments that used the item method. This was the
selective rehearsal account (see Bjork 1972). The argument was that each item was held in abeyance in working memory until the instruction applied to that item
was presented; as little processing as possible was
undertaken on an item prior to instruction. Then, if
the instruction was to remember, active rehearsal of
that R item was undertaken; if the instruction was to
forget, that F item received no further processing. The
selective rehearsal account and the selective search
account became the principal theoretical combatants
in the early days. In part, this may have stemmed from
these two views fitting both the working memory and

993

D

994

D

Directed Forgetting

the long-term memory studies then ongoing in investigations of intentional forgetting.
This was the state of theoretical affairs from the late
1960s until the late 1980s, although throughout this
period the selective rehearsal account usually was seen
as the preferred explanation. This preference likely
rested on three factors: (1) that rehearsal was a more
established process in the broader memory literature,
and one that seemed more observable, (2) that most of
the studies used the item method, for which the
rehearsal account was optimally suited, and (3) that
emphasis had shifted primarily to studies involving
long-term memory. But then, coincident with the rise
in studies using the list method, the inhibition hypothesis made a comeback (see Bjork 1989), displacing the
selective search hypothesis as the primary alternative to
the selective rehearsal hypothesis.
Most recently, a new account has emerged. Like the
inhibition hypothesis, the contextual change hypothesis
(Sahakyan and Kelley 2002) is particularly well matched
to the experiments using the list method. The idea is that
context plays a key role in the directed forgetting effect as
it is known to do in many memory phenomena – indeed,
memory can be seen in general as exquisitely contextual.
Under this view, a context is in place as the first half of
the list of items is presented. Then an instruction to
forget those items is delivered, which disrupts the initial
context such that a new context is established. The
remainder of the list is then presented and studied
under this new context. So the initial F items are linked
to one context and the subsequent R items are linked to
another context. Because there is no context disruption
as the individual segues from study to test, the R items
have the advantage of a context match between study
and test, whereas the F items have the disadvantage of
a context mismatch. This discrepancy, it is argued, produces the directed forgetting effect.

Important Scientific Research and
Open Questions
From the beginning of research on directed forgetting,
there had been a key empirical mystery (see MacLeod
1998, for a review). This pertained to the type of test
used to assess the Remember–Forget manipulation.
Researchers had long reported that a recall test –
where the individual must try to recover the items
from memory unaided – always revealed a directed
forgetting effect. But this regularity was absent on

a recognition test – where the individual is shown
some items that were studied and some that were not
and is asked to indicate for each item whether it was in
fact studied. On recognition tests, the directed forgetting effect was sometimes present and sometimes
absent. Finally, in the late 1980s/early 1990s,
a solution to the mystery was offered: Recognition
showed a quite consistent directed forgetting effect
when the item method was used, but the effect rarely
appeared when the list method was used.
This empirical resolution led to the proposal that
the two methods were perhaps not just minor variations on each other. Basden et al. (1993) suggested that
selective rehearsal was responsible for the directed forgetting effect under the item method but that inhibition was responsible for the effect under the list
method. Selective rehearsal was seen as eminently possible when instructions were provided on an item-byitem basis, but as not possible when a quite large set of
items preceded a single forget cue, as in the list method.
The idea was that the long list prior to the instruction
in the list method would result in F items already
having been rehearsed, making it too late for selective
rehearsal to operate successfully. Therefore, another
factor would have to come into play to produce list
method directed forgetting: The prime candidate was
inhibition.
It is important to note that the directed forgetting
effect is not due to demand characteristics, where the
argument might be made that people simply do not try
to recall the F items, essentially doing what they think
the experimenter wants them to do. In the item
method, even when the instructions themselves cannot
be remembered, there is still a directed forgetting effect.
And enticing people to try to recall more F items, such
as by offering incentives (e.g., money for F items but
not for R items), does not diminish the effect. It would
appear that memory is indeed not as good for F items
as it is for R items.
The directed forgetting effect appears to be limited to
explicit memory tests such as recall and recognition,
where remembering is done consciously. On implicit
memory tests – such as completing partial words that
were or were not studied, or reading aloud as quickly
as possible words that were or were not studied – the
evidence is quite consistent that there is no difference
between F items and R items. As well, the effect disappears
if the F items are meaningfully linked to the R items. Such

Discontinuities for Mental Models

results implicate conscious encoding and retrieval processes as the locus of the directed forgetting effect.
There is now reasonable consensus that selective
rehearsal during study underlies the directed forgetting
effect when the item method is used (see MacLeod
1998). Debate centers on the best explanation for the
effect when the list method is used. Are the F items
inhibited or do they suffer from a contextual mismatch
between study and test? Both mechanisms, instituted
during the study phase, would make later retrieval of
F items less successful than that of R items. (Indeed, it is
even conceivable that a modified rehearsal account
could explain list method directed forgetting: Given
the smaller effect under the list method, it might be
that there is selective rehearsal of only some of the
items, such as just those few immediately preceding
the mid-list instruction to forget.)
In the end, “directed forgetting” may well be a
misnomer. For the item method, it would appear that
people simply do not learn the F items as well as the
R items: Obeying instructions, they give F items less
attention and less rehearsal, resulting in weaker learning. As for the list method, although its explanation is
more debatable, both the context change account and
the inhibition account rely at least in part on actions
taking place during encoding that therefore influence
learning. To the extent, then, that directed forgetting is
an encoding effect under both methods, it might better
be labeled “directed learning,” where the instructions
result in more (R items) or less (F items) learning of the
material being studied.

D

Bjork, R. A. (1989). Retrieval inhibition as an adaptive mechanism in
human memory. In H. L. Roediger III & F. I. M. Craik (Eds.),
Varieties of memory and consciousness: Essays in honour of Endel
Tulving (pp. 309–330). Hillsdale: Lawrence Erlbaum.
MacLeod, C. M. (1998). Directed forgetting. In J. M. Golding & C. M.
MacLeod (Eds.), Intentional forgetting: Interdisciplinary
approaches (pp. 1–57). Mahwah: Lawrence Erlbaum.
Muther, W. S. (1965). Erasure or partitioning in short-term memory.
Psychonomic Science, 3, 429–430.
Sahakyan, L., & Kelley, C. M. (2002). A contextual change account of
the directed forgetting effect. Journal of Experimental Psychology.
Learning, Memory, and Cognition, 28, 1064–1072.

Direction
▶ DICK Continuum in Organizational Learning
Framework

Disciplinary Identity
▶ Identity and Learning

Disciplinary Literacy
▶ Content-Area Learning

Cross-References
▶ Episodic Learning
▶ Inhibition and Learning
▶ Intentional Learning
▶ Memory Consolidation and Reconsolitation
▶ Retention and Learning
▶ Selective Attention in Learning
▶ Verbal Learning

References
Basden, B. H., Basden, D. R., & Gargano, G. J. (1993). Directed
forgetting in implicit and explicit memory tests: A comparison
of methods. Journal of Experimental Psychology. Learning, Memory, and Cognition, 19, 603–616.
Bjork, R. A. (1972). Theoretical implications of directed forgetting. In
A. W. Martin & E. Melton (Eds.), Coding processes in human
memory (pp. 217–235). Washington, DC: Winston.

Discontinuities for Mental
Models
SUSANNE PREDIGER
IEEM – Institute for Development and Research in
Mathematics Education, TU Dortmund University,
Dortmund, Germany

Synonyms
Cognitive dissonances; Incoherences as driving forces
for concept development

995

D

996

D

Discontinuities for Mental Models

Definition
Along with other types of cognitive structures (such as
schemes), a mental model can be thought of as
a specific representation in the human mind of an
individual’s experience or insight. One substantial
challenge and catalyst for constructing adequate mental models are discontinuities of conceptual models.
A conceptual model (i.e., the target model being the
prescriptive counterpart of a mental model) is defined
to have a discontinuity, when an extension of its scope
of application necessarily implies changes in properties
or meanings. These discontinuities in the subject matter structure can offer obstacles and opportunities for
students’ development of conceptual understanding by
altering their mental models.

Theoretical Background
Constructing adequate mental models and schemata is
a major goal for subject-matter learning processes in
many domains, for example science education or
mathematics education. Individuals construct new
mental models in situations of accommodation, when
an experience cannot be assimilated to existing schemata (see entry ▶ Mental Model or Schema Construction). Mental models and schemata in the individual
cognitions of learners are said to be adequate if they are
congruent with corresponding conceptual models in
the subject-matter domain. This definition uses
a distinction between mental models (in a descriptive
mode as those being cognitively constructed by individuals) and conceptual models (in a prescriptive mode
as those intended to be constructed from an instructional perspective) (Seel 2003).
One substantial catalyst and challenge for
constructing adequate mental models (among others)
are discontinuities of conceptual models. This theoretical
construct offers a subject-related background for locating and explaining conceptual challenges that learners
encounter in longer-term subject-matter learning processes, as they are often described in conceptual change
approaches (Prediger 2008). Typical discontinuities
appear, for example in mathematics, when extending
number domains from natural to rational numbers.
For example, a fraction does not have a unique neighbor on the number line whereas every natural number
has. In the transition from natural to rational numbers,
students have to restructure their order schema for
numbers and have to build a mental model on the

density property of fractions and decimals and realize
the limit of scope of their former mental model: The
idea of neighborhood must be constrained to natural
numbers (Vosniadou and Verschaffel 2004).
In order to explain the sources of difficulties,
researchers in domain-specific education research
areas focus on epistemologically specific mental
models, namely, those that concern the meaning of
domain-specific phenomena or concepts (e.g., the
interpretation of a mathematical concept or biological
phenomena in real-world situations). They are important because the mental construction of adequate
meanings is a precondition for dealing with domainspecific concepts adequately, flexibly, and with conceptual understanding (see entry ▶ Mathematical Learning). In mathematics education research, conceptual
models concerning meaning have been conceptualized
as “Grundvorstellungen” (vom Hofe 1998) or have
simply been termed as “models,” defined in the epistemologically restricted way as “meaningful interpretation of a phenomenon or concept” (Fischbein 1989, p.
12).

Important Scientific Research and
Open Questions
The impact of discontinuous conceptual models on
students’ thinking can be illustrated by an often
documented prototypical example, namely, the case
of multiplication of numbers. For natural numbers,
multiplication always makes bigger, but when multiplication is extended to rational numbers, the property
is not satisfied anymore. This discontinuity may cause
difficulties for students when they mathematize realworld situations. For example, when calculating the
cost of 0.7 kg potatoes at £1.50 per kg, many students
chose the operation division, because for them, multiplication appears to make the 1.50 bigger, but the cost is
supposed to be less (see Bell et al. 1981). In this case, the
wrong order property is used as a non-adequate
schema that guides the (non-adequate) operation
choice. Researchers have suggested to confront the
students’ wrong schema with new examples and to
initiate processes of accommodation, but in many
cases, the construction of new mental models seems
to be not stable. Other studies have shown that it is
crucial to consider the discontinuity not only on the
level of order properties alone but on the level of
meaning (Prediger 2008): It is the most important

Discourse

conceptual model on the meaning of multiplication,
the repeated addition model, that does not apply for
rational numbers between 0 and 1. An adequate mental
model for multiplication making bigger or not should
include a model for the meaning of multiplication, like
the interpretation of multiplying as scaling up (and
down, for numbers smaller than 1). Students can construct these alternative interpretations of multiplication in suitable learning environments.
The example shows that some cognitive difficulties
in constructing adequate mental models can be traced
back to discontinuities and continuities of meanings as
crucial parts of the inherent structure of the subject
matter. On the other hand, for investigating processes
of mental model construction, discontinuities of conceptual models offer instructive subjects of research.
Whereas some discontinuities of mathematical
properties and propositions have been subject to
research in the conceptual change tradition (see
Vosniadou and Verschaffel 2004), discontinuities of
meanings have only rarely been the subject of empirical
research (one example is the case of the meaning of the
equal sign in the transition from arithmetic to algebra).
In science education, the conceptual change approach
was more consequently applied to discontinuities
between everyday and scientific conceptions, whereas
changes within the physicists’ or chemists’ theory
buildings have – so far – gained less attention. That is
why many important research questions in all these
areas are still open, for example:
How do individual constructions of mental models
proceed on the microlevel when students encounter
discontinuities? Which conditions in a learning environment and which instructional strategies do facilitate
or hinder successful constructions?
How do constructed models interact with each
other and which mental resources do students go
back to when constructing new mental models?
These questions can be answered on a general level,
but it is sure that the answers will vary in different
subject domains. It is therefore an important task for
domain-specific learning sciences to research for different scientific subjects.

Cross-References
▶ Conceptual Change
▶ Mathematical Learning
▶ Mental Model

D

997

References
Bell, A., Swan, M., & Taylor, G. M. (1981). Choice of operation in
verbal problems with decimal numbers. Educational Studies in
Mathematics, 12, 399–420.
Fischbein, E. (1989). Tacit models and mathematical reasoning. For
the Learning of Mathematics, 9(2), 9–14.
Prediger, S. (2008). The relevance of didactic categories for analysing
obstacles in conceptual change: Revisiting the case of multiplication of fractions. Learning and Instruction, 18(1), 3–17.
Seel, N. M. (2003). Model-centered learning and instruction. Journal
of Technology, Instruction, Cognition and Learning, 1(1), 59–85.
vom Hofe, R. (1998). On the generation of basic ideas and individual
images: Normative, descriptive and constructive aspects. In
A. Sierpinska & J. Kilpatrick (Eds.), Mathematics education as a
research domain: A search for identity. An ICMI study (pp. 317–
331). Dordrecht: Kluwer.
Vosniadou, S., & Verschaffel, L. (2004). The conceptual change
approach to mathematics learning and teaching. Learning and
Instruction, 14(5), 445–548.

Discourse
ANNA SFARD
Department of Mathematics Education,
University of Haifa, Haifa, Israel

Synonyms
Communication (type of); Conversation; Talk

Definition
The English word discourse comes from the Latin
discursus, a derivative of the verb discurrere, to run
about. Initially, the term meant the act, or faculty, of
producing a logical argument in an orderly fashion,
and was therefore almost equivalent to the English
word reasoning. In modern everyday English, discourse
is synonymous with verbal interaction or simply conversation. In modern linguistics, it signifies the unit of
analysis that goes beyond the sentence. In social sciences, the use of the word is polysemic. In some contexts, it signifies a segment of connected speech or
written text produced in the course of verbal interaction. In other instances, it refers to the activity of
communicating rather than to its products. In still
other contexts, it is a higher-order term that designates
a category of specific instances of communicational acts
and products, unified by some common features, such

D

998

D

Discourse

as specialized vocabulary, distinct patterns of communicational actions, the type of statements produced by
participants, etc. In this case, therefore, the word discourse is understood as referring to the type of communication rather than any specific instance of
communicating. One can speak, for example, about
mathematical discourse, a discourse of physics, of history, or of fine arts, as well as about discourses of
different political factions, social classes, and professional or ethnic groups. Although originally associated
mainly with spoken interactions, the word discourse is
increasingly used in the broader context of communication at large, not necessarily verbal, vocal, or
synchronous.

Theoretical Background
In the recent decades, there was a significant increase in
the amount of explicit attention paid by social scientists
to human communication. The change is not purely
quantitative. Rather, the reorientation toward discourse indicates a revolutionary shift not only in
research methods, but also in the epistemological
underpinnings of the contemporary social thought. It
is now widely agreed among social scientists that
understanding human communication, its inner working, and its interaction with other human activities is
crucial to our grasp of all uniquely human phenomena.
In the times of unconstrained connectivity, when
people spend unprecedented proportion of their waking time interacting with others in unprecedented multiplicity of ways, the wide consensus about centrality of
communication is not surprising. The origins of the
discursive turn in social sciences, however, go back
farther than the latest technological advances, to
a number of interrelated developments in the contemporary scientific thought. Probably the most significant
among these developments was the postmodern rejection of the notion of absolute truth, the attainment of
which was the declared aim of positivistic science.
Rather than seeing the products of scientist’s work as
originating in the nature itself and arising in the direct
interaction between the researcher’s mind and reality,
postmodern thinkers began to speak about knowledge
construction as a “conversation of mankind” (Rorty
1979, p. 389), an interpersonal process of telling stories
about the world, accompanied by the constant effort to
refine the forms of communication that made these
stories possible. Many influences came together in the

metaphor of knowledge-as-conversation, among them
Ludwig Wittgenstein’s (1889–1951) criticism of the
positivist science, grounded in his observations on
how our “language games” would often lead us astray;
Alfred Schutz’s (1899–1959) emphasis on the study of
social interactions as a basis for understanding all
human experience; Thomas Kuhn’s (1922–1996) vision
of science as a succession of paradigms, none of which
can aspire to the title of “ultimate one”; and Michael
Foucault’s (1926–1984) work on discourse as the principal arena of all those phenomena that give meaning
to the term social.
The ontological and epistemological upheaval
caused by all these developments had numerous entailments. For many social researchers, the revolutionized
vision of science brought a new message about the
status and methods of their study. If the product of
the researcher’s investigations are narratives about the
world, then by virtue of their having human authors
and addressees, these narratives can make no claims to
“full” objectivity. As cogent as they may appear, they
will always remain contestable and subject to revision.
Moreover, since the protagonists of the researcher’s
stories are themselves active storytellers, the researcher
needs to ask about the status of his or her own narratives versus those offered by the study participants.
Thus, the social scientist who takes discourses as the
object of his or her research does not ask whether
informants’ narratives are “objectively true.” Rather,
the investigator inquires about what makes people say
what they say, how those who speak convince others to
endorse their stories, and how the things said affect
interlocutors’ lives. The search after answers is
supported by recent developments in linguistics,
where the focus has shifted from the language as
a tool, with the single sentence serving as an object of
study, to the study of language in use, with entire
discursive episodes (sequences of utterances) constituting units of analysis. This transformation is particularly
visible in the relatively young but burgeoning domains
of study such as ▶ systemic functional linguistics,
▶ discourse analysis, and ▶ conversation analysis.
As a result of this foundational shift in social studies, discourse began occupying center scene also in the
sciences of learning. In spite, however, of the wide
consensus among researchers about discourses as the
primary medium for studying human cognitive
growth, the central question of the relation between

Discourse

communicating and learning, and even more fundamentally, between communication and thinking, is
rarely addressed by researchers in a direct manner. At
a closer look, a variety of answers seem to be underlying
diverse research efforts. These answers are spanning
a wide range of possibilities, delineated by two extreme
doctrines. On one end of the spectrum, there is the
conviction that thought and communication, although
interrelated and often concomitant, are distinct types
of human activity, with discourse playing the secondary role of the “carrier” of one’s thoughts. The other
extreme is marked by Wittgenstein’s denial of the primacy of thought over speech and by his rejection of the
idea of “pure thought” that would preserve its identity
through a variety of verbal and nonverbal expressions.
This radical position is in concert with the work of Lev
Vygotsky (1896–1934), who illustrated the inseparability of thought (or meaning) and speech by saying that
conducting study of thought independently of the
study of words is comparable to investigating properties of water by focusing separately on hydrogen and on
oxygen.
In spite of the fact that the rejection of the thinking–
communicating dichotomy has been heralded by some
writers as the beginning of the “second cognitive revolution” (Harré and Gillett 1995), the moderate discursive approaches did not disappear from the current
sciences of learning and they seem to coexist peacefully
with the more radical ones. Treating discourse as
a “window” to the contents of human mind is in tune
with the foundational tenets of the traditional cognitive
psychology and with its sustained Cartesian belief in
the ontological uniqueness of mental phenomena. By
equating thinking with communicating and thus disposing of the Cartesian split between body and mind,
adherents of the radical position render the sciences of
learning clear sociocultural slant: they recognize the
fact that even the most private of human activities,
such as thinking or feeling, can be understood only if
conceived as a part of wider collective activities. Here,
learning mathematics or physics becomes one’s
attempt to become a participant of the historically
developed forms of discourse known as mathematics
or physics. As made clear by the word participation,
and as famously stated by Vygotsky, learning originates
on the “social plane” rather than directly in the world.
The diversity of the foundational positions notwithstanding, there seems to be a general consensus that the

D

999

active engagement in conversation with others is
a necessary condition for learning.

Important Scientific Research and
Open Questions
Almost any question about learning can be recast as
a question about discourse. According to the perspective adopted and aspects considered, one can probably
sort all the existing discourse-oriented studies of learning into three thematic strands. The first two of these
distinct lines of research are concerned with different
features of the discourse under investigation and can
thus be called intra-discursive or inward looking. The
third one deals with the question of what happens
between discourses or, more precisely, how interdiscursive relations impact learning.
The first intra-discursively oriented strand of
research on learning focuses on learning–teaching interactions, whereas its main interest is in the impact of
these interactions on the course and outcomes of learning. Seminal events that initiated this type of research
include, among others, the studies that brought to the
fore the ubiquity of ▶ Initiation-Response-Evaluation
(IRE) sequence in a traditional classrooms (see, e.g.,
Mehan 1979) and those that focused on teacher’s discursive routines, such as ▶ revoicing or scaffolding.
Today, when inquiry learning, collaborative learning,
computer-supported collaborative learning, and other
conversation-intensive pedagogies become increasingly
popular, one of the main questions asked by researchers
is that of what features of small group and whole-class
interactions make these interactions conducive to highquality learning. Whereas there is no doubt about theoretical and practical importance of this strand of
research, some critics warn against the tendency of
this kind of studies for being too generic, which is
what happens when findings regarding patterns of
learning–teaching interactions are presented as if they
were independent of the topic learned in classrooms.
The second intra-discursively oriented line of
research on learning inquires about development of
discourses. For those who equate thinking with communicating, asking about the development of, say, one’s
mathematical discourse is tantamount to asking about
this person’s learning of mathematics. This time, the focus
is on the uniquely mathematical features of the discourse
and on these features’ gradual evolution. Comparable in
its aims to research conducted within the tradition of

D

1000

D

Discourse

conceptual change, this relatively new type of study on
learning is made distinct by its foundational assumption on the unity of thinking and communicating and
its use of methods of discourse analysis. It owes its
growing popularity, among others, to the pioneering
contributions by Jay Lemke (1993), whose work on
learning science was supported by analytic techniques
of systemic functional linguistics. One of the main tasks
yet to be undertaken is to develop subject-specific
method of discourse analysis, tailored according to
the distinct needs of the discourse under study.
Finally, the inter-discoursively oriented studies
inquire about interactions between discourses and their
impact on learning. This type of research is grounded in
the recognition of the fact that one’s access to a particular
discourse, say mathematics or science, may be supported
or hindered by other discourses. Of particular significance among these learning-shaping forms of communication are those that carry specific cultural norms and
values or distinct ideological messages. One of the earliest, but still influential exemplars of this kind of research
is the work of Basil Bernstein (1971). Studies belonging
to this tradition are often concerned with issues of
power, oppression, equity, social justice, and race,
whereas the majority of researchers whom this research
brings together do not hesitate to openly admit their
ideological commitment. The notion of identity is often
used here as the conceptual device with which to
describe the way cultural, political, and historical narratives impinge upon individual learning (see, e.g., Gee
2001). Methods of ▶ critical discourse analysis are particularly useful in this kind of study.
As different as these three lines of research on learning may be in terms of their focus and goals, their
methods have some important features in common. In
all three cases, the basic type of data is the carefully
transcribed communicational event. A number of
widely shared principles guide the processes of collection, documentation, and analysis of such data. Above
all, researchers need to keep in mind that different
people may be using the same linguistic means differently, and that in order to be able to interpret other
person’s communicational actions, the analysts have to
alternate between being insiders and outsiders to their
own discourse: they must sometimes look “through”
the word to what they usually mean by it, and they also
must be able to ignore the word’s familiar use, trying to
consider alternative interpretations. For the same

reason, events under study have to be recorded and
documented in their entirety, with transcriptions
being as accurate and complete records of participants’
verbal and nonverbal actions as possible. Finally, to be
able to generalize their findings in a cogent way,
researchers should try to support qualitative discourse
analysis with quantitative data regarding relative frequencies of different discursive phenomena.
Because of these and similar requirements, the
discourse-oriented research is much more demanding
and time-consuming than many other types of studies.
If a researcher is still ready to engage in this kind of
investigations, it is because of their unique payoffs.
True, the task of discourse analyst is not much different
from that of any person trying to take part in verbal or
nonverbal exchanges: in both cases, one has to make
sense of other people’s discursive actions. And yet, the
admittedly demanding methods of discourse analysis,
when at their best, allow the analyst to see what inevitably escapes one’s attention in real-time conversations. The resulting picture of learning is
characterized by high-resolution: one can now see as
different things or situations that, so far, seemed to be
identical; and is able to perceive logic in discursive
actions that in real-time exchange appeared as nonsensical. Moreover, discourse-oriented research brings
a promise of a unifying framework, free from dichotomous divides such as thinking versus communicating,
form versus content, cognition versus affect, or individual versus social. In this framework, all aspects of
learning would be seen as members of a single ontological category, to be studied with one integrated
system of tools. The task of constructing such framework is likely to preoccupy discourse-oriented investigators of learning in years to come.

Cross-References
▶ Collaborative Learning
▶ Computer Supported Collaborative Learning
▶ Discourse and the Production of Knowledge
▶ Discourse Processes and Learning
▶ Inquiry Learning

References
Bernstein, B. (1971). Class, codes, and control. New York: Schocken
Books.
Gee, J. P. (2001). Identity as an analytic lens for research in education.
Review of Research in Education, 25, 99–125.

Discourse and the Production of Knowledge
Harré, R., & Gillett, G. (1995). The discursive mind. Thousand Oaks:
Sage.
Lemke, J. L. (1993). Talking science: Language, learning, and values.
Norwood: Ablex.
Mehan, H. (1979). Learning lessons: Social organization in the classroom. Cambridge, MA: Harvard University Press.
Rorty, R. (1979). Philosophy and the mirror of nature. Princeton:
Princeton University Press.

Discourse Analysis

relationship between these two central notions of
the humanities and social sciences, despite the fact
that we acquire most knowledge by text and talk, and
that in order to produce and understand discourse language users need vast amounts of knowledge. This article
summarizes some of the current theoretical and empirical studies that have contributed to this insight, especially in contemporary Discourse Studies and
Cognitive Science.

Discourse Studies

Also known as DA or discourse studies, is a generic
name for different approaches to analyzing written,
spoken or signed communication. DA focuses on
speech units larger than the sentence and takes into
account the contexts in which discourse occurs.

Discourse and the Production
of Knowledge
TEUN A. VAN DIJK
Department of Translation and Language Sciences,
Pompeu Fabra University, Barcelona, Spain

Synonyms
Beliefs; Communication; Discourse;
processing; Knowledge; Talk; Text

D

Discourse

Definitions
Social knowledge is here defined as the shared, justified
beliefs held by the members of an (epistemic)
community.
Discourse is variously defined as a communicative
event, a form of interaction and as a situated unit of
language use.

Theoretical Background
Introduction
Both on discourse and on knowledge there is a vast
amount of research since classical rhetoric and epistemology. Yet, there is as yet not a single monograph that
explores the obvious insight of the fundamental

Since the 1960s, the cross-discipline of Discourse Studies has vastly extended our understanding of text and
talk in all disciplines of the humanities and social
sciences, and beyond the psycholinguistics as well as
the traditional, structural, and generative grammars of
isolated sentences. Discourse today is analyzed as
a complex, multimodal object, as a form of social
interaction and as a communicative event in its sociocultural context, managed by socially shared underlying cognitive strategies and representations – some of
which are to be dealt with in this article (Schiffrin et al.
2001; Van Dijk 2011).

The Theory of Knowledge
Classical as well as much of modern epistemology fundamentally defines (declarative) knowledge as justified
true beliefs, with many variations as to the nature and
conditions of justification (among a vast numbers of
books in epistemology, see, e.g., Bernecker and Dretske
2000). In this article, our approach to the theory of
knowledge will be more natural and pragmatic, namely
as a multidisciplinary account of the cognitive, social,
and cultural properties and functions of the shared
beliefs of an (epistemic) community, justified by the
variable (epistemic) standards or criteria of that community. This approach implies that knowledge is both
contextual and relative: What is assumed to be knowledge now by the members of an epistemic community
maybe seen as mere or false belief, or as superstition or
prejudice, by members of another community, or by
those of the same community later. As a practical test,
we assume that beliefs count as knowledge of
a community if they are presupposed and taken for
granted in the social practices, and hence in the public
discourse, of the community. We here find a first and
fundamental relationship between discourse and
knowledge.

1001

D

1002

D

Discourse and the Production of Knowledge

The psychological study of knowledge, since the
cognitive revolution of the 1960s and 1970s, analyzed
knowledge as organized networks of concepts and categories in semantic memory, and as part of Long-Term
Memory, for instance, in terms of schemas, scripts, and
prototypes (for review, see Wilkes 1997). It did so
largely in isolation from the obvious social psychological insight that most knowledge is not acquired and
used by isolated individuals, but shared by, or distributed over the minds of the members of a community.
Under the influence of the emerging neurosciences
in the 1990s, psychology today is developing new
insights into knowledge defined as an embodied, multimodal system “grounded” in various brain regions,
such as those processing vision, movement, and emotion, involved in the acquisition and uses of knowledge
in the experiences of everyday life (Barsalou 2008,
among many other papers).

Discourse Processing
It is within this broad, multidisciplinary framework
that we need to account for the cognitive production
and comprehension of discourse, and for the role of
knowledge both as a condition as well as a consequence
of these processes (for reviews and introductions on
discourse processing, see, e.g., Graesser et al. 1997,
2003; Kintsch 1998; McNamara and Magliano 2009;
Van Dijk and Kintsch 1983).

Discourse Production and Knowledge
Management
Given the multimodal and multilevel nature of discourse, the production of text or talk is a situated social
practice organized by semiotic (phonological, visual,
etc.), syntactic, semantic, pragmatic, and interactional
structures based on various kinds of mental representations and organized by cognitive strategies that make
sure that the discourse is understandable, well-formed,
meaningful, appropriate, and efficient in its communicative situation (despite the vast literature on discourse
processing, there are hardly specialized monographs
focusing on the production of discourse).
At all these levels of discourse production, first of
all, socially shared knowledge of the language, consisting
of the lexicon, the grammar, as well as the rules of
discourse, interaction and context, obviously plays
a central role. At the same time, language users need
to activate and apply their knowledge of the world, that

is, their general, socially shared knowledge about the
objects, people, actions, events or situations talked or
written about (for references, see below).
Given the shared nature of social knowledge of the
world, as Common Ground (Clark 1996), language
users need not express all information in discourse
they may assume can be inferred by the recipients
from the knowledge they have in common with the
speaker or writer. In other words, discourse is essentially
incomplete, because many of the propositions that
define its local and global meaning and coherence are
left implicit in the process of production.
Despite the vast amount of knowledge language
users of the same community have in common, there
are obviously personal and social differences as to the
knowledgeability or the expertise of individual language users. Hence, speakers and authors need to contextually adapt this knowledge management during
discourse production to their assumptions about the
knowledge of the recipients, or the lack of knowledge of
new members of the epistemic community (children,
students, foreigners, etc.), as is also the case in the
popularization of science. For didactic, persuasive, or
emotional reasons, speakers may of course repeat some
information they know recipients might or should
already have. And conversely, recipients may be manipulated or otherwise abused if the speaker presupposes
knowledge they do not have – but still is taken for
granted indirectly, even when in fact the beliefs are
false.
Further dependent on many contextually variable
strategies and constraints, the general pragmaticepistemic rule of discourse production is that speakers
or writers assert propositions they assume recipients do
not yet know and cannot infer themselves from their
own knowledge. This is at the same time the basic
condition of (new) knowledge production as well as
of knowledge distribution and reproduction in the
community.

Context Models
Language users are only able to epistemically adapt
their text or talk to the recipients if they know what
the recipients know. Such assumptions are part of their
subjective representation of recipients and other relevant aspects of the communicative situation, called
their context model, stored in episodic memory, part
of Long-Term Memory (Van Dijk 2008, 2009).

Discourse and the Production of Knowledge

A dynamically changing context model controls the
many variable aspects of discourse that make sure the
discourse (fragment) is communicatively appropriate,
such as its genre, style, register, and topics. Such
a context model consists of a relatively simple schema
with categories such as Setting (Place, Time), current
Social Action, as well as the Participants (and their
current social identities, roles, and relations, and their
current cognitive properties, such as their goals and
knowledge, as well as ideologies if they speak as group
members).
Current multimodal knowledge theories suggest
that these context models, defined as models of communicative experience just like any other experience,
may well have a multimodal nature, featuring auditory
aspects of speech (such as a special tone, stress, or
intonation) or the environment (e.g., noise), visual
information about participants and the setting, body
movements of interaction (gestures, position), as well
as opinions and emotions about the participants, the
topics of discourse, or the whole speech event
(Barsalou 2008).
Central in the context model is a knowledge device
that dynamically and ongoingly hypothesizes what
recipients already know or may infer from their knowledge, so that the speaker can strategically adapt the
discourse to this assumed knowledge of the recipients,
by being more or less explicit or implicit, and manage
what information must be asserted and what information may be presupposed, both locally within or
between sentences, as well as globally as in the discourse
as a whole. One powerful strategy is that if recipients are
members of the same knowledge community, recipients
are assumed to have the same general knowledge as the
speaker or author, except new knowledge the speaker or
author has recently acquired by reliable observation,
sources (speakers, media), or inference.

Discourse Comprehension
Discourse comprehension has many properties in common with discourse production, and is a process that is
based on (more or less) the same knowledge of the
language and the world as they are used and applied
by speakers and writers (Kintsch 1998; Britton and
Graesser 1996). The obvious difference is that speakers
and writers in principle know what they mean and
want to convey, and need to find an appropriate discursive expression to these meanings, whereas

D

recipients start with this discursive expression and
need to figure out what the speaker or writer means.
Recipients have their own context model of the
communicative situation, with their own information
and opinions about the setting, the participants (and
their identities, roles, relations, goals, knowledge, etc.)
and the ongoing social action. Discrepancies with the
context model of the speaker or writer, for instance,
about the goal of the communicative interaction, may
thus lead to communicative conflicts.
Especially relevant for the topic of this article is the
role of knowledge in the construction of the meaning of
the discourse (see Kintsch 1998; Van Dijk and Kintsch
1983). Since speakers or writers assume recipients are
able to infer much information from their (shared)
social knowledge, this is precisely what recipients
(have to) do: Together with the information derived
from what is explicitly expressed in discourse, they
ongoingly must generate at least those inferences
from their knowledge that are needed to produce
a meaningful and coherent semantic interpretation of
the discourse (Graesser and Bower 1990). Typically,
they may thus generate plausible causes or consequences of events or reasons of action, or fill in many
details of socioculturally well-known episodes, such as
going to work or to school, shopping, eating in restaurants, birthday parties, or demonstrations, among
many others. Obviously, the nature and amount of
these inferences crucially depend on the abilities (literacy, etc.), knowledge, goals, or tasks of the recipients
(for details, see, e.g., Graesser and Bower 1990).

Situation Models
This knowledge-based process of discourse comprehension appears to go far beyond the mere interpretation of words, clauses, or sentences and even beyond
the construction of locally and globally coherent discourse meanings. Indeed, the goal of discourse comprehension is not merely to understand the discourse
itself, but rather what the discourse is about: what it
tells us about some event or situation or the world. It is
therefore assumed that besides construing a semantic
representation of the discourse (its intension), language
users also construe a subjective, multimodal mental
model of these events, situations, or episodes, referred to
or spoken about (its extension). In other words, to
understand discourse means to be able to construe
a mental model for it. This model may feature the

1003

D

1004

D

Discourse and the Production of Knowledge

visual, auditory, sensorimotor, emotional, and other
modal aspects that are associated with the way recipients imagine or simulate the event talked or written
about (instead of mental models, Barsalou (2008)
speaks of simulations to refer to situated comprehension and experiences).
As is the case for (pragmatic) context models and
other models of personal experience, these (semantic)
models of events or situations are also stored in Episodic Memory (for details on mental models, see Johnson-Laird 1983; Van Dijk and Kintsch 1983: Van
Oostendorp and Goldman 1999).
As suggested, general, sociocultural knowledge
plays a central role in the construction of this mental
model, together with the (new) information of the
discourse, and possibly with information derived
from old mental models (previous experiences, previous discourses), for instance, by supplying missing
inferences about conditions, consequences, participants, details, and other plausible elements of the
situation.
Note that the way knowledge is (partly) activated
and applied in the construction or updating of such
models of the event or situation referred to is controlled
by information in the pragmatic context model. In
other words, different recipients may interpret the
same discourse in a different way by constructing different (semantic) situation models. And conversely, for
the same contextual reasons, different readers may also
acquire different (new) knowledge from the same discourse, depending on their previous knowledge, interest, motivation, and current goals.

Knowledge Production
Crucial at this point is not only that shared knowledge is
strategically (partly) activated to construe semantic representations and mental models, but also may be
transformed (formed, changed) by discourse. Indeed,
information that is not implied or presupposed by text
or talk may be used to build and socially distribute
mental models about unknown events, as is the case in
everyday personal storytelling as well as in news reports.
When repeated, such discourses and their mental
models may be generalized and abstracted from so as to
form more general knowledge about this type of event.
For instance, news about specific terrorist attacks may
be used to build knowledge and attitudes about terrorism. This is a special (discourse) way of learning from

personal experience, and a condition for the social
reproduction of knowledge as well as other forms of
social cognition (attitudes, ideologies, norms, values)
in society.
Obviously, besides model-based (i.e., experiencebased) acquisition of knowledge, new knowledge may
also be produced more directly, as is the case in many
forms of pedagogical or expository discourse (Britton
and Black 1985), for instance, by generic descriptions
of events, objects, or phenomena; by definitions of
terms; the use of metaphors; schemas; etc. As is the
case for all discourse, such discourse presupposes the
general, shared knowledge of the community, but strategically expands this knowledge by various multimodal strategies of knowledge transformation. These
may include information about (a) categorical relationships (such as higher-level categories or lowerlevel subcategories), (b) visual or other perceptual
appearances, (c) parts or components, (d) relationships with other objects or phenomena, (e) functions
or uses, and so on.

Important Scientific Research
Most of the theoretical issues of discourse processing
and the role of knowledge discussed above have been
shown to be empirically warranted by (mostly experimental) research. Thus, many studies have shown that
discourse comprehension crucially depends on the
activation and application of what is usually called
“prior knowledge” – although such knowledge is not
usually precisely defined (McNamara and Kintsch
1996; see also below).
Thus, perhaps trivially, people who know more about
a domain or topic, usually better understand a discourse
about such a topic or domain – if only because they are
able to derive more inferences and hence are able to
construe more detailed mental models of specific events
or new schemas of new, generic knowledge.
But, as is generally the case, both outside as well as
within in the laboratory, actual knowledge acquisition
depends on the structures and strategies of text and context. For instance, because of their larger knowledge,
experts may pay less attention to the specific details of
text or talk and hence may hardly do better than nonexperts in specific tasks, such as recall or recognition.
Similarly, if texts are very explicit they may be less
interesting for experts, and hence they may pay less
attention and again recall fewer details than

Discourse and the Production of Knowledge

nonexperts. And in all cases, it depends on the tasks
and hence the goals of the participants: Someone who
must correct the style or translate a news report may
well learn less about some news event than a reader or
a political activist who is specifically motivated and
interested in news about a specific topic or domain.
Among the vast number of studies on the role of
knowledge in discourse comprehension and hence on
learning from text, here is a summary of some of the
findings in addition to those that have been mentioned
above (for detail, and further references for each result,
see especially the chapters in Graesser et al. 2003).

Context Variables
● People in general learn more from text when they

●

●
●
●

have more prior knowledge about the domain or
topic of the text (among many studies, see, also
Kendeou and Van den Broek 2007).
People in general have a memory bias for the information with which they agree. However, people
with more knowledge about an issue are able to
better reproduce two sides of a controversial
argument.
Experts versus nonexperts (high- or low-knowledge
subjects) learn differently from texts.
People learn more from text when they do so interactively, e.g., by discussion about the text.
More generally, people learn more when they
explicitly (must) think about the way they learn
from the text (metacognition).

Text Variables
● More cohesive, more coherent, more explicit, and

better organized text (e.g., with summaries,
headers, conclusions) generally favor comprehension and hence knowledge acquisition.
● Inaccurate prior knowledge needs to be explicitly
rejected – it is less efficient to simply present correct
knowledge.
● Images may help understanding and learning from
text.

Combined Text and Context Variables
● In general, people learn more from cohesive, more

coherent, and well-organized text, especially if they
are less-skilled readers, but the interaction between
text structure, prior knowledge, and reading ability
is more complex than that.

D

Unfortunately, most experimental work in the
laboratory is focused more on “learning from text”
in the narrow sense of what (new) information can
be recalled, recognized, reproduced, or applied in specific ad hoc laboratory tasks (see also Kintsch 1991,
1998). Socially shared knowledge, however, should
be defined in broader terms, and at least involve relatively long-term or even permanent transformation of
our socioculturally shared knowledge as members
of epistemic communities. Outside educational
situations (classrooms, exams, etc.), few controlled
experiments offer insight into these long-term constructions and transformations of our knowledge.
Most likely, such new, socioculturally shared, knowledge is acquired and integrated within the knowledge
system if it is repeatedly situationally relevant,
namely if it is often presupposed for the understanding
of public discourse (as is the case for our generally
knowledge about computers, the Internet, and DNA,
for instance) and if it is taken for granted in other
social practices.
Most experimental studies on the role of knowledge
in discourse production and comprehension, or on the
acquisition of (new) knowledge from discourse, barely
reflect on the nature, the structure, and the organization of knowledge in memory, and how such knowledge is changed. In order to examine how exactly
people acquire knowledge from discourse, we need to
know much more about how the structures of discourse are related to the structures of knowledge, as
well as about the many context variables that affect this
relationship in actual learning and the use and reproduction of knowledge in society.

Cross-References
▶ Cognitive Models of Learning
▶ Discourse
▶ Discourse Processes and Learning
▶ Knowledge Acquisition
▶ Knowledge and Learning in Natural Language
▶ Knowledge Integration
▶ Knowledge Representation
▶ Learning from Text
▶ Memory Dynamics
▶ Mental Models
▶ Mental Models in Discourse processing
▶ Naturalistic Epistemology

1005

D

1006

D

Discourse in Asynchronous Learning Networks

References
Barsalou, L. W. (2008). Grounded cognition. Annual Review of
Psychology, 59, 617–645.
Bernecker, S., & Dretske, F. I. (Eds.). (2000). Knowledge:
Readings in contemporary epistemology. Oxford: Oxford
University Press.
Britton, B. K., & Black, J. B. (Eds.). (1985). Understanding expository
text. A theoretical and practical handbook for analyzing explanatory text. Hillsdale: Lawrence Erlbaum Associates.
Britton, B. K., & Graesser, A. C. (Eds.). (1996). Models of understanding text. Mahwah: Lawrence Erlbaum Associates.
Clark, H. H. (1996). Using language. Cambridge: Cambridge
University Press.
Graesser, A. C., & Bower, G. H. (Eds.). (1990). Inferences and text
comprehension. The psychology of learning and motivation (Vol.
25). New York: Academic.
Graesser, A. C., Millis, K. K., & Zwaan, R. A. (1997). Discourse
comprehension. Annual Review of Psychology, 48, 163–189.
Graesser, A. C., Gernsbacher, M. A., & Goldman, S. R. (Eds.).
(2003). Handbook of discourse processes. Mahwah: Lawerence
Erlbaum.
Johnson-Laird, P. N. (1983). Mental models. Towards a cognitive science of language, inference, and consciousness. Cambridge, MA:
Harvard University Press.
Kendeou, P., & van den Broek, P. (2007). The effects of prior
knowledge and text structure on comprehension processes
during reading of scientific texts. Memory & Cognition, 35(7),
1567–1577.
Kintsch, W. (1991). The role of knowledge in discourse comprehension: A construction-integration model. In G. Denhiere & J. P.
Rossi (Eds.), Text and text processing (pp. 107–153). Oxford:
North-Holland.
Kintsch, W. (1998). Comprehension. A paradigm for cognition. Cambridge: Cambridge University Press.
McNamara, D. S., & Kintsch, W. (1996). Learning from texts: Effects
of prior knowledge and text coherence. Discourse Processes, 22(3),
247–288.
McNamara, D. S., & Magliano, J. (2009). Toward a comprehensive
model of comprehension. Psychology of Learning and Motivation,
51, 297–384.
Schiffrin, D., Tannen, D., & Hamilton, H. E. (Eds.).
(2001). The handbook of discourse analysis. Malden: Blackwell
Publishers.
Van Dijk, T. A. (2008). Discourse and context. A socio-cognitive
approach. Cambridge: Cambridge University Press.
Van Dijk, T. A. (2009). Society and discourse: How context controls text
and talk. Cambridge: Cambridge University Press.
Van Dijk, T. A. (Ed.). (2011). Discourse studies. A multidisciplinary
introduction. New, one-volume edition. London: Sage.
Van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse comprehension. New York/Toronto: Academic.
Van Oostendorp, H., & Goldman, S. R. (Eds.). (1999). The construction of mental representations during reading. Mahwah: Lawrence
Erlbaum.
Wilkes, A. L. (1997). Knowledge in minds. Individual and collective
processes in cognition. Hove: Psychology Press.

Discourse in Asynchronous
Learning Networks
ALLAN JEONG
Educational Psychology & Learning Systems, Florida
State University, Tallahassee, FL, USA

Synonyms
Computer-mediated communication

Definition
Discourse – communication of thought via written text
and speech. In the field of linguistics, discourse is any
unit of connected speech or writing longer than
a sentence.
Asynchronous learning – A network of people
engaged in peer-to-peer interaction and information
sharing via online resources (e.g., email, electronic mail
lists, threaded discussion boards, blogs, wikis, twitter,
document sharing systems) that overcome the constraints of time and place.
Network – a collection of electronic tools (e.g.,
email, discussion boards, text messaging, video conferencing), and devices (e.g., computers, cell phones,
mobile devices) interconnected by multiple communication channels (e.g., text, voice, visual) used to facilitate information sharing and communications.
Discourse in asynchronous learning networks –
The communicative behaviors exhibited in peer-topeer interactions observed in electronic media.

Theoretical Background
The earliest form of asynchronous learning was
established with the introduction of distance education
and correspondence courses in the early 1800s and
increased access to postal mail. Although postal mail
enabled learners to overcome the constraints of time
and place, asynchronous learning was primarily
a solitary activity where communication with instructors and other learners were limited by the cost and
time delays associated with postal mail. With limited
opportunities for learners to engage in discourse with
instructors and peers, the instructional content and
instructional activities were by necessity carefully
developed and determined by the course instructor

Discourse in Asynchronous Learning Networks

and/or sponsoring program to ensure that learners are
able to work independently to successfully achieve the
learning objectives. As a result, asynchronous learning
was primarily if not exclusively instructor-led.
With advances in technologies over the last century,
learners are gaining increased access to tools and
resources that support multi-channel discourse (text,
voice, visuals). The steady improvements in tools and
tool integration enables better coordination, communication, and information sharing between peers and at
the same time enables learners to pool their collective
knowledge and experiences in ways that is decreasing
learners’ reliance on the input and guidance of instructors. As a result, asynchronous learning not only incorporates methods that support self-directed learning but
also incorporates methods that support collaborative
learning. Collaborative learning is more a studentcentered than an instructor-led approach to learning
given that the ▶ discourse between learners reflect the
learner’s interests, motivations, prior knowledge and
experiences, and ways of knowing – forming much of
the foundation on which knowledge and meaning is
constructed. Some of the common goals of discourse
(when conducted within an asynchronous learning
network or ALN) are to promote and increase the
level of critical thinking, meaningful problem solving,
and knowledge construction.
With the development of the World-wide web, ALN
discourse has become a key component in most online
courses. As a result, online discourse has been the focus
of much research among educational and instructional
psychologists. The ultimate goal of the research is to
examine and better understand how different tools can
be used and further refined to facilitate discourse in
ways that increases learning and performance. However, researchers have faced difficult challenges in the
attempt to determine which tools help produce better
discourse because there are too many independent
variables that must be controlled – variables such as
group size, composition of group, and nature of tasks.
Furthermore, these variables interact in such ways that
it is difficult if not impossible to establish cause and
effect between choice of tools, quality of discourse, and
learning outcomes.
To address these challenges, researchers are examining how specific attributes of a tool changes and
mediates the interactions exhibited in online discourse
(Dillenbourg et al. 1996). In order to examine how tool

D

attributes mediate learner interactions, new conceptual
frameworks, methods, and tools have been developed
to analyze and/or model learner interactions and learning processes exhibited in the interactions. At the heart
of this research is the issue of what (e.g., cognitive,
meta-cognitive, social behaviors; quantitative vs qualitative; individual vs group; message vs sentence units;
data from self-reports vs discussion transcripts) to
examine and code from the discourse, and how to
analyze the data (e.g., frequency counts, response probabilities, Markov chains). A myriad of models and
approaches have been developed and used to elucidate,
make more explicit, and operationally measure the
form, function, and/or the dynamic and interactive
nature of ALN discourse.

Important Scientific Research and
Open Questions
Despite the complexity and scope of the research in this
field, three critical overarching questions are identified
and presented below – questions that reflect current
challenges and suggest directions for future research.
Included are brief descriptions of studies that illustrate
how some of the latest tools and methods have been
used to address these questions and that can be used to
make further advancements in the field.
What discourse models or typologies are most useful
for identifying the interactions that produce higher gains
in learning? Although an abundance of studies have
developed and/or implemented various models to
code and analyze online discourse (brief descriptions
of some of the existing models presented in Marra et al.
[2004]), the focus of the analysis needs to be centered
foremost around the cognitive operations exhibited
within each dialog move – cognitive operations that
learners must perform to complete the learning task
and achieve the desired learning outcome. Other dialog
moves associated with the social and meta-cognitive
behaviors (or any other dimensions of discourse)
should be analyzed in terms of how they influence the
sequence of cognitive actions exhibited by learners as
they engage in discourse. For example, Garrison et al.’s
(2010) structural equation analysis produced a model
that revealed the extent to which learners’ social interactions and interactions with instructors impacted the
cognitive interactions performed by learners. However,
this study like most studies that examine online discourse did not determine how different types of

1007

D

1008

D

Discourse in Asynchronous Learning Networks

.01

.66

.73

ARG
143->153

.52

ARGc
32->21

BUT
239->77

.90

BUTc
174->44

.69

.82

.59

.75

.16
.09

EXPL
50->17

+.31

.18
+.28

.11

+.23

.18

.10

EVID
81->29

.14
Without conversational language

EXPLc
37->13

.08

EVIDc
30->4

.25

.23
With conversational language

Discourse in Asynchronous Learning Networks. Fig. 1 Transitional state diagrams illustrating the response patterns
produced from messages with versus without conversational language. ARG = argument, BUT = challenge, EVID =
supporting evidence, EXPL = explanation, “c” denotes a message presented in a conversational style, “+” denotes
transitional probabilities that were significantly higher than expected (based on z-scores at p<0.01)

cognitive interactions affect the learning outcome. In
order to do this, particular methods must be used to
precisely identify and convey the similarities and differences in interaction patterns produced by high versus low performing learners.
Which methods and tools can identify, convey, and
model interactions patterns and identify which interactions produce higher gains in learning? Given the complexity and dynamic nature of discourse, dialog move
sequences do not always unfold in orderly and predictable ways. Soller (2004) believed that this is one reason
why the simple frequencies of each dialog move
performed by learners did not distinguish learners
that scored high versus low on a posttest measuring
knowledge acquisition. As a result, Soller incorporated
a process-oriented approach that examined how interactions unfold over time by producing transitional
state diagrams (often referred to as Markov chains) to

convey how likely (or the probability) one dialog move
was followed by another dialog move (e.g., inform,
acknowledge, request information, discuss with doubt,
agree). This interaction data combined with posttest
scores were analyzed using multidimensional scaling
to reveal three to four interactions (spanning from
one to three conversational turns) that distinguished
the groups that collectively scored high versus low on
the posttest. Using a similar approach to examine how
social interactions influence cognitive interactions,
Jeong (2006) used transitional state diagrams (Fig. 1)
to determine how the use of conversational language
(e.g., making references to participants by name, saying
thank you, and use of greetings and emoticons)
affected the probabilities of certain responses elicited
by arguments, challenges, explanations, and presentation of supporting evidence. The findings reveal that
the (argument ! challenge ! explanation) interaction

Discovery Learning

was more likely to emerge from students’ interactions
when students used conversational language while
presenting arguments, challenges, and explanations.
How do communication tools change and mediate the
interactions in ways that produce higher gains in learning? Olson et al. (1992) conducted one study that examined both the relationship between tool and group
interactions, and the relationship between group interactions and learning outcome. Their study compared the
effects of using a shared document editor, ShrEdit,
against the use of whiteboard with paper and pencil on
group interaction patterns and performance on a group
paper. Their findings revealed that students using
ShrEdit produced significantly higher quality papers
even though the interaction patterns produced in both
groups (visually conveyed in transitional state diagrams)
appeared to be very similar. However, Olson did not
perform any statistical tests on the transitional probabilities (e.g., z-score tests) to determine which of the interaction patterns occurred at rates that were significantly
higher or lower than expected. Nevertheless the study
did find significant group differences in the amount of
time spent on performing specific actions (discussing
issues, actions, and alternatives) and some significant
differences in the frequency of actions performed by
the groups between conditions.
Future efforts to integrate the methods and tools
used in the studies described above will provide the
basis to conducting more complete investigations into
the relationship between technology, discourse, and
learning. At the same time, future investigations will
need to determine how differences in socio and cultural
contexts affect how well discourse models can be used to
accurately explain and predict learning outcomes. These
efforts all together will produce the empirical research
needed to arrive at a better understanding of communication technologies and how to refine the technologies to
promote the type of discourse that can optimize learning
processes and maximize learning outcomes.

D

▶ Discourse and the Production of Knowledge
▶ Distance Learning
▶ Learning with Collaborative Mobile Technologies
▶ Online Collaborative Learning
▶ Rapid Collaborative Knowledge Building

D

References
Dillenbourg, P., Baker, M., Blaye, A., & O’Malley, C. (1996). The
evolution of research on collaborative learning. In E. S. P. Reiman
(Ed.), Learning in humans and machine: Towards an interdisciplinary learning science (pp. 189–211). Oxford: Elsevier.
Garrison, D. R., Cleveland-Innes, M., & Fun, T. S. (2010). Exploring
causal relationships among teaching, cognitive and social presence: Student perceptions of the community of inquiry framework. Internet and Higher Education, 13, 31–36.
Jeong, A. (2006). The effects of conversational styles of communication on group interaction patterns and argumentation in online
discussions. Instructional Science, 34(5), 367–397.
Marra, R. M., Moore, J., & Klimczak, A. (2004). Content analysis of
online discussion forums: A comparative analysis of protocols.
Educational Technology Research and Development, 52(2), 23–40.
Olson, G., Olson, J., Carter, M., & Storrosten, M. (1992). Small group
design meetings: An analysis of collaboration. Human-Computer
Interaction, 7(4), 347–374.
Soller, A. (2004). Computational modeling and analysis of knowledge
sharing in collaborative distance learning. The Journal of Personalization Research, 14(4), 351–381.

Discourse Processing
▶ Discourse and the Production of Knowledge
▶ Language/Discourse
Comprehension
Understanding

and

Discovery Learning
HEINZ NEBER
University of Munich (LMU), Munich, Germany

Cross-References
▶ Collaboration Scripts
▶ Collaborative Learning
▶ Collaborative Learning Strategies
▶ Collaborative Learning Supported by Digital Media
▶ Computer-Supported Collaborative Learning
▶ Discourse

1009

Synonyms
Example-based learning; Guided discovery learning;
Inductive teaching; Inquiry learning; Learning by
design; Learning by experimentation; Socratic
questioning

1010

D

Discovery Learning

Definition
Discovery Learning denotes a general instructional
approach that represents the first broad development
of constructivist learning for school-based learning
environments. Jerome Bruner (1961) derived discovery
learning from contemporary studies in cognitive psychology and stimulated the development of more specific instructional methods. The most important
defining characteristic of discovery learning is that
learners have to generate units and structures of
abstract knowledge like concepts and rules by their
own inductive reasoning about nonabstracted learning
materials (Holland et al. 1986). Only such materials are
provided by the learning environment. The learning
materials may consist in examples of general concepts,
cases for general approaches and procedures (e.g.,
teaching method, management style), ill-defined questions and kinds of situated problems (e.g., how to
motivate students in a passive classroom), or phenomena that have to be causally explained by the learners
(e.g., why a liquid gets hard). Another characteristic is
the amount of guidance of the learner’s required inductive reasoning processes. In discovery learning situations, the level of guidance offered may vary adaptively,
depending on the difficulty of the learning material, the
complexity of the intended conceptual and procedural
knowledge, and the cognitive and motivational prerequisites of the learners. For this reason, the level of
guidance or structuredness of the learning environment is not fixed, but represents a variable, nondefining characteristic of discovery learning. This is
a necessary statement because in some recent discussions, discovery learning has been confused with
unguided instruction which has to be considered as
a big flaw (Hmelo-Silver et al. 2007). In fact, in all
actually implemented versions of discovery learning,
the reasoning processes of the learners are considerably
guided or scaffolded.
Learning by examples as the probably earliest
method of discovery learning exemplifies a high level
of guidance. This method of discovery learning represents a rather direct instructional application of
Bruner’s procedure for investigating concept attainment by children. Concept attainment requires the
induction of defining and non-defining characteristics
or attributes from examples or instances of the concept
or category (e.g., concepts like pet, metacognition,
cooperative learning). In the corresponding

instructional method, carefully selected examples and
non-examples are presented in planned and prescribed
sequences (e.g., always beginning with a positive example). The learners analyze the features of each presented
example, compare examples, decide on the definitional
status of each feature, and get a feedback on the correctness of the decision. This cycle of searching, deciding, and testing is repeated until the learners are able to
define the concept in terms of its induced abstract
characteristics. Guidance is provided by presenting
series of well-designed instances, by prescribing and
facilitating the steps of searches and comparisons to
the learners (e.g., by providing prompts and rubrics),
as well as by providing feedback on the correctness of
the assumptions and hypotheses of the learners. Learning by examples as a basic form of discovery learning
belongs to the repertoire of well-established models of
teaching that are derived from information-processing
studies in cognitive psychology (Joyce and Weil 2008).
Meanwhile, the spectrum of methods of discovery
learning has been considerably expanded. The most
widely used and investigated method is learning by
experimentation (Neber 2010) or synonymously called
scientific inquiry (Joyce and Weil 2008; de Jong et al.
2005). In experimenting, learners generate causal
knowledge for explaining instructionally provided phenomena. In contrast to learning by examples, they do
not get an automatic feedback on the correctness of
their assumptions or hypotheses about causal variables.
This method is challenging for learners because they
additionally have to design situations (experiments)
for testing their provisional explanations (hypotheses)
on their own. In terms of Herbert Simon’s Dual Space
Search Discovery (DSSD) model, rules (causal knowledge) and examples or instances to test the rules have to
be coordinatively generated by the learners in laboratory-based hands-on experimentation or in technology-based virtual learning environments (e.g.,
microworlds)(de Jong et al. 2005).
Learning by Designing or Design-based Learning
represents an even more challenging discovery learning
method. In contrast to the other two methods, learners
do not get issues (e.g., examples, phenomena) that have
to be retrospectively analyzed, defined, or causally
explained. With this method, learners have to create
a product (e.g., a machine, program, a text, or a model)
that attains prescribed functions or meets given criteria
and constraints. For designing products that attain the

Discovery Learning

provided prespecified functions or criteria, domainspecific knowledge has to be searched and generated
by the learners.
The shortly described methods of discovery learning may be realized in isolation. On the other hand,
these methods and, in particular, learning by experimentation and by designing, constitute components in
more complex instructional approaches like ProjectBased Learning or Problem-Based Learning (PBL). As
in stand-alone solutions as well as in integrated instructional approaches, methods of discovery learning can
be implemented in learning environments that are
augmented by educational technology and by collaboratively distributed knowledge-generation processes.
Both, computer-based technology and collaboration
scripts may be used as tools and supports for the
inductive reasoning processes that are required in all
methods of discovery learning.

Theoretical Background
In general, discovery learning expands the range of
cognitive processes for learners and contributes to promote cognitively driven activities of the learners. Thus,
the implementation of discovery learning methods may
contribute to attain higher levels of thinking that represents an important general goal of education. The
cognitive processes that are required in all methods of
discovery learning may be conceived and investigated
on different levels of decomposition. On a macrolevel,
the sequence of cognitive or inductive reasoning processes is analyzed in terms of inquiry cycles, and analogously as investigation webs, learning cycles, or as
cyclical regulatory phases (Neber 2010). These cycles
consist in sequences of macroprocesses like accessing
prior knowledge, questioning, hypothesizing,
searching for information, deciding, explaining, and
reviewing. This expanded range of processes that is
required to generate knowledge by the learners is difficult, at least for novice learners. Many studies revealed
that in discovery-learning environments, learners often
remain non-intentional, do not frame questions and
hypotheses, have deficits in coordinating evidence and
explanations, e.g., do not provide arguments for their
explanations in terms of available data or by referring
to the generated domain-specific knowledge. In these
cases, discovery learning remains ineffective.
As a consequence, some of these macroprocesses of
inquiry cycles have further been decomposed and

D

investigated on microlevels that inform about the procedural complexity and provide the basis for developing instruments and tools for procedural support and
scaffolding. The actual research focuses in particular on
questioning, hypothesizing, argumenting, and
explaining (Neber and Anton 2008).
Principles for guiding and supporting discovery
processes on different levels have been theoretically
derived from conceptions of learning, and further
developed by empirical studies. A first principle is
based on positive findings by presenting discovery
tasks in stages of increasing complexity of the concepts,
rules, or theories that the learners have to generate
(e.g., well-established approaches are progressive
inquiry, or the concept of model progression in technology-based virtual learning environments; de Jong
et al. 2005). A second principle consists in the stepwise,
adaptive, and systematic fading or reduction of structure provided for the macrophases of inquiry cycles
(e.g., as proposed by the National Science Foundation
for discovery learning by experimentation in all science
subjects). An example for a fading procedure for framing questions by the learners is first prescribing
a question to the learners, then offering alternative
questions for selecting, followed by only providing
a script or prompt for formulating own questions by
the learners (Neber and Anton 2008).
For realizing these principles in guiding and structuring cycles of discovery processes, scaffolding techniques and support tools have been developed. The
broad range of such techniques can be distinguished
into three categories (Neber 2010): Techniques for
supporting knowledge-generation processes (e.g.,
prompts like question stems, prespecified templates
for hypotheses, planning grids for exploring the spaces
of knowledge and examples, or self-explanation
prompts and modeling tools), techniques for
supporting regulative processes (e.g., metacognitive
and reflection prompts), and techniques for
supporting collaborative activities (e.g., constructive
controversy and other cooperative scripts for distributing roles for knowledge-generation and regulative processes). Computer-based learning environments offer
special advantages for proximally adapting techniques
and scaffolds to online information about frequencies
and qualities of learners’ activities (e.g., SimQuest as an
example for designing discovery-learning environments with integrated guidance and support facilities).

1011

D

1012

D

Discovery Learning

Important Scientific Research and
Open Questions
Bruner (1961) argued that discovery learning is more
effective than what is called didactic teaching or nonconstructivist receptive learning. Positive effects should
be attained for memorization of knowledge, for solving
transfer problems, for general learning or self-regulation
strategies, and for intrinsic motivation. Bruner derived
these assumed effects from laboratory-based studies in
cognitive psychology. However, the range of variables
and their interactions are much less constrained in learning environments like classrooms. This may explain the
many discrepant findings about such effects of discovery
learning in school-based investigations. Compared to
other instructional approaches (e.g., direct instruction),
the attainment of positive cognitive and motivational
effects in discovery learning may be dependent on further conditions (of the learners and the instructional
environment) and seems to require additional resources
(e.g., more time-consuming, more teacher preparation,
and more other costs). The ongoing discussion about
guided-versus-pure discovery learning is a consequence
of the discrepant findings and indicates the need for
further studies on what and how to guide and scaffold
in all forms of discovery learning, in particular if it is
conducted under less structured conditions.
A second open question is if the assumed positive
effects of discovery learning have really been investigated by adequate research designs and measurement
instruments used to assess the effects. What seems to be
neglected are long-term effects, effects on general competencies, including the self-system development of the
learners, and effects on the attainment of non-inert
knowledge, including its structure.
A third category of open questions concerns the
context or environment of discovery learning. Only few
studies have focused on collaborative discovery learning,
and on how to adequately integrate discovery learning
with and into other instructional approaches. Even that
a considerable spectrum of scaffolding and support tools
have already been developed and tested, further investigations seem to be reasonable. In particular, how such
tools relate and interact with inquiry cycles, and determine cognitive processes on microlevels of decomposition. Finally, further studies on teacher development and
teacher education for discovery learning may help in
implementing this instructional approach and in
attaining the expected effects.

Cross-References
▶ Concept Formation: Characteristics and Functions
▶ Constructivist Learning
▶ Creative Inquiry
▶ Generative Learning
▶ Humanistic Approaches to Learning
▶ Inductive Reasoning
▶ Learning Cycles
▶ Learning from Questions
▶ Problem-Based Learning

References
Bruner, J. S. (1961). The act of discovery. Harvard Educational Review,
31, 21–32.
De Jong, T., Beishuizen, J., Hulshof, C., Prins, F., van Rijn, H.,
van Somereren, M., Veenman, M., & Wilhelm, P. (2005). Determinants of discovery learning in a complex simulation environment. In P. Gärdenfors & P. Johansson (Eds.), Cognition,
education, and communication technology (pp. 257–284).
Mahwah: Lawrence Erlbaum.
Hmelo-Silver, C. E., Duncan, R. G., & Chinn, C. A. (2007). Scaffolding and achievement in problem-based and inquiry learning:
A response to Kirschner, Sweller, and Clark (2006). Educational
Psychologist, 42, 99–107.
Holland, J. H., Holyoak, K. J., Nisbett, R. E., & Thagard, P. R. (1986).
Induction. Processes of inference, learning, and discovery. Cambridge: MIT Press.
Joyce, B., & Weil, M. (2008). Models of teaching (8th revised ed.).
Boston: Ally & Bacon.
Neber, H. (2010). Entdeckendes Lernen (Discovery learning). In
D. H. Rost (Ed.), Handwörterbuch Pädagogische Psychologie (Dictionary of educational psychology) (pp. 124–132). Weinheim:
Beltz.
Neber, H., & Anton, M. A. (2008). Promoting pre-experimental
activities in high-school chemistry: Focusing on the role of
students’ epistemic questions. International Journal of Science
Education, 30, 1801–1821.
Internet Sources
General description and illustration of discovery learning; live from
classroom. Discovery Learning Center. Retrieved from, http://
www.youtube.com/watch?v=bcuyHnVRJcE (for use, observe
YouTube regulations. Retrieved from, http://www.youtube.com/
t/terms).
Example of discovery learning by experimentation (scientific discovery): SimQuest as an authoring system for simulation-based
discovery-learning environments. Retrieved from, http://www.
simquest.nl/simquest/index.htm (developed by T. De Jong
et al., University of Twente, NL).
Example of discovery learning by examples: Concept attainment
model of teaching [Powerpoint presentation]. Retrieved from,
http://imet.csus.edu/imet3/drbonnie/portfolio/conceptattain/
conattainppt.pdf (designed by B. Drumright according to Joyce,
& Weil, 2008)

Discrimination Learning Model

Discovery Learning Model
▶ Guided Discovery Learning

Discrimination
▶ Big Five Personality and Prejudice

Discrimination Learning Model
JONAS ROSE1, ROBERT SCHMIDT2
1
The Picower Institute for Learning and Memory,
Massachusetts Institute of Technology, Cambridge,
MA, USA
2
Department of Psychology, University of Michigan,
Ann Arbor, MI, USA

Definition
The term “discrimination learning” refers to the formation of associations between different stimuli and
corresponding outcomes or behaviors. It enables animals to choose different responses for different stimuli.
Models of discrimination learning address (a) the algorithmic principles underlying the implementation of
discrimination learning in the brain and (b) the related
neurophysiological processes that enable the
implementation.

Theoretical Background
Learning to discriminate objects, stimuli, situations,
etc., is a fundamental ability of all animals including
humans. Distinguishing good things from bad things is
essential for survival and any directed behavior. Scientific research on the governing principles of discrimination learning has a long tradition in neuroscience.
Examples include forms of classical and operant conditioning (▶ differential conditioning), where arbitrary sensory stimuli (so-called conditioned stimuli)
and unconditioned stimuli are presented to the animal.
Unconditioned stimuli can be rewarding (e.g., food) or
aversive (e.g., an electric shock). In classical

D

conditioning, the conditioned stimulus is always
followed by the unconditioned stimulus. In operant
conditioning, the animal is required to perform some
action (e.g., a lever press) in order to receive the unconditioned stimulus (i.e., the reward). In differential conditioning, only some stimuli are followed by
unconditioned stimuli, but others are not. For example, when the animal sees a red square appearing on
a monitor, it always receives some food reward a short
time later. However, when a green circle appears instead
on the monitor, the animal does not receive the food
reward. Thereby, in such paradigms, the animal learns
to discriminate stimuli based on their contingencies
with the unconditioned stimuli, so that it will respond
differently to a rewarded stimulus than to an
unrewarded stimulus. In this example, the animal
might increase saliva production only if the red square
appears on the monitor.
Most behavioral studies require some form of discrimination learning and many behavioral paradigms
have been developed to study behavioral and neural
properties of discrimination learning. The most fundamental paradigm used for discrimination learning is
the go/nogo procedure, a form of operant differential
conditioning. In this paradigm, the subject is rewarded
for responding to some stimuli (go) and not rewarded
for responding to other stimuli (nogo). In some variants of this paradigm, not responding to nogo stimuli is
also rewarded. Another common paradigm employing
discrimination learning is choice discrimination. In
choice discrimination, two types of stimuli (one paired
with reward and one not paired with reward) are
presented simultaneously. The animal learns to choose
stimuli associated with reward over the unrewarded
stimuli. Countless variations of these paradigms exist,
including probabilistic reward delivery (e.g., one stimulus yields a reward in 20% of the trials, the other
stimulus in 80% of the trials), switching task contingencies (the rewarded stimulus becomes the
unrewarded one and vice versa), and differential
reward delays or magnitudes (one stimulus yields
a small reward immediately, the other stimulus yields
a larger reward after some time).
In discrimination learning, subjects learn about the
contingencies of specific entities. For efficient behavior,
however, discrimination learning alone would leave
animals inflexible and unable to cope with the sensory
complexity of their environment. In order to behave

1013

D

1014

D

Discrimination Learning Model

efficiently, animals must be able apply their knowledge
to novel stimuli. They do this using ▶ stimulus generalization (see also Kehoe 2008) or by ▶ categorization
(Cook and Smith 2006). Only generalization and categorization allow to respond efficiently to unfamiliar
stimuli and to behave adequately in a complex everchanging environment.

Important Scientific Research and
Open Questions
In general, discrimination learning is subject to three
(overlapping) lines of scientific research. The first line
of research tries to identify the algorithm that underlies
discrimination learning. Here, models of discrimination learning implement different learning algorithms
and compare them to the behavior of the animal.
Different learning algorithms employ, for example, different learning strategies and make different predictions on animal behavior. Discrimination learning
can easily be implemented with machine reinforcement-learning algorithms, such as the prominent temporal-difference algorithm (Sutton and Barto 1998).
This algorithm uses a reward prediction-error as
a teaching signal to detect which stimuli or actions
are consistently followed by rewards. Thereby, the algorithm is able to predict future rewards and also to direct
decision-making toward the most rewarding actions.
Recently, Rose et al. (2009) found that standard reinforcement-learning algorithms can account for the
effect of the magnitude of rewards on learning. In
reinforcement-learning algorithms, the magnitude of
the reward affects how fast a task is learned. Therefore,
if animals use a similar algorithm, the learning speed
should similarly depend on the reward magnitude. In
fact, Rose et al. (2009) showed that animals learned
a task faster when they received a larger amount of food
as a reward, compared to a task where they received
a smaller amount of food as a reward. Thus, the predictions of standard reinforcement-learning algorithms matched the animal behavior.
The second line of research on discrimination
learning tries to identify the neurophysiological processes in implementing some form of reinforcement
learning. In this context, models of discrimination
learning hypothesize on the role of different neural
structures. For example, the dopaminergic system is
a good candidate for discriminating rewarding from
neutral stimuli (see Phillmore 2008). Dopamine is

a modulatory neurotransmitter that plays an important role in synaptic plasticity. A series of studies supports the notion that the activity of dopamine neurons
shows a resemblance to a reinforcement-learning
teaching signal (e.g., see Schultz et al. 1997). Effectively,
dopamine release might tag an arbitrary sensory stimulus as rewarding, and thereby play a crucial role in
learning to discriminate rewarding from nonrewarding stimuli. While the exact role of dopamine
in reward-related learning is still under debate,
a corresponding teaching signal for aversive learning
has not yet been found.
The third line of research aims to reveal how discrimination learning is related to the more complex
process of categorization. Mainly two models have
been discussed in the past, assuming that categorization may be based on exemplars or on a prototype.
Exemplar-based models assume that categorization is
achieved through an extended form of discrimination
learning: by learning a set of stimuli, or exemplars, that,
if taken together, represent the entire category. New
stimuli can then be categorized by comparing them to
the set of exemplars. Prototype-based models, on the
other hand, assume that during learning of a new category one abstract prototype is generated that represents all of the defining features of the category. Recent
research suggests that none of the models alone can
fully account for learning to categorize and that both
processes are involved in categorization (Cook and
Smith 2006). While early stages of learning are dominated by prototype-based categorization, later in learning exemplars are added, which might replace the
prototype.

Cross-References
▶ Adaptive Learning Systems
▶ Anticipatory Learning
▶ Association Learning
▶ Categorical Learning
▶ Computational Models of Human Learning
▶ Formal Learning Theory
▶ Neural Networks of Classical Conditioning
▶ Reinforcement Learning

References
Cook, R. G., & Smith, J. D. (2006). Stages of abstraction and exemplar
memorization in pigeon category learning. Psychological Science,
17, 1059–1067.

Disposition to Understand
Kehoe, E. J. (2008). Discrimination and generalization. In J. H. Byrne
& R. Menzel (Eds.), Learning and memory: A comprehensive reference (Vol. 1, pp. 123–150). Oxford: Elsevier.
Phillmore, L. S. (2008). Discrimination: From behaviour to brain.
Behavioural Processes, 77(2), 285–297.
Rose, J., Schmidt, R., Grabemann, M., & Güntürkün, O. (2009). Theory
meets pigeons: The influence of reward-magnitude on discrimination-learning. Behavioural Brain Research, 198, 125–129.
Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate
of prediction and reward. Science, 275, 1593–1599.
Sutton, R. S., & Barto, A. G. (1998). Reinforcement learning: An
introduction. Cambridge, MA: MIT Press.

Discriminative Law of Effect
▶ Matching

Discussion Group
▶ Group Dynamics and Learning

D

Dishabituation
▶ Habituation and Sensitization

D

Disinhibition
▶ Impulsivity and Reversal Learning

Disinterest
▶ Boredom in Learning

Disorder of Written Expression
▶ Language-Based Learning Disabilities

Disengaged Learners
▶ Apathy in Learning

Disposition
General, relatively stable inclination to approach new
learning tasks and situations in a particular way.

Disengagement
▶ Boredom in Learning

Disequilibrium
Disequilibrium is a situation where internal and/or
external forces prevent system equilibrium from being
reached or cause the system to fall out of balance. This
can be a short-term by-product of a change in variable
factors or a result of long-term structural imbalances.

Cross-References
▶ Cognitive Conflict and Learning

1015

Cross-References
▶ Attitudes – Formation and Change
▶ Personality Effects on Learning

Disposition to Understand
Disposition to understand is a form of thinking disposition identified in university education that suggests
a continuing, strong form of a deep approach to learning. It brings together ability and the willingness to use
it, together with sensitivity to context that allows
a person to recognize opportunities to develop and
use their current understanding.

1016

D

Dispositional Interest

Dispositional Interest
Dispositional interest is a particular way of conceptualizing personal or individual interests as a relatively
stable characteristic of a person or a general orientation
to specific actions. Dispositional interests are considered to be part of a child’s self-concept that influences
participation and learning in activities where the child
has opportunities to initiate interactions with the social
and nonsocial environment.

●

Dispositions for Learning
JEANNE ELLIS ORMROD
School of Psychological Sciences (Emerita), University
of Northern Colorado, Greeley, CO, USA

●

Synonyms
Habits of mind

Definition
A ▶ disposition is a general, relatively stable inclination
to approach new learning tasks and situations in
a particular way. Researchers have identified a variety
of dispositions that have an impact on learning and
performance, sometimes for the better and sometimes
for the worse. Typically, dispositions reflect an
intermingling of cognitive processes, motivational factors, and personality characteristics. To date, the disposition construct has been used primarily within the
context of human (rather than nonhuman animal)
learning.

●

Theoretical Background
Psychologists have long observed individual differences
in how people approach and thereby benefit from new
activities and learning opportunities. Some of these
individual-difference variables are relatively stable
over time and across diverse contexts. Some examples
of these variables are described and discussed below:
● Stimulation seeking: Some people are more inclined

than others to actively seek out new information
and learning experiences. This disposition may

●

reflect individual differences in the need for arousal
and in the optimal level of arousal at which people
feel most comfortable. For example, any given level
of stimulation might feel boring to some individuals, comforting to others, and overwhelming to
still others.
Need for cognition: Just as people vary in their need
for stimulation in general, they also vary in the
extent to which they actively seek out and engage
in challenging cognitive tasks. For example,
a person with a strong need for cognition might
avidly read books on a broad range of topics,
eagerly seek out brainteaser puzzles, or voluntarily
engage in debates about controversial issues. This
disposition, too, may reflect varying levels of the
need for arousal.
Conscientiousness: Some people consistently tackle
learning tasks and activities in a deliberative, careful, thorough manner. For example, they are apt to
plan ahead, and they can be relied on to get a job
done. Although conscientiousness does, in general,
lead to enhanced learning, in its extreme form –
perfectionism – it can lead to debilitating anxiety
levels and overly harsh self-evaluations. Conscientiousness is one of the “Big Five” orthogonal personality traits that some personality theorists have
described.
Learned industriousness: People differ in the degree
to which they persevere in attempting to master
a new topic or skill even when they need to exert
considerable effort or face obstacles to their success.
People with a low level of learned industriousness
give up quickly in the face of failure; those with
a high level “try, try again.” Quite possibly people
acquire learned industriousness when past experiences have taught them that successes often come
only with effort and effective strategies. In contrast,
people who have previously been accustomed to
easy, effortless successes (e.g., as might be true for
especially gifted individuals) may abandon an
endeavor at the first stumbling block and may possibly develop learned helplessness about the domain
involved.
Open-mindedness: People also differ in the extent to
which they can flexibly consider alternative, potentially contradictory perspectives and evidence.
Open-minded individuals typically suspend judgment about topics until they have enough

Dissatisfaction

information to make an informed decision; in some
cases they delay on choosing among conflicting
ideas indefinitely. In contrast to open-mindedness,
people with a strong need for closure tend to jump
very quickly to conclusions about which facts, perspectives, theories, and so on are accurate or reasonable and which are not, often with little or no
consideration of supporting or contradictory
evidence.
● Critical thinking : Critical thinking involves not
only a set of fairly sophisticated cognitive skills but
also a disposition to use those skills. People
who have a disposition for critical thinking consistently evaluate new information and arguments
in terms of their accuracy, logic, and credibility.
In contrast, noncritical thinkers tend to accept
new ideas at face value, with little or no reflection
and analysis.
● Consensus seeking : People with a high need for
consensus seeking strive to synthesize diverse perspectives into a more complex understanding of
a topic or phenomenon than any single perspective
could provide. People without this disposition are
more likely to assume that diverse perspectives
must necessarily be mutually exclusive and thus
are apt to determine that only one perspective can
have validity.

Important Scientific Research and
Open Questions
Studies of the nature and effects of dispositions have
been relatively small in number, but it has become
increasingly clear that these cognition–motivation–
personality blends can have a significant impact on
learning. In fact, dispositions sometimes overrule intelligence in their impact on learning and achievement
(Perkins and Ritchhart 2004). For example, people who
show a consistent tendency to seek out physical and
cognitive stimulation tend to learn more from what
they read and to be higher achievers in instructional
settings (Cacioppo et al. 1996; Raine et al. 2002). And
people who are predisposed to critically evaluate new
information are more likely to revise existing beliefs to
be in line with scientifically validated theories
(Southerland and Sinatra 2003).
The origins of various dispositions remain
largely unexplored. Possibly inherited temperamental
differences play a role in such dispositions as

D

stimulation seeking and need for cognition
(Raine et al. 2002). It appears, too, that people’s beliefs
about the nature of particular academic disciplines and
about knowledge in general – that is, people’s epistemological beliefs (also known as epistemic beliefs) –
predispose them to think critically and analytically
about diverse explanations, on the one hand, or to
“zero in” very quickly on just a single explanation, on
the other (DeBacker and Crowson 2009). Culture, too,
seems to play a role; for example, whereas some
cultural groups encourage critical thinking, others
emphasize that wisdom is best obtained from authority
figures or religious teachings and that certain
perspectives must never be questioned (Kuhn and
Park 2005).

Cross-References
▶ Conditions of Learning
▶ Epistemological Beliefs and Learning
▶ Openness to Experience

References
Cacioppo, J. T., Petty, R. E., Feinstein, J. A., & Jarvis, W. B. G. (1996).
Dispositional differences in cognitive motivation: The life and
times of individuals varying in need for cognition. Psychological
Bulletin, 119, 197–253.
DeBacker, T. K., & Crowson, H. M. (2009). The influence of need for
closure on learning and teaching. Educational Psychology Review,
21, 303–323.
Kuhn, D., & Park, S.-H. (2005). Epistemological understanding and
the development of intellectual values. International Journal of
Educational Research, 43, 111–124.
Perkins, D., & Ritchhart, R. (2004). When is good thinking? In D. Y.
Dai & R. J. Sternberg (Eds.), Motivation, emotion, and cognition:
Integrative perspectives on intellectual functioning and development (pp. 351–384). Mahwah: Erlbaum.
Raine, A., Reynolds, C., & Venables, P. H. (2002). Stimulation seeking
and intelligence: A prospective longitudinal study. Journal of
Personality and Social Psychology, 82, 663–674.
Southerland, S. A., & Sinatra, G. M. (2003). Learning about biological
evolution: A special case of intentional conceptual change. In
G. M. Sinatra & P. R. Pintrich (Eds.), Intentional conceptual
change (pp. 317–345). Mahwah: Erlbaum.

Dissatisfaction
▶ Boredom in Learning

1017

D

1018

D

Dissociation

Dissociation
Impairment in one aspect of a normally integrated cognitive function, with preservation of other aspects of the
function, as a result of a lesion in the brain. A study of
lesion-induced dissociations (or fractionations) is helpful in suggesting whether the lesioned brain structure is
a necessary substrate of the function in question.

Definition
Distance learning is an outcome of distance education.
Where learners and teachers are separated by geographical and/or temporal distance, a form of mediated learning can be achieved using a combination of technologies.
Distance learning can be differentiated from e-learning,
which may be undertaken at a distance or contiguously,
or as a combination of both (blended learning). Moore
and Kearsley define distance education as:
"

Dissonance Reduction Theory
Cognitive dissonance is a communication theory
developed by Leon Festinger (1957) which contrasts
with behaviorist conditioning or reinforcement theories.
The dissonance reduction theory considers individuals as
purposeful decision makers who strive for harmony in
their beliefs. If presented with decisions or information
that evoke dissonance, individuals apply dissonance
reduction strategies in order to regain a state of equilibrium, especially if the dissonance affects their self-esteem.

References
Festinger, L. (1957). A theory of cognitive dissonance. Stanford:
Stanford University Press.

Distance Education
▶ Affective and Cognitive Learning in the Online
Classroom
▶ Distance Learning
▶ Online Learning

Distance Learning

. . .planned learning that normally occurs in a different
place from teaching and as a result requires special
techniques of course design, special instructional techniques, special methods of communication by electronic and other technology, as well as special
organizational and administrative arrangements.
(Moore and Kearsley 1996, p. 2)

Some of the earliest forms of distance education,
known as correspondence courses, were achieved using
a combination of printed material and the postal service. An early example of this was the Pitman shorthand course set up in Victorian Britain, but some argue
that correspondence courses can be traced back as far as
the instructional writings (epistles) of St Paul to the
early Christian church. In current practice, distance
learning may be supported through a combination of
technologies that include broadcast TV and radio,
printed materials, Web-based instruction, videoconferencing, and mobile communication. A number of new
tools, notably social web services including wikis, blogs,
podcasts, and social networking are also being investigated (Wheeler et al. 2008; Selwyn and Grant 2009).
In the most generally accepted approaches to distance education, the emphasis has shifted away from
the institute and onto the learner. Learning opportunities are provided at convenient times and places for the
learner, rather than for the institution. Keegan (1990)
has identified four key elements of distance learning:
● The teacher is separated from the student by

distance.
STEVE WHEELER
University of Plymouth, Plymouth, Devon, UK

● The influence of an educational organization and

the provision of student evaluation.
● The use of educational media to carry the course

content.

Synonyms
Blended learning; Correspondence courses; Distance
education; Distributed learning; Remote education

● Provision of two-way communication between

teacher and student and between student and
student.

Distance Learning

Theoretical Background
Some see distance learning as an extension of traditional provision, and some may even view it as a means
to widen traditional student catchment areas. Often
distance learning methods are employed as a way to
improve access for previously disenfranchised individuals. However, such views may be an oversimplification. In the past two decades, distance education has
become an important discipline in its own right and
a significant educational movement, as the vast growth
in the literature confirms (Bernath et al. 2009). Indeed,
there are many specialist peer-reviewed academic publications in circulation and the growth of interest in
distance learning can be evidenced in the high attendance at international distance education and educational technology conferences around the world in
recent years.
Hoffman et al. (2000) argue that regardless of the
time and place constraints imposed upon distance
learners, the goal of distance education is still to bring
about positive changes in student behavior just as with
traditional on campus education. Distance education
can and does promote positive changes. Instruction
becomes more learner centered, with students enjoying
access to learning events and resources that can be
adjusted to meet their individual learning needs and
styles (Simonson et al. 2000). Distance education,
when managed effectively, can provide remote learners
with an equivalent quality of experience to learners
studying in more traditional settings.

Important Scientific Research and
Open Questions
Distance education is often characterized by the extent
to which students can choose their mode, place, and
pace of study. This kind of flexibility is almost unheard
of in many traditional forms of education, and questions arise about how much autonomy students should
be given by the institution, and to what extent student
agency can be reconciled with the requirements and
regulations of the institution.
Distance learning can be conceptualized as an activity that is location and time independent, encouraging
students to assume the responsibility for their own learning. In many cases, the teacher is more remote and less
accessible than the learning materials, acting as another
learning resource rather than as a central component in
the learning process. The extent to which the available

D

learning materials will be used is dependent upon how
highly the students value their usefulness. The learning
materials may often be accessed and read in whatever
order and depth the student chooses, particularly if they
are presented in web-based formats. These attributes run
counter to the traditional model, where the course material selected by the teacher is transmitted to the student
sequentially leaving little room for student agency
(Valjataga and Laanpere 2010).
The tension between the two models requires investigation. Those institutions that allow students complete autonomy within the learning process are few and
far between. Those that do offer complete freedom
generally remove rules such as completion times and
set marking periods, preferring to steer clear of any
intervention that would constrain freedom in learning.
Opposed to this approach are those institutions that
insist upon a semblance of structure within the freedom that derives from distance learning. Such course
providers generally consider the imposition of some
structure as a means of preventing academic failure. It
is well documented in the literature that there can be
high levels of attrition rates amongst distance learning
populations. Some have speculated that this is due to
the reduced social contact with tutors and peers,
resulting in loss of motivation. Another theory is that
students who study away from an institution may be
disadvantaged through lack of access to the vital learning resources enjoyed by on-campus learners.
Other issues under investigation also center upon
student experiences, where the transactional distance,
or the “instructional gap” may play a part in creating
a psychological distancing between students and their
teachers, leading to misunderstandings and perceptions of social isolation (Moore and Kearsley 1996).
There is also a question whether distance learning is
more or less effective than traditional, campus-based
learning. Several sources indicate that there is no significant difference between the two modes of learning
(Merisotas and Phipps 1999). However, the recent
emergence of Web 2.0-based tools and services (social
networking, wikis, blogs, and podcasts) has prompted
increased research activity into the use of social media
and their effectiveness in distance education. Furthermore, the growing number of user-generated content
repositories such as Wikipedia has raised concerns over
accuracy and veracity of information found on the web,
and draw new attention to the blurring of boundaries

1019

D

1020

D

Distant Associations

between what it means to be a knowledgeable contributor and a credentialed expert.
A number of tensions therefore exist between an
idealistic approach to distance education where students are given complete control over their own study
schedules, and can select their own personal media and
technologies; and a practical approach where the
learners are provided with a certain amount of structure designed to maintain the impetus of their studies
and to motivate them toward completing their assignments within a set time. It is probable that many
distance educators will opt for the middle ground and
borrow from the best of both philosophies to provide
the best balance of freedom and control.

Cross-References
▶ Distributed Technologies
▶ Tele-Learning

Distinction
▶ Simultaneous Discrimination Learning in Animals

Distributed Cognition
▶ Shared Cognition
▶ Social Construction of Learning

Distributed Intelligence/
Cognition
A shared task between the learner and another person,
tool, artifact, or resource.

References
Bernath, U., Szucs, A., Tait, A., & Vidal, M. (Eds.). (2009). Distance
and e-learning in transition. London: Wiley.
Hoffman, S. Q., Martin, M. S., & Jackson, J. E. (2000). Using the
theory of equivalency to bring onsite and online learning
together. Quarterly Review of Distance Education, 1, 327–335.
Keegan, D. (1990). Foundations of distance education. London:
Routledge.
Merisotas, J. P., & Phipps, R. A. (1999). What’s the difference? Outcomes of distance v. traditional classroom-based learning.
Change, 31, 13–17.
Moore, M. G., & Kearsley, G. (1996). Distance education: A systems
view. Belmont: Wadsworth.
Selwyn, N., & Grant, L. (2009). Researching the realities of social
software: An introduction. Learning, Media and Technology, 34,
79–86.
Simonson, M., Smaldino, S., Albright, M., & Zvacek, S. (2000).
Teaching and learning at a distance: Foundations of distance education. Upper Saddle River: Merrill.
Valjataga, T., & Laanpere, M. (2010). Learner control and personal
learning environment: A challenge for instructional design. Interactive Learning Environments, 18, 277–292.
Wheeler, S., Yeomans, P., & Wheeler, D. (2008). The good, the bad
and the wiki: Evaluating student generated content for collaborative learning. British Journal of Educational Technology, 39,
987–995.

Distant Associations
▶ Learning Nonadjacent Dependencies

Distributed Learning
JAMIE KIRKLEY
Information in Place Inc., Indiana University Research
Park, Bloomington, Indiana, USA

Usually distributed learning or spaced learning is
defined as opposed to massed learning. Distributed
learning means that the material to be learned is distributed over a long period of time so that the learner
must integrate the various separated parts of material
into a unique entity. Contrarily, massed learning means
that the material to be learned is provided within
a short period of time.
Distributed learning is grounded on the assumption that long-term memory will be improved when
there is more time between acquisition and retrieval of
information. Accordingly, it has been argued (Litman
and Davachi 2008) that it would be better for exams to
be taken after a break than before, assuming there was
a review before the exams, because of the spacing effect.
Closely related with distributed or spaced learning
is the idea of technology-based distributed learning
environments which integrate the interactive

Distributed Technologies

capabilities of networking, computing, and multimedia with learner-centered collaboration and discovery
learning. It can be argued that distributed learning
environments require not only spaced learning
but also information aggregation aiming at collecting
relevant information from multiple sources
(Dutta et al. 2005).

Cross-References
▶ Distance Learning

References
Dutta, P. S., Jennings, N. R., & Moreau, L. (2005). Cooperative
information sharing to improve distributed learning in multiagent systems. Journal of Artificial Intelligence Research, 24,
407–463.
Litman, L., & Davachi, L. (2008). Distributed learning enhances
relational memory consolidation. Learning and Memory, 15(9),
711–716.

Distributed Learning
Environments
▶ Interactive Learning Environments

Distributed Learning Model
▶ Advanced Distributed Learning

Distributed Practice
▶ Trial-Spacing Effect in Associative Learning

Distributed Scaffolding
Multiple forms of support in the forms of tools, artifacts, resources, and people within the classroom.

D

1021

Distributed Technologies
JOSEPH PSOTKA
Basic Research Unit, US Army Research Institute for
the Behavioral and Social Sciences, Arlington, VA, USA

D
Synonyms
Adaptive intelligent web–based teaching and learning;
Asynchronous learning; Blended learning; Distance
learning

Definition
Distributed technologies take advantage of computational environments to bring students and teachers
together, either synchronously or asynchronously, in
ways that bring all the advantages of face-to-face collaboration, along with the augmented capabilities of
computer mediation, analysis, search of online
resources, and faithful recording of all interactions.
The Internet has encompassed older distributed technologies, such as video, print, movies, radio, and television, and integrated them thoroughly into one
medium, so that Web 2.0 (or even 3.0) technologies
now constitute the core of distributed learning and
blogging, videoconferencing, and e-mail; all parts of
distributed technologies for teaching and learning.
However, these technologies are continuing to transform at a rapid pace, with handheld, wireless devices
creating the most profound opportunities and issues
for the next few years.

Theoretical Background
Distributed technologies, particularly those on the
Internet (Web 2.0 and 3.0) are the hottest innovation
in education, and the source of new online industries,
whether for full-fledged higher education Universities
or tools and technologies to support classroom instruction. Sadly, many thousands of new web-based educational applications are little more than static web pages
of hypertext that are widely giving computer-based
learning a bad name. What is possible and what is
actual are divided by a coulisse of grand proportions.
Prior to the easy distributed access provided by the
Internet, computer-based education was dominated by
single user, standalone systems, where the computer
had only to build a single user model, provide guidance

1022

D

Distributed Technologies

and problem-solving assistance to a manageable set
of errors and misconceptions, and guide the course of
instruction through only one level and area of a
curriculum, usually in one class or one schoolhouse.
Now with distributed technologies on the web,
possibilities and problems have multiplied manifold.
The new critical research issues center around
cooperative and collaborative learning, but not of
facts and carefully digested knowledge; but instead, of
creativity, investigation, connection, integration, and
synthesis. Problem solving and cognitive flexibility
have become the driving force of educational research
using distributed technologies, but remain largely
invisible in educational practice. As part of the leading
edge of implementing these technologies in classrooms,
learning management systems (LMS) have to integrate
and manage students nationwide. Increasingly LMS
have to do more than simply tabulate grades and test
results for teachers. Instead they must provide efficient
communication among students, through blogs (web
logs of written materials), wikis, and threaded discussion systems, as well as provide teaching and learning
tools of increasing sophistication. Social networking
aspects of instruction and learning have become paramount within LMS, with group projects, WebQuests,
and active online engagement that is the hallmark of
constructivist learning activities. LMS are continuing
to undergo technological improvements in order to
provide support for online collaboration and knowledge construction.
Distributed technologies provide the opportunity
to enrich the range and immediacy of resources for
students in conjunction with interactive classroom lectures and activities. Naturally, this opportunity
increases in scope with increasing age and sophistication of students from elementary, secondary, and
higher education. Distributed technology creates the
unique and largely unexplored possibility of bringing
the real world into the classroom and taking the classroom into the real world. The additional resources it
brings range from interactivity based on multimedia
and online semantic content to interactivity based on
peers, other cultures, and world-class experts. Given
the richness of these possibilities, it is little wonder
that educational systems have only explored a small
part of this vast space. One of the barriers to its effective
implementation is just this richness, the overload of
information it offers, and the difficulty of making the

right information at the appropriate level of complexity and difficulty available to students at the moment
they need it. Peers can share their information in
just this way, but their information may be too limited.
Collaborative filtering is one common approach to
making information available by a consensus of peers.
As search and retrieval technologies become ever
more sophisticated so that search is not Boolean, or
based on keywords, but is instead leveraged with
broad natural language capabilities that cross international boundaries, or semantic search derived from
the gist of entire text or video paragraphs or chapters,
these resources will become more effectively integrated
into distributed interactive learning environments.
They will create new challenges for teachers and the
educational system and leadership: how to make use of
this knowledge outside the curriculum; how to generate challenging topics that generate collaboration
among students; how to create novel assessments
that accurately measure what students have learned;
and how to measure growth in accuracy of opinions
and judgments, not just facts.
One of the most widely adopted activities in K-12
classrooms using distributed technologies has been the
WebQuest. Of the conclusions reached by diverse
research studies conducted about WebQuests, the
most common findings are positive attitudes and perceptions among students, increases in their motivation,
and improved collaboration skills. However, rarely
have they found a significant impact on learning and
achievement, since these critical attributes of motivation, attitudes, and collaboration are seldom measured
as part of classroom skills and objectives.
One of the most exciting areas of distributed technology development is occurring with handheld, wireless streaming augmented reality. Carrying cell phones
or pdas, students can learn in real time about the
geographic areas and specific locations, artifacts, and
objects in their surroundings. Using techniques such as
geospatial gps systems, local infrared, radio frequency,
or wireless signals, or short-range Bluetooth technology, students can engage in real situated learning, interact naturally and manipulate physical objects, or enter
shared spaces for collaborative transactions. Creating
these designed spaces for effective learning represents
yet another challenge for educators. The widespread
popularity of flash mobs and swarms of people engaged
in fantasy skits suggests real possibilities for historical

Divergent Probabilistic Judgments Under Bayesian Learning with Nonadditive Beliefs

reenactment, informal science ecological environments, and distributed museums and zoos; but all
this remains to be explored.

Important Scientific Research and
Open Questions
The critical research issues are embedded in this whole
entry, but the largest issue is how education will be
transformed by these new technologies. Will schools
and universities diversify to put education into homes
and communities, with special social activities
performed in group places? Will stratified, lock step
grades be replaced by individualized diplomas and
certification of accomplishments on a domain-bydomain and level-by-level basis? Will it become more
commercial and dominated by large corporations? Will
it exploit these new social, distributed, and interactive
games and simulations to transform education and
make it more effective for all ranges of abilities? Will
leadership and policy makers begin to understand the
vast potential of distributed technologies, to encourage
huge new investments in research that will be evaluated
and moved rapidly into fundamental new structures
and organizations for education?

Cross-References
▶ Advanced Distributed Learning
▶ Asynchronous Learning
▶ Asynchronous Learning Networks
▶ Classroom Management and Motivation
▶ Collaborative Knowledge Building
▶ Collaborative Learning
▶ Collaborative Learning and Critical Thinking
▶ Collaborative Learning Strategies
▶ Collaborative Learning Supported by Digital Media
▶ Collective Development and the Learning Paradox
▶ Collective Learning
▶ Community of Learners
▶ Computer-Supported Collaborative Learning
▶ Discourse in Asynchronous Learning Networks
▶ Distributed Learning
▶ Distributed Technologies
▶ Learning Management System
▶ Learning Object Evolutions in a Distributed
Environment
▶ Learning with Collaborative Mobile Technologies
▶ Online Collaborative Learning
▶ Online Learning

D

▶ Online Learning with Monte Carlo Methods
▶ Ontology of Learning Objects Repository for Knowledge Sharing
▶ Open Learning
▶ Open Learning Environments
▶ Peer Learning and Assessment
▶ Rapid Collaborative Knowledge Improvement
▶ Zone of Proximal Development

References
Brusilovsky, P. (1999). Adaptive and intelligent technologies for webbased education. In C. Rollinger., & C. Peylo (Eds), Künstliche
Intelligenz (4), Special Issue on Intelligent Systems and
Teleteaching (pp. 19–25). http://www2.sis.pitt.edu/~peterb/
papers/KI-review.html. Accessed 28, April 2011.
Fletcher, J. D. (2009). Education and training technology in the
military. Science, 353(2), 72–79.
Herlocker, J. L., Konstan, J. A., & Riedl, J. (2000). Explaining collaborative
filtering recommendations. Proceedings of the 2000 ACM conference on Computer supported cooperative work (pp. 241–250).
Philadelphia. [doi:10.1145/358916.358995]
Landauer, T. K., & Psotka, J. (2000). Simulating text understanding
for educational applications with latent semantic analysis: introduction to LSA. Interactive Learning Environments, 8, 73–76.
Mayer, R. E., Dow, G. T., & Mayer, S. (2003). Multimedia learning in
an interactive self-explaining environment: what works in the
design of agent-based microworlds? Journal of Educational Psychology, 95(4), 806–813.
Reily, K., Ludford Finnerty, P., & Terveen, L. (2009). Two peers are
better than one: aggregating peer reviews for computing assignments is surprisingly accurate. Proceedings of the ACM 2009
international conference on Supporting group work (pp. 10–13).
Sanibel Island. [doi:10.1145/1531674.1531692]

Divergent Probabilistic
Judgments Under Bayesian
Learning with Nonadditive
Beliefs
ALEXANDER ZIMPER
School of Economic and Business Sciences, University
of the Witwatersrand, Johannesburg, Gauteng, South
Africa

Synonyms
Attitude polarization; Irrational belief persistence;
Myside bias

1023

D

1024

D

Divergent Probabilistic Judgments Under Bayesian Learning with Nonadditive Beliefs

Definitions
– A nonadditive probability measure n defined on the
measurable space (O, F ) satisfies (1) normalization,
i.e., n(O) = 1, nð;Þ ¼ 0 as well as (2) monotonicity,
i.e., AB implies n(A)n(B) for all A, B 2 F .
– The Choquet expected value of a bounded random
variable Y : O !  with respect to a nonadditive
probability measure n is defined as the following
Riemann integral:
Z 0
E½Y ; n ¼
ðnðfo 2 OjY ðoÞ  zgÞ  1Þ dz
1
Z þ1
nðfo 2 OjY ðoÞ  zgÞ dz:
þ
0

Theoretical Background
Standard Bayesian decision theory in the tradition of
Ramsey, de Finetti, and Savage considers a decision
maker whose uncertainty is comprehensively described
by some additive probability space (m, O, F ). The
concept of Bayesian learning refers to the specific situation where the state space O is rich enough to include
a data-space – consisting of (possibly infinite) sequences
of outcomes of an i.i.d. random process – as well as
a parameter space that determines the distribution of
this process. New information in terms of new data
then gives rise to probabilistic learning in the sense that
the decision maker revises his prior estimate of the distribution parameter. The resulting posterior estimate is
thereby uniquely determined as the expected value of the
random distribution parameter with respect to the probability measure m conditioned on the newly received
information. As a consequence, models of Bayesian
learning provide a decision-theoretically sound answer
to the question of how an agent learns probabilistic
judgments under the assumption that sample data is
drawn from an i.i.d. process. Moreover, celebrated consistency results on Bayesian estimators (e.g., Doob
1949) ensure that these probabilistic judgments will
almost certainly coincide with the true parameter
values of standard i.i.d. processes if the learning process
incorporates sufficiently many data observations.
In spite of the elegance of standard Bayesian models
of probabilistic learning, two critical remarks about the
descriptive shortcomings of these models are in order.
Firstly, several studies in the psychological literature
demonstrate that people’s learning behavior may be

prone to effects such as “myside bias” or “irrational
belief persistence” which may give rise to “attitude
polarization” (cf. Baron 2008). The learning behavior
elicited in these experiments cannot be explained by
standard models of Bayesian learning according to
which differences in agents’ probabilistic judgments
must decrease rather than increase whenever the agents
receive identical information. Secondly, recent decision-theoretic developments suggest that additive
probability measures describe subjective uncertainty
in a rather unsatisfactory way since they neglect ambiguity attitudes that are relevant to real-life probabilistic
judgments. This entry introduces a closed-form model
of Bayesian learning that addresses in a unified way
both descriptive shortcomings of standard models of
Bayesian learning. Our formal approach is based on
Choquet Expected Utility (CEU) theory, which considers nonadditive probability measures, i.e., capacities,
in order to describe violations of Savage’s (1954) surething principle as elicited by paradoxes of the Ellsberg
type (Schmeidler 1986). More specifically, the agent’s
estimate about the parameter value is given as the
parameter’s Choquet expected value with respect to
a conditional neo-additive capacity in the sense of
Chateauneuf et al. (2007) according to which an agent’s
nonadditive belief about the likelihood of an event is
a weighted average of an ambiguous part and an additive
part. The key to our model is the existence of several
perceivable Bayesian update rules for nonadditive probability measures that may express different psychological
attitudes toward the interpretation of new information.
More precisely, we consider the so-called full Bayesian as
well as the optimistic and the pessimistic update rule
(Gilboa and Schmeidler 1993). As explained below, this
“indeterminacy” of update rules for nonadditive probability measures is a direct consequence of the violation
of Savage’s sure-thing principle.
Finally, using this Bayesian learning model with conditional neo-additive capacities, we then analyze the
behavior of revised probabilistic judgments of two heterogeneous agents. Two main findings on the possibility
of diverging judgments emerge from our formal model:
1. We may observe divergent probabilistic judgments
for agents who have identical attitudes with respect
to the interpretation of new information but have
different initial attitudes with respect to optimism,
or pessimism, under ambiguity.

Divergent Probabilistic Judgments Under Bayesian Learning with Nonadditive Beliefs

2. We may observe divergent probabilistic
judgments in case the agents have identical
initial attitudes with respect to optimism, or
pessimism, under ambiguity but have different
attitudes with respect to the interpretation of new
information.

Important Scientific Research and
Open Questions
Bayesian Learning with Additive
Probability Measures
We present at first a specific closed-form model that
will serve as our benchmark model of Bayesian learning
with an additive probability measure. Suppose that an
agent observes an arbitrary number of independent
trials in which a specific outcome – say Heads – occurs
with identical probability. Formally, we describe this
situation by a probability space (m, O, F ) where p
denotes the event in F such that p 2 [0, 1] is H’s true
probability, i.e.,
p ¼ fo 2 Oj~
pðoÞ ¼ pg
~ denotes a random variable with range [0, 1].
where p
~ is given as
We assume that the agent’s prior over p
a Beta distribution with parameters a, b>0 so that,
for all p 2 [0, 1],
mðpÞ ¼ Ka;b pa1 ð1  pÞb1
where Ka,b is a normalizing constant. While the agent
will never observe any direct information about H’s
true probability, he receives in period n sample information Ikn which is formally defined as the event in
F such that H has occurred k-times in the n first trials,
i.e.,
Ikn ¼ fo 2 OjIn ðoÞ ¼ kg
whereby the random variable In counts the number of
occurrences of the outcome H in the first n trials. By
our i.i.d. assumption, In is, conditional on the parameter value p, binomially distributed with probabilities
 



n k
m Ikn jp ¼
p ð1  pÞnk for k 2 f0; . . . ; ng:
k
By Bayes’ rule, we then obtain the following posterior probability that p is the true parameter value
conditional on information Ikn

D

1025





 k  m p \ Ikn

 
m pjIn ¼
m Ikn



m Ikn jp mðpÞ



¼R
k
½0;1 m In jp mðpÞ dp
¼ Kaþk;bþnk paþk1 ð1  pÞbþnk1 :

D

The agent’s prior probabilistic judgment, i.e., his
prior estimate of the true probability of outcome H, is
~ with respect to the
defined as the expected value of p
a
prior Beta distribution, i.e., E½~
p; m ¼ aþb
. Accordingly,
the agent’s revised probabilistic judgment in the light of
~
information Ikn is defined as the expected value of p
with
 respect

 	to the resulting posterior distribution,

  i.e.,
~; m 	jIkn . Since the agent’s posterior m 	jIkn over p
~
E p
is itself a Beta distribution
with
parameters
a+k,
b+


 	
aþk
~; m 	jIkn ¼ aþbþn
nk, we have E p
or, equivalently,





 
aþb
n
k
~; m 	jIkn ¼
:
E ½p
E½~
p; m þ
aþbþn
aþbþn n
ð1Þ
Suppose now that there are two agents with different subjective probability measures m1 and m2, respectively. Since both agents’ posterior estimates put
increasing weight on the commonly observed samplemean nk , the following convergence result for the Bayesian learning model (1) follows immediately.
Proposition 1. Consider additive probability measures
m1 and m2 such that E½~
p; m1  > E½~
p; m2 . Then the difference in agents’ probabilistic judgments strictly decreases
in the number of observations, i.e., for all n,



	



	


 	
~; m1 	jIknþ1  E p
~; m2 	jIknþ1 < E p
~; m1 	jIkn
E p


 	
~; m2 	jIkn :
E p

Updating Nonadditive Probability
Measures
As a generalization of an additive probability space, we
now consider a nonadditive probability space (n, O, F ).
Properties of nonadditive probabilities are used in the
literature on Choquet expected utility (CEU) theory
for formal definitions of, e.g., ambiguity and uncertainty attitudes, pessimism and optimism as well as
sensitivity to changes in likelihood.
CEU theory has been developed in order to accommodate paradoxes of the Ellsberg type which show that
real-life decision makers violate Savage’s sure-thing

1026

D

Divergent Probabilistic Judgments Under Bayesian Learning with Nonadditive Beliefs

principle. The abandoning of the sure-thing principle
implies that there exist several perceivable Bayesian
update rules for nonadditive probability measures. To
see this, define the Savage-act fBh: O!X such that
(
f ðoÞ for o 2 B
fB hðoÞ ¼
hðoÞ for o 2 :B
where B is some nonempty event. That is, the act fBh
gives the same consequences as the act f in all states
belonging to event B and it gives the same consequences as the act h in all states outside of event B.
Recall that Savage’s sure-thing principle states that, for
all acts f, g, h, h0 and all events B 2 F ,
fB h 
 gB h implies

fB h0 
 gB h0 :

ð2Þ

Let us now interpret event B as new information
received by the agent. The sure-thing principle then
implies a straightforward way for deriving ex post
preferences 
B , conditional on the new information
B, from the agent’s original preferences 
 over Savageacts. Namely, we have
f 
B g if and only if fB h 
 gB h for any h:

ð3Þ

For a subjective EU decision maker (3) implies the
familiar definition of a conditional additive probability
measure, i.e., for all A, B 2 F such that m(B)>0, by
mðAjBÞ ¼

mðA \ BÞ
:
mðBÞ

In case the sure-thing principle does not hold, the
specification of act h in (3) is no longer arbitrary. For
CEU preferences, there therefore exist several possibilities of deriving ex post preferences from ex ante preferences. Let us at first consider conditional CEU
preferences satisfying, for all acts f, g,
f 
B g if and only if fB h 
 gB h

ð4Þ

where h is the so-called conditional certainty equivalent
of g, i.e., given information B the agent is indifferent
between the act g and the act h that gives in every state
of B the same consequence. The corresponding Bayesian update rule for the nonadditive beliefs of a CEU
decision maker is the so-called full Bayesian update rule
which is given by
nFB ðAjBÞ ¼

nðA \ BÞ
;
nðA \ BÞ þ 1  nðA [ :BÞ

ð5Þ

where nFB (A|B) denotes the conditional capacity for
event A 2 F given information B 2 F .
In addition to the full Bayesian update rule, we also
consider the optimistic and the pessimistic update rule
as introduced by Gilboa and Schmeidler (1993). For the
so-called optimistic update rule, h in (4) is the constant
act that gives in every state the worst possible consequence so that the impossible event NOT B becomes
associated with the worst outcome possible. As
corresponding optimistic Bayesian update rule for conditional beliefs of CEU decision makers, we obtain
nopt ðAjBÞ ¼

nðA \ BÞ
:
nðBÞ

ð6Þ

For the pessimistic update rule, h is the constant act
associating with the impossible event the best outcome
possible. The corresponding pessimistic Bayesian
update rule for CEU decision makers is
npess ðAjBÞ ¼

nðA [ :BÞ  nð:BÞ
:
1  nð:BÞ

ð7Þ

Bayesian Learning with Neo-additive
Probabilities
Let us now formally link the updating of nonadditive
probabilities to Bayesian learning behavior. Our own
approach focuses thereby on nonadditive probability
measures that are defined as neo-additive capacities in
the sense of Chateauneuf et al. (2007).
Definition (Neo-additive Capacities). Given the measurable space (O, F ) the neo-additive capacity, n, is
defined, for some d, l 2 [0, 1] by
nðAÞ ¼ d 	 l þ ð1  dÞ 	 mðAÞ

ð8Þ

for all A 2 F such that A 2
= f;; Og.
Neo-additive capacities can be interpreted as nonadditive beliefs that stand for deviations from additive
beliefs such that a parameter d, the degree of ambiguity,
measures the lack of confidence the decision maker has
in some subjective additive probability distribution m.
The second parameter l is typically interpreted as the
degree of optimism under ambiguity whereby l = 1 and l
= 0 correspond to extreme optimism and extreme
pessimism, respectively.
The Choquet expected value of the random variable
~ with respect to a neo-additive capacity gives the
p

Divergent Probabilistic Judgments Under Bayesian Learning with Nonadditive Beliefs

following prior (Choquet) estimate of the true parameter value


~ðoÞ þ ð1  lÞ min p
~ðoÞ
E½~
p; n ¼ d l max p
o2O

o2O

þ ð1  dÞE½~
p; m
¼ dl þ ð1  dÞE½~
p; m:
The following observation characterizes posterior
(Choquet) estimates in the light of sample information
Ikn for the different update rules discussed in the previous section.
Observation. Contingent on the applied update rule the
agent’s posterior estimate conditional on information Ikn
is given as follows.
1. Full Bayesian updating




 FB 
 k 	


 	
FB
~; n 	jIn ¼ dFB
~; m 	jIkn
E p
	E p
I k l þ 1  dI k
n

n

whereby
dFB
Ik ¼
n

d

 :
d þ ð1  dÞ 	 m Ikn

2. Optimistic Bayesian updating

  
 	
 opt 
 k 	
opt
~;n 	jIn ¼ dopt
~; m 	jIkn
E p
	E p
k þ 1d k
I
I
n

n

whereby
opt

dI k ¼
n

d	l

 :
d 	 l þ ð1  dÞ 	 m Ikn

3. Pessimistic Bayesian updating
  
 	
 pess 
 k 	 
pess
~;n
~; m 	jIkn
E p
	jIn ¼ 1  dIk 	 E p
n

whereby
pess

dIk ¼
n

d 	 ð1  lÞ

 :
d 	 ð1  lÞ þ ð1  dÞ 	 m Ikn

Observe that the agent’s posterior estimates are given as
a weighted
 
 average
	 of the additive benchmark estima~; m 	jIkn , as given by (1), and of the numbers l
tor E p
(for full Bayesian learning), 1 (for optimistic Bayesian
learning), and 0 (for pessimistic Bayesian learning),
respectively. Thus, in contrast to the benchmark case
of the standard model of Bayesian learning, the above
learning rules exhibit an additional bias that reflects the
agent’s ambiguity attitudes.

D

1027

Divergent Probabilistic Judgments
We are now ready to state our main result which identifies conditions such that Bayesian learning with neoadditive beliefs may result in divergent probabilistic
judgments in the following sense.
Definition (Divergent Probabilistic Judgments). We
say that the difference in the agents’ probabilistic judgments strictly increases (in the number of observations)
iff, for all n,



	



	


 	
~; n1 	jIknþ1  E p
~; n2 	jIknþ1 > E p
~; n1 	jIkn
E p


 	
~; n2 	jIkn
E p
ð9Þ
whereby


 	


 	
~; n2 	jIkn :
~; n1 	jIkn  E p
E p

ð10Þ

For the sake of expositional clarity, we restrict attention
to the case in which differences in initial beliefs of
agents can only be due to their respective optimism
parameters li , i 2 {1, 2}, under ambiguity whereby we
assume that agent 1 is more optimistic than agent 2
ensuring (10).
Proposition 2. Let the neo-additive probability measures n1 and n2 satisfy di = d > 0, mi = m for i 2 {1, 2}
as well as l1 l2.
1. Suppose that both agents use the full Bayesian update
rule. Then the difference in the agents’ probabilistic
judgments strictly increases if and only if l2.
2. Suppose that agent 1 uses the optimistic whereas
agent 2 uses the pessimistic Bayesian update rule.
Then the difference in the agents’ probabilistic judgments strictly increases if and only if l1 l2.
By proposition 2, our stylized model of Bayesian learning formally accommodates two alternative scenarios of
diverging probabilistic judgments. In a first scenario,
divergence arises because of different personal attitudes
toward the resolution of ambiguity. In a second scenario,
divergence corresponds to personal attitudes toward the
interpretation of information. While existing psychological studies provide empirical evidence for the phenomenon of diverging probabilistic judgments, they do not
differentiate between these two alternative explanations
for the phenomenon. It would therefore be interesting to
gather more empirical evidence on updating and learning with nonadditive beliefs.

D

1028

D

Divergent Thinking and Learning

Cross-References
▶ Attitudes – Formation and Change
▶ Bayesian Learning
▶ Belief Formation
▶ Belief-Based Learning Models
▶ Bounded Learning – Rational Learning
▶ Divergent Thinking and Learning

References
Baron, J. (2008). Thinking and deciding. New York/Melbourne/
Madrid: Cambridge University Press.
Chateauneuf, A., Eichberger, J., & Grant, S. (2007). Choice under
uncertainty with the best and worst in mind: Neo-additive
capacities. Journal of Economic Theory, 127, 538–567.
Doob, J. L. (1949). Application of the theory of martingales. In
Colloques Internationaux du Centre National de la Recherche
Scientifique (Ed.), Le Calcul des Probabilite’s et ses Applications
(Vol. 13, pp. 23–27). Paris: Author.
Gilboa, I., & Schmeidler, D. (1993). Updating ambiguous beliefs.
Journal of Economic Theory, 59, 33–49.
Savage, L. J. (1954). The foundations of statistics. New York/London/
Sydney: Wiley.
Schmeidler, D. (1986). Integral representation without additivity.
Proceedings of the American Mathematical Society, 97, 255–261.

Divergent Thinking and
Learning
OLGA M. RAZUMNIKOVA
Cognitive Physiology Lab, Department of Pedagogy
and Psychology, Research Institute of Physiology SB
RAMS, Novosibirsk State Technical University,
Novosibirsk, Russia

Synonyms
Convergent thinking

Definition
Divergent thinking is a thought process used to generate
diverse and numerous ideas on some mental task,
implying that not only one solution may be correct.
The term divergent thinking is used in the sciences of
learning and cognition to designate a psychological
construct that accounts for the specific form of
human thinking. The goal of divergent thinking is to
generate many different ideas about a topic in a short

period of time. It involves breaking a topic down into
its various component parts in order to gain insight
about the various aspects of the topic.

Theoretical Background
The concept of divergent thinking was developed by
psychologist J. P. Guilford, who saw it as a major component of creativity and associated it with four main
characteristics (Guilford 1967):
● Fluency, the ability to rapidly produce a large num-

ber of ideas to problem solution
● Flexibility, the ability to generate multiple problem

solutions from different semantic categories
● Originality, the ability to generate the number of

unique or unusual ideas
● Elaboration, the ability to develop a detail of the

decision with their association at a final embodiment of the general idea
The elements that make up a divergent thinking are
lateral thinking, deducting and inducting, identifying,
synthesizing, differentiating, critical thinking, and
problem solving. The choice of the final decision can
unconsciously occur on the basis of insight. Heuristics
strategy is a relevant method for effective divergent
thinking. It is possible to also use another strategy
when the critical judgment of a set of spontaneously
generating ideas is used. Then along with divergent
thinking the creative problem solving can demand
convergent thinking, intrinsic motivation, persistence,
openness to experience, willingness to take risks, functional nonconformity, and others. Therefore many
researchers consider that divergent thinking only is
not enough for achievement of creative efficiency and
it is not an equivalent of creativity.
However divergent thinking tests are often used,
though of course they really just estimate the potential
for creative thought. Early versions of divergent thinking tasks focused on ideational processes of verbal and
figural content. More recently efforts have been made
to design specific divergent thinking tests relevant to
various domains, including art, science, management,
and engineering or complex and real world issues and
problem-finding skills in children and students (e.g.,
Basadur et al. 2000; Scott et al. 2004). A number of
studies reported more valid divergent thinking scores
when explicit instructions to be original or creative
were given to the examinees and when administration

Divergent Thinking and Learning

of the tests was untimed. There are positive and statistically significant relationships between various divergent thinking test scores and reasonably acceptable
nontest indices of creative behavior and achievement.
Both creativity and divergent thinking, as assessed
through open-ended tests such as consequences,
incomplete figures, and alternative uses, where
responses are scored for fluency (number of responses),
flexibility (category shifts in responses), originality
(uniqueness of response), and elaboration (refinement
of responses), do represent a distinct capacity contributing to many forms of creative performance. Based on
these characteristics of divergent thinking, it can be
discriminated by different paths of learning to this
kind of thinking.
The first level is learning to fluency in thinking – to
train students’ thinking speed so that they may propose
more concepts and more answers in the possible
shortest time.
The second level is cognitive flexibility training to
accommodate themselves to different changes.
The third level is novelty – to train students’ ability
to boldly break away from conventions and to bravely
develop their creative spirit. The capacity to apply ideas
creatively in new contexts, referred to as the ability to
“transfer” knowledge, requires that learners have
opportunities to actively develop their own representations of information to convert it to a usable form.
As in divergent thinking, different cognitive processes are involved, such as information gathering,
comparing, observing, hypothesizing, associating, classifying, generalizing, interpreting, conceptual combination synthesizing ideas, idea evaluation, and
implementation planning, thinking development can
go different ways and include training programs
such as:
● Speed up one’s mental operations and help utilize

all of one’s resources.
● Assist the assimilation of new information, develop

one’s susceptibility to new information, and
encourage one’s conscious curiosity facilitating in
this way the spotting of relationship among distant
areas of interest and operation.
● Eliminate a hindrance to the free functioning of
mental operations and associative mechanisms.
● Extend the area of creative processes for the effective promotion of creative power.

D

With the identification of divergent thinking as
a distinct capacity making a unique contribution to
creative thought, scholars interested in the development of creativity began to apply divergent thinking
tasks in the design of training. Divergent thinking
learning can beginning with discussion on the value
of a task performance strategy, such as heuristic strategy in identifying alternative uses, and then providing
practice in the application of this strategy. Divergent
thinking tasks intended to elicit alternative uses, story
completion, and question generation provided the
basis for training. Novelty of ideas was measured
based on different quality attributes, such as originality,
effectiveness, and implementability.
Many techniques will make the generation of original ideas easier so let us start learning about them now:
random word and picture, false rules, challenge facts,
analogies, wishful thinking, and others. A specific tool
designed to enhance divergent thinking in groups is
brainstorming developed by A. Osborn as a method by
which a group tries to find a solution for a specific
problem by amassing a list of ideas spontaneously
contributed by its members. Osborn’s rules for brainstorming sessions are as follows:
● Judgment of ideas is not allowed (this comes later).
● Outlandish ideas are encouraged (these can be

scaled back later).
● A large quantity of ideas is preferred (quantity leads

to quality).
● Members should build on one another’s ideas
(members should suggest idea improvement).
Brainstorming is used extensively as a technique for
group idea generation in marketing strategy, research
and development procedures, written documents and
articles, engineering components, government policies,
management methods, and company structure and
policy.
The other methodic of divergent thinking training
to aid effective discussions and individual thinking by
systematically promoting different thinking states of
mind was a tool The Six Thinking Hats developed by
E. de Bono. Using different states of mind, de Bono’s
methodic gives a systematic way of considering the
subject from different perspectives and in doing so to
be more complete and effective. Another form of brainstorming development is Gordon’s Sinectics, which
means a connection of different elements of the

1029

D

1030

D

Divergent Thinking and Learning

topic. The divergent thinking in engineering activity
can develop on a basis of G.C. Altshuller’s TRIZ.
Especially when a knowledge domain is complex
and fraught with ill-structured information, activelearning strategies is demonstrably more effective
than traditional linear teaching. Ideation can be training by modeling divergent thinking in students and
clear-cut reinforcement for their successful divergent
thinking.

Important Scientific Research and
Open Questions
Using different neurophysiological measurement
methods, such as EEG, fMRI, and PET, neuroscientific
studies have yielded evidence of possible brain correlates underlying creative and divergent thinking (e.g.,
Fink et al. 2007; Howard-Jones et al. 2005;
Razumnikova et al. 2007). Analysis of the coherence
of an EEG, with its multichannel leads from different
parts of the cortex, makes it possible to clarify the
correspondence of originality in problem solving with
the divergent nature of the interaction between different regions of the cortex. It was found that creative
people are distinguished by their high ability to change
the frequency-spatial organization of cortical activity,
specific forms of which are dependent on sex, intelligence, and other individual characteristics, especially
the person’s emotional sensitivity (e.g., Razumnikova
et al. 2007). Defining the personality profile of an
individual makes it possible to predict his success in
divergent thinking and creative problem solving that
require specific thought strategies in professional work.
The interaction of the left hemispheric frontal with the
parietal part of the attention system, which is more
pronounced in women, may indicate that conscious
control of such activity is more important for women
than for men. Consequently, increasing the significance
attributed to intellectual abilities in modern professional work, as well as the prestige of innovative
forms of behavior, should be seen as a most significant
step toward increasing the creativity of women.
The specific nature of the interaction between the
anterior and posterior cortical regions is just as necessary
a condition for organizing different thinking strategies
as is the interaction of the two hemispheres, and this
interaction is related to the type of problem being
solved and to individual distinctions in the regulation
of functional brain activity (figure). Clearly expressed

Sex differences in:
Arousal
Reticular – thalamic
–cortical interaction
(6-10 Hz rhythms)

Divergent
thinking

ion
n
ers tuitio
n
trav
Ex on - i
ti
nsa
Se
Emotional activation
ism
Neurotic
Cortical - limbic interaction
g
in
nk
hi
(4-6 and 10-30 Hz rhythms)
Feeling -t
Int

ell

ige

nc

e

Strategy of idea search
(information selection)
Cortico-cortical interactions:
Hemispheric asymmetry and
frontal-TPO interaction
(4-30 Hz rhythms)

Divergent Thinking and Learning. Fig. 1 TPO-temporoparietal-occipital

intellectual and character traits are formed as the result
of a specific individual mosaic of both activating and
inhibitory interactions of cortical and subcortical
structures, establishing conditions for future thinking
strategies: divergent/convergent, rational/irrational,
logical/intuitive, and verbal/figural.
A great deal of evidence has also been developed for
the weakening of functional asymmetry, as the reserves of
the right hemisphere are drawn upon for divergent thinking. Right-hemisphere dominance in creative activities is
linked to the fact that its functions include not only visualspatial, but also verbal processes, such as the construction
of metaphors or semantic operations that require a wide
net of associations (Howard-Jones et al. 2005).
The EEG study revealed that the generation of original ideas was associated with alpha synchronization in
frontal brain regions and with a widespread pattern of
alpha synchronization over parietal cortical regions
(Fink et al. 2007). Attention to alpha oscillations
(7–13 Hz) in the studies of EEG correlates of creativity
is based on the fact that the power of the alpha rhythm
is an indicator of cortical activation, a low level of
which, according to C. Martindale, defines the
“defocused” state of attention required for effective creative thinking.
Considering the functional significance of the separate brain structures in divergent thinking, it is worth

Dogmatism

mentioning that it is not a single necessary and sufficient area of the brain that ensures the final result, but
rather a topographically extensive neural network,
which dynamically changes its characteristics
depending on the stage of the creative process or its
character. Ideas about the interaction of the anterior
and posterior cortical areas during a spontaneous or
deliberative search for a creative solution to a problem
seem promising for further study of the patterns of the
neurobiological bases of creative abilities, including
effective divergent thinking. The parietal system, in
this case, represents the “searching” part of the creative
process and provides the generation of multiple ideas
through a variety of associations of visual, auditory,
and symbolic representations. The frontal system performs a “critical/initiating” function and, according to
individual goals and interests, selects ideas received
from the parietal system by developing those that are
acceptable and suppressing those that do not seem
necessary.

Cross-References
▶ Brainstorming and Learning
▶ Creativity and Learning Resources
▶ Creativity, Problem Solving, and Learning
▶ Flexibility in Learning and Problem Solving
▶ Heuristics and Problem Solving
▶ Nature of Creativity

References
Basadur, M., Runco, M. F., & Vega, L. (2000). Understanding how
creative and thinking skills, attitudes and behaviors work
together: A casual process model. The Journal of Creative Behavior, 34, 77–100.
Fink, A., Benedek, M., Grabner, R. H., Staudt, B., & Neubauer, A. C.
(2007). Creativity meets neuroscience: Experimental tasks for the
neuroscientific study of creative thinking. Methods, 42, 68–76.
Guilford, J. P. (1967). The nature of human intelligence. New York:
McGraw-Hill.
Howard-Jones, P. A., Blakemore, S. J., Samuel, E. A., Summers, I. R.,
& Glaxton, G. (2005). Semantic divergence and creative story
generation: An fMRI investigation. Cognitive Brain Research, 25,
240–250.
Razumnikova, O. M., Volf, N. V., & Tarasova, I. V. (2007). Gender
differences in creativity: A psychophysiological study. In
C. Martindale, V. Petrov, & L. Dorfman (Eds.), Aesthetics and
innovation (pp. 445–468). Newcastle: Cambridge Scholars Press.
Scott, G., Leritz, L. E., & Mumford, M. D. (2004). The effectiveness of
creativity training: A quantitative review. Creativity Research
Journal, 16, 361–388.

D

1031

Diversive Exploration
▶ Curiosity and Exploration
▶ Play, Exploration, and Learning

D

Divided Attention
▶ Dual-Task Performance in Motor Learning

Division of Labor
▶ Altruistic Behavior and Cognitive Specialization in
Animal Communities

Dogmatism
ADAM BROWN
School of Education, Elementary Education,
St. Bonaventure University, St. Bonaventure, NY, USA

Synonyms
Belief formation; Belief system; Open- and Closedmindedness

Definition
The word dogmatism comes from the Greek word for
dogma “dόgma,” which means that which seems to one,
opinion, or belief. Dogmas are commonly described in
religion, in which they are the fundamental tenets and
beliefs of that religion. Dogmatism can also relate to how
readily one receives novel information. Those who are
open to new information are considered to be low in
terms of dogmatism and those who are typically more
closed-minded are higher in terms of dogmatism. Dogmatism is defined by Rokeach as “a relatively closed
cognitive organization of beliefs about reality focused
around a central set of beliefs about absolute authority
which, in turn, provides a framework for patterns of

1032

D

Dogmatism and Learning

intolerance and qualified tolerance toward others”
(Rokeach 1954, p. 195). It is further defined as “positiveness in assertion of opinion especially when unwarranted
or arrogant” (Merriam-Webster Dictionary).

Theoretical Background
Milton Rokeach imparts a seminal contribution to the
subject of dogmatism. According to Rokeach (1960),
people’s schemas or cognitive structures are separated
into two well-organized belief systems: the belief system
and disbelief system. A belief system is a working schema
of all the beliefs an individual accepts as truthful.
A disbelief system is constructed of a cluster of subsystems that an individual would deem to be false. Rokeach
describes dogmatism as a rather narrow arrangement of
beliefs and disbeliefs about reality, which provides a
framework for acceptance or rejection of the beliefs of
others. The separate systems of belief and disbelief work
to provide individuals with a means to organize their
interpretations of truth and falsehood. These systems
can also be described as open or closed. “The closed
nature of the belief systems of individuals high in dogmatism can be observed in their tendency to compartmentalize and isolate their beliefs and disbeliefs, whereas
the more open belief systems of individuals low in dogmatism can be observed in their readiness to make
connections between disparate beliefs” (Davies 1998, p.
456). Isolation occurs when two beliefs are not seen as
interrelated and, hence, no communication between
the two subsystems takes place.

Important Scientific Research and
Open Questions
Dogmatism seems to be closely tied to environmental
factors such as parental upbringing (Lesser 1985),
moral reasoning (Nichols and Stults 1985), and personality characteristics such as tender mindedness
(Rokeach and Hanley 1956). However, there also
seem to be several cognitive aspects to dogmatism
(e.g., Rokeach et al. 1955). Throughout the 1950s and
1960s, there was considerable investigation of these
aspects of the construct (e.g., Long and Ziller 1965).
Dogmatism is a topic that has regained interest as of
late. This newfound popularity is due partially to its
interrelatedness to contemporary areas of research and
interest (e.g., conflict resolution, behavior problems in
school, etc.). Dogmatism has implications for research
in the area of cognitive functioning as well.

Cross-References
▶ Dogmatism and Learning

References
Davies, M. F. (1998). Dogmatism and belief formation: Output interference in the processing of supporting and contradictory cognitions. Journal of Personality and Social Psychology, 75, 456–466.
Lesser, H. (1985). The socialization of authoritarianism in children.
The High School Journal, 68(3), 162–166.
Long, B., & Ziller, R. (1965). Dogmatism and predecisional information search. The Journal of Applied Psychology, 49, 376–378.
Nichols, D. P., & Stults, D. M. (1985). Moral reasoning: Defining
issues in open and closed belief systems. The Journal of Social
Psychology, 125, 535–536.
Rokeach, M. (1954). The nature and meaning of dogmatism. Psychological Review, 61, 191–204.
Rokeach, M. (1960). The open and closed mind. New York: Basic
Books.
Rokeach, M., & Hanley, C. (1956). Care and carelessness in psychology. Psychological Bulletin, 53, 183–186.
Further Reading
Dogmatism (2010) In Wikipedia. Retrieved 20 Apr 2010, from http://
en.wikipedia.org/wiki/Dogmatism
Rokeach, M., McGovney, W., & Denny, M. (1955). A distinction
between dogmatic and rigid thinking. Journal of Abnormal and
Social Psychology, 51, 87–93.

Dogmatism and Learning
ADAM BROWN, ANNA PRUDENTE
School of Education, Elementary Education,
St. Bonaventure University, St. Bonaventure, NY, USA

Synonyms
Belief formation; Belief system; Open- and closedmindedness

Definition
Dogmatism is defined by Rokeach (1954) as “a relatively closed cognitive organization of beliefs and disbeliefs about reality, organized around a central set of
beliefs about absolute authority which, in turn, provide
a framework for patterns of intolerance and qualified
tolerance toward others” (p. 195). The more openminded an individual is, the lower that individual is
said to be in terms of dogmatism. Conversely, more
closed-minded individuals are higher in dogmatism.

Dogmatism and Learning

Learning is “the process by which changes in behavior arise as a result of experience interacting with
the world” (Gluck et al. 2008, p. 2). It is “moving
information from working memory – where we actively
think about it – to long-term memory, where information is stored indefinitely” (Martinez 2010, p. 62).
We do not learn everything that we think about; rather,
we learn only a small portion of the great bulk of
information that is presented to us. To learn information, we must first perceive incoming information,
consciously attend to that information, and move
it from temporary working memory to enduring
long-term memory.

Theoretical Background
According to Rokeach (1960), people’s schemas or cognitive structures are separated into two well-organized
belief systems: the belief system and disbelief system.
A belief system is a working schema of all the beliefs an
individual accepts as truthful. A disbelief system is
constructed of a cluster of subsystems that an individual would deem to be false. The degree of dogmatism
one exhibits determines their belief systems and subsequently determines one’s style and capability of learning. Highly dogmatic individuals tend to isolate their
beliefs and disbeliefs with little room for integrating
new information, while individuals with low dogmatism are more open to the possibilities and potential of
new information.
Working memory is the system for temporary
maintenance and manipulation of information. One’s
working memory allows for the playing out of complex
cognitive actions, such as comprehension, reasoning,
and learning (Baddeley 1994). The capacity of working
memory differs from individual to individual, with
each of us having a restricted amount of attentional
resources. This availability of working memory space is
unique for each of us (Tirre and Pena 1992).
It is with consideration to working memory capacity that the relationship between dogmatism and learning comes into play. Let us consider the phenomenon
of belief formation, which is, in essence, synonymous
with learning. It is contended that general rules exist
which impact cognitive processing as it is related to
dogmatism and belief formation. When an individual
is presented with new information, that information is
utilized to form a provisional belief. Whether or not
that provisional belief is substantiated is determined by

D

the incorporation of supporting and/or contradictory
information. Highly dogmatic individuals tend to seek
out supporting information at the expense of contradictory information; this is in an effort to strengthen
their tentative belief, granting it the support in their
eyes necessary to strengthen and/or confirm it (Davies
1998). When one has limited availability for the attentional resources necessary to integrate any new contradictory information, that is, limited working memory
space for considering alternative arguments, the individual tends to be more dogmatic. The more evidence
one is presented with, the more confident that individual is in his or her belief (Kelly 2008).
Strength of argument impacts belief formation
and subsequently, learning. Individuals low in dogmatism tend to be influenced by an argument regardless of
the expertise displayed by the individual imparting
the information. The low-dogmatic individuals are
persuaded by the strength of the argument, not by
the individual providing that information. These individuals are able to reason through the information
and determine the value of it with minimal consideration granted to the source. Conversely, highly
dogmatic individuals are inclined to believe an
argument based on source expertise alone. That is, if
the source appears to be a legitimate informant of the
argument at hand, highly dogmatic individuals
deem the validity of the argument therefore strengthened, regardless of the actual information imparted
(DeBono and Klein 1993).

Important Scientific Research and
Open Questions
Dogmatism is a topic that has regained interest as of
late. This newfound popularity is due partially to its
interrelatedness to contemporary areas of research and
interest (e.g., conflict resolution, behavior problems in
school, cognitive functioning, etc.).
Many advocate the necessity of developing critical
thinking skills, and the reasons for this promotion are
extensive. Dogmatism is not typically incorporated
into the idea of critical thinking, yet its relationship to
learning warrants consideration. If, as a group, we
simply accept what is told to us as truth, without
considering the possibility that the information is
potentially fallacy, we run the risk of being both
uninformed and easily manipulated (White 2006). In
doing so, we perceive all new incoming information as

1033

D

1034

D

Domain of Movement

fact without any justification for this belief other than
the simple reality that it was told to us. Thus, belief
could in essence turn into a mechanism of power and
influence if individuals do not critically evaluate for
themselves. If we are able to critically think about
a situation or proposition, considering both sides,
however, and replace fallacies with progressive truths,
we are able to enhance our own learning as well as the
social good (Harrigan 2008).
Tolerance is advanced when one is willing to consider new information as possible truth (Leone 1989).
If we are willing to remove ourselves from our own
inflexible belief systems and consider the worth in
assessing alternative views, we are exhibiting tolerance
and opening ourselves up to the potential benefits of
accepting new information. Essentially, if we are willing
to open our minds and be less dogmatic, we become
more accepting of the notion that other actions and
beliefs are perhaps justifiable.
It is arguable that those who are very willing to
readily accept new information as truths open the
door to manipulation and may become the clay with
which others can easily mold. On the other hand,
however, it is erroneous to doubt simply for doubt’s
sake alone. Doubt must be grounded in logic, facts, and
evidence, and simply opting to disbelieve for the sake of
disbelieving is in itself a form of belief – it becomes
one’s mantra – and makes one more
dogmatic. Therefore, it is debatable that a balance
must be achieved in which the dichotomy of acceptance
versus rejection of arguments is weighed.

Cross-References
▶ Belief Formation
▶ Dogmatism and Belief Formation
▶ Learning About Learning
▶ Learning and Thinking

DeBono, K. G., & Klein, C. (1993). Source expertise and persuasion:
The moderating role of recipient dogmatism. Personality and
Social Psychology Bulletin, 19(2), 167–173.
Gluck, M. A., Mercado, E., & Myers, C. E. (2008). Learning and
memory: From brain to behavior. New York: Worth.
Harrigan, C. (2008). Against dogmatism: A continued defense of
switch side debate. Contemporary Argumentation and Debate,
29, 37–66.
Kelly, T. (2008). Disagreement, dogmatism, and belief polarization.
The Journal of Philosophy, 105(10), 243–252.
Leone, C. (1989). Self-generated attitude: Some effects of thought and
dogmatism on attitude polarization. Personality and Individual
Differences, 10, 243–252.
Martinez, M. E. (2010). Learning and cognition: The design of the
mind. New Jersey: Merrill.
Rokeach, M. (1954). The nature and meaning of dogmatism. Psychological Review, 61, 194–204.
Rokeach, M. (1960). The open and closed mind. New York: Basic
Books.
Tirre, W., & Pena, C. (1992). Investigation of functional working
memory in the reading span test. Journal of Educational Psychology, 84, 462–472.
White, R. (2006). Problems for dogmatism. Philosophical Studies, 131,
525–557.

Domain of Movement
▶ Impaired
Learning

Multidimensional

Motor

Sequence

Donation
▶ Cognitive Aspects
Nonhuman Primates

of

Prosocial

Behavior

in

References
Baddeley, A. (1994). Working memory: The interface between
memory and cognition. In D. L. Schacter & E. Tulving (Eds.),
Memory systems 1994 (pp. 351–367). Cambridge, MA: MIT
Press.
Davies, M. F. (1998). Dogmatism and belief formation: Output interference in the processing of supporting and contradictory cognitions. Journal of Personality and Social Psychology, 75(2),
456–466.

Double-Blind Experiment
Is an experimental procedure in which neither the subjects of the experiment nor the persons administering it
know the critical aspects of the experiment; a doubleblind procedure is used to guard against experimenter
bias, the “Clever Hans effect,” and placebo effects.

Double-Loop Learning

Double-Loop Learning
MOHAMED AMINE CHATTI1, MATTHIAS JARKE1,
ULRIK SCHROEDER2
1
Informatik 5, RWTH Aachen University,
Aachen, Germany
2
Lehr- und Forschungsgebiet Informatik 9,
RWTH Aachen University, Aachen, Germany

Definition
Double-loop learning is a core concept in organizational learning. It is defined in (Argyris and Schön
1996, p. 21) as “learning that results in a change in
the values of theory-in-use, as well as in its strategies
and assumptions.”

Theoretical Background
The concept of double-loop learning was introduced by
Argyris and Schön (1978) within an organizational
learning context. A starting point for Argyris and
Schön’s double-loop learning is the notion of theory
of action. A theory of action includes “strategies of
actions, the values that govern the choice of strategies
and the assumptions in which they are based” (Argyris
and Schön 1996, p. 13). The authors make a distinction
between two types of theories of action: espoused theory
and theory-in-use. They write:
"

Theory of action, whether it applies to organizations or
individuals, may take two different forms. By “espoused
theory” we mean the theory of action which is
advanced to explain or justify a given pattern of activity. By “theory-in-use” we mean the theory of action
which is implicit in the performance of that pattern of
activity. (p. 13)

An organization/individual theory-in-use includes
norms for corporate/individual performance, strategies
for achieving values, and assumptions that bind strategies and values together.
Argyris and Schön (1978, 1996) draw upon the
notion of theory-in-use to present their view of organizational learning. According to the authors, organizational learning is the process of detecting and
correcting errors. It takes account of the interplay
between the actions and interactions of individuals
and higher-level organizational entities such as

D

departments, divisions, or groups of managers. Each
member of an organization constructs his or her own
representation of the theory-in-use of the whole. Organizational learning then occurs when individuals
within an organization experience a problem (error
detection) and work on solving this problem (error
correction). Error correction happens through
a continuous process of organizational inquiry, where
everyone in the organizational environment can
inquire, test, compare, and adjust his or her theoryin-use, which is a private image of the organizational
theory-in-use. Effective organizational inquiry then
leads to a reframing of one’s theory-in-use, thereby
changing the organizational theory-in-use.
Argyris (1991) asserts that most people define
learning too narrowly as mere “problem solving,” so
they focus on identifying and correcting errors in the
external environment. This is what Argyris calls singleloop learning. But, in the words of Argyris:
"

if learning is to persist, managers and employees must
also look inward. They need to reflect critically on their
own behavior, identify the ways they often inadvertently contribute to the organization’s problems, and
then change how they act. (p. 99)

This deeper form of learning is what Argyris terms
double-loop learning. Argyris further claims that highly
skilled professionals are frequently very good at singleloop learning:
"

After all, they have spent much of their lives acquiring
academic credentials, mastering one or a number of
intellectual disciplines, and applying those disciplines
to solve real-world problems. But ironically, this very
fact helps explain why professionals are often so bad at
double-loop learning. Put simply, because many professionals are almost always successful at what they do,
they rarely experience failure. And because they have
rarely failed, they have never learned how to learn from
failure. So whenever their single-loop learning strategies go wrong, they become defensive, screen out
criticism, and put the “blame” on anyone and everyone
but themselves. In short, their ability to learn shuts
down precisely at the moment they need it the most.
(p. 99)

To put it simply, single-loop learning differs from
double-loop learning in that the former aims at efficiency (i.e., doing the things right) and the latter

1035

D

1036

D

Double-Loop Learning

focuses on effectiveness (i.e., doing the right things).
Argyris and Schön (1996) define single-loop learning
as “learning that changes strategies of actions or
assumptions underlying strategies in ways that leave
the values of a theory of action unchanged” (p. 20),
and double-loop learning as “learning that results in
a change in the values of theory-in-use, as well as in its
strategies and assumptions” (p. 21).
As depicted in Fig. 1, single-loop learning refers to
a single feedback loop that connects detected error to
organizational strategies of action and their underlying
assumptions. These strategies and assumptions are
modified, but the norms and values of the theory-inuse remain unchanged. In double-loop learning, by
contrast, correction of error requires inquiry through
which organizational norms and values themselves are
modified. Double-loop learning refers to the two feedback loops that connect detected error not only to
strategies and assumptions but also to norms and
values of the theory-in-use. Strategies and assumptions
may then change concurrently with, or as
a consequence of, change in values of the theory-inuse (Argyris and Schön 1996).
In other words, Argyris and Schön differentiate
between learning that does not change the underlying
mental models of the learners but merely revises their
application scenarios (single-loop), and learning which
does affect such changes (double-loop). Double-loop
learning starts from a learner’s mental model (i.e.,
theory-in-use) defined by base norms, values, strategies, and assumptions, and suggests critical reflection
in order to challenge, invalidate, or confirm the used

theory-of-use. Double-loop learning also encourages
genuine inquiry into and testing of one’s actions and
requires self-criticism, that is, the capacity for
questioning one’s theory-in-use and openness to
change the same as a function of learning. The result
of reflection, inquiry, testing, and self-criticism would
then be a reframing of one’s norms and values, and
a restructuring of one’s strategies and assumptions,
according to the new settings.
To explain the distinction between single- and double-loop learning, Argyris and Schön (1996) give the
example of the behavior of a heating or cooling system
governed by a thermostat:
"

In an analogy to single-loop learning, the system
changes the values of certain variables (for example,
the opening or closing of an air valve) in order to keep
temperature within the limits of s setting. Double-loop
learning is analogous to the process by which
a change in the setting induces the system to maintain
temperature within the range specified by a new
setting. (p. 21)

Argyris and Schön (1996) stress that double-loop
learning is essential for productive organizational
learning within rapidly changing and uncertain settings. As they put it “long-term effectiveness depends
on the possibility of double-loop learning” (p. 96) and
“values that govern double-loop organizational inquiry
are foundational to sustained productive organizational learning” (p. 246).
Argyris and Schön (1996) further distinguish
between Model I and Model II theories-in-use as two

Theories-in-use

Norms/
values

Action
strategies/
assumptions

Consequences
/outcomes of
action

Single-loop learning

Double-loop learning

Double-Loop Learning. Fig. 1 Single- and double-loop learning

Error
detection/
correction

Dreaming: Memory Consolidation and Learning

models of theories-in-use that either inhibit or
enhance double-loop learning. The authors point
out that when individuals deal with issues that are
embarrassing or threatening, their reasoning and
action conform to Model I theories-in-use. Theoriesin-use consistent with Model I are often triggered by
defensiveness, competitiveness, and mutual mistrust
and are shaped by an implicit willingness to protect
oneself, avoid embarrassment, maintain control, maximize winning, and minimize losing. Such Model
I theories-in-use often lead to defensive reactions,
which, in turn, reduce the likelihood that individuals
will engage in the kind of organizational inquiry that
leads to productive learning outcomes. Model
I theories-in-use are thus said to inhibit double-loop
learning.
Model II theories-in-use, by contrast, value processes that foster double-loop learning, such as sharing
control, participation in design and implementation of
action, and encouraging frequent organizational
inquiry as well as public testing of actions and
their underlying assumptions. The consequence of
using Model II theories-in-use should then be “an
enhancement of the conditions for double-loop learning in organizational inquiry where assumptions and
norms central to organizational theory-in-use can be
surfaced, publicly confronted, tested, and restructured”
(Argyris and Schön 1996, p. 119).
Model I and Model II theories-in-use build the base
for a distinction between O-I and O-II organization
learning systems. According to Argyris and Schön
(1996), Model I theories-in-use lead to an organization
with an O-I learning system, which is “dominated by
organizational defensive routines” (p. 106) and “highly
unlikely to learn to alter its governing variables, norms,
and assumptions, because this would require organizational inquiry into double-loop issues, and O-I systems
militate against such inquiry” (p. 111). Model II theories-in-use, on the other hand, help organizations
move toward O-II learning systems with a capacity
for productive organizational inquiry and doubleloop learning.

Cross-References
▶ Knowledge Management
▶ Learning Cycle
▶ Organizational Learning

D

1037

References
Argyris, C. (1991). Teaching smart people how to learn. Harward
Business Review, 69(3), 99–110.
Argyris, C., & Schön, D. A. (1978). Organizational learning, a theory of
action perspective. Reading: Addison-Wesley.
Argyris, C., & Schön, D. A. (1996). Organizational learning II: Theory,
method and practice. Reading: Addison-Wesley.

Dreaming
▶ Dreaming: Memory Consolidation and Learning

Dreaming: Memory
Consolidation and Learning
QI ZHANG
Sensor System Inc, Madison, WI, USA

Synonyms
Dreaming; Learning; Memory consolidation

Definition
Dreaming refers to the subjective conscious experience
we have during sleep. The experience is often
vivid, intense, bizarre, and emotional. Memory
consolidation is considered a neural process by
which hippocampus-dependant memory becomes
independent and is consolidated into the general neocortex over a long period of time from weeks to years.
Learning, in here, refers to the acquisition process
of long-term memory for semantic knowledge and
implicit knowledge. Numerous psychological and
functional studies have concluded that the quality of
learning depends on the quality of dream sleep.
The discovery of hippocampal reactivation supports
the theory of memory consolidation, that is, daily
experience is reactivated and consolidated during
dream sleep. The reactivated past experience is also
believed to be the source from which the neocortex
learns knowledge.

D

1038

D

Dreaming: Memory Consolidation and Learning

Theoretical Background
Mankind is not the sole species that is able to dream.
Almost all mammals have rapid eye movement (REM)
during sleep that is considered the signature of dreaming. Why do we dream? Many functions of dreaming
have long been proposed to answer this question. The
functions are initially based on psychophysiological
studies that traditionally study dream reports (dream
recalls) and impacts of dream deprivation. The proposed functions include learning, memory consolidation, problem solving, creativity, psychological balance
and growth, and disguised attempt of wish fulfillment,
etc. For example, unprepared learning, which is difficult to master, is especially dependent of the quality of
REM sleep; skill learning that requires significant concentration is followed by increased REM sleep; semantic knowledge learned in daytime is recalled better the
next morning after a night of good sleep.
However, some researchers conclude that dream may
not have any cognitive function because dreams are random thoughts or random impulses. For example, when
studying dream reports, it is often found that daily experiences are replayed in the form of segments rather than
entire episodes, which corresponds to the bizarreness of
dreaming. The randomness of dreaming is firstly reported
by Hobson and McCarley (1977), who conclude that
dreams are caused by random signals arising from the
pontine brainstem during REM sleep, and the forebrain then synthesizes the dream and tries its best to
make sense out of the nonsense it is presented with.
Although later studies suggest that the brainstem is not
the key source of signal that causes dreaming and REM
is not the only sleep state that generates dream report,
the random activation is widely agreed as the common
characteristic of dreaming, and becomes a required mechanism of any proposed cognitive function of dreaming.
Neuropsychological study, equipped with single neuron recording and neuroimaging technologies, has
revealed detailed neuronal and cortical activities that
reinforce the view that dreaming may play an important
part in memory consolidation. Firing of the neural assemblies within the ▶ hippocampus, which was activated in
daily activities, has been recorded and reported by
Pavlides and Winson (1989) and many other researchers.
The firings were recorded in humans and rats during both
REM and non-REM (NREM) sleep. Synchronized activity in the hippocampus and neocortex in sleep is also
reported. In many of these studies, the hippocampal firing

is explicitly attributed to the reactivation of episodic
memory and considered the source of dreaming.
Since dreaming is considered a cognitive process that
can result in memory consolidation and learning, the
understanding of these functions becomes an understanding of human memory system, its organization
and learning process. Human long-term memory can
be fractionated into explicit (declarative) memory and
implicit (nondeclarative) memory. Implicit memory
encompasses priming, perceptual learning, and procedural skills, etc. Explicit memory can be further divided
into ▶ episodic memory and ▶ semantic memory
(Tulving 1972). Episodic memory refers to memory
for events (past experiences); one must retrieve the
time and place of occurrence in order to retrieve the
event. Semantic memory refers to relatively permanent
and generic knowledge of the world, or factual knowledge. The division has its neurobiological basis: episodic
memory is correspondent to the hippocampal complex
(including the hippocampus and its surrounding areas),
and semantic memory to the general neocortex.
The hippocampus is extremely important to one’s
explicit memory. When a person’s hippocampus is
entirely damaged, the person will never be able to recall
any experience that occurred after the onset of the damage, and become almost impossible to learn new semantic
knowledge. However, some of past experience that
occurred before the onset may still be recalled, which is
attributed to memory consolidation. There are two kinds
of views about memory consolidation. In one view (e.g.,
Squire and Alvarez 1995), episodic memory is initially
stored in the neocortex as discrete pieces of information,
and the associations (traces) among the pieces are stored
in the hippocampus. During consolidation, the discrete
information slowly becomes associated under the influence of the activated traces. When the association is fully
established, episodic memory becomes independent of
the hippocampus. In the other view (e.g., McClelland
et al. 1995; Zhang 2009), episodic memory is initially
stored in hippocampus. Memory consolidation is
a training process in which the hippocampus slowly
teaches the hippocampal representations (episodic
memory) into the neocortex. In such a slow process
of consolidation, episodic memory has to be revived
constantly and repeatedly. Squire and Alvarez (1995)
suggest that dreaming may be the best answer to
explain why the consolidation process does not regularly intrude into our waking consciousness (Fig. 1).

Dreaming: Memory Consolidation and Learning

Dreaming: Memory Consolidation and Learning. Fig. 1
The neocortex and hippocampus. Both the neocortex and
hippocampus are split into left and right hemispheres

Human semantic memory is considered meaningbased and is abstracted factual knowledge acquired
from past experience. The acquisition process involves
meaning perception, abstraction, generalization, and
organization. The mechanisms of the subprocesses are
fundamental and open questions of cognitive science.
Since episodic memory is the prerequisite of semantic
learning, semantic knowledge may only be acquired
from the reactivations of episodic memory associated
with or stored in the hippocampus. Thus, dreaming (as
well as, memory consolidation) is also considered the
process of semantic learning (e.g., Zhang 2009). During
semantic learning, reactivated information repeatedly
accesses the neocortex and slowly causes long-lasting
synaptic modification within associated neural assemblies, which is believed to be necessary and sufficient
for the acquisition of long-term memory in the general
view of neural learning mechanism.

Important Scientific Research and
Open Questions
As we go to sleep, we slowly sink down into NREM sleep.
After an hour or two, the first REM period begins and
lasts a few minutes. Then, we sink back into NREM sleep
again. This cycle occurs about every 90 min. Toward the
end of the night, the REM periods get longer. Almost all
mammals have the NREM–REM cyclic alternation in
sleep, which suggests not only a shared mechanism

D

across species but a universal functional significance.
Dreams are reported in 70–95% of REM awakenings,
and in 5–10% of NREM awakenings. Thought-like
recalls are reported in 43–50% of NREM awakenings.
The proposed functions of dreaming require an activated brain during dream sleep, which is proved to be the
case by functional imaging studies. In REM sleep, the
brain is almost as active as awaking, except for the
primary sensory areas and motor output areas. Specifically, activity in the hippocampus, entorhinal cortex,
and other parahippocampal regions is increased during
REM sleep relative to both waking and NREM sleep. The
following areas are relatively less active in NREM than in
REM sleep: brainstem, midbrain, anterior hypothalamus, hippocampus, caudate, medial prefrontal, caudal
orbital, anterior cingulate, parahippocampal, and inferior temporal cortices.
The discovery of the reactivation of recent waking
patterns of neuronal activity within the hippocampus
(i.e., hippocampal firing) in dream sleep plays an
extremely important role to reinforce the proposals of
dream’s cognitive functions and to identify the source of
dreaming. In such studies, the subjects (e.g., rats) are
implanted with microelectrode arrays to record multiple
single-cell activity in the hippocampus. The hippocampal
firing patterns are recorded during specific training tasks,
performances of the tasks, and in REM and NREM sleep.
The distinctive firing pattern corresponding to the specific
task is replayed and recorded in REM and NREM sleep.
Memory consolidation and knowledge learning in
dreaming has a causal relationship with the dissociation between episodic memory and semantic memory.
The most studied amnesic patient, H.M., has a special
position in the understanding of the dissociation. After
H.M.’s hippocampus was surgically removed due to
intractable epilepsy, he became amnesic and “every
day is a living hell” as he described. He could not recall
any experience that occurred after the onset, but he
could recall certain experience that occurred before
the onset (the earlier the experience from the onset,
the better the recall), and he hardly learned any semantic knowledge thereafter.
Several studies have simulated the memory consolidation process using connectionist networks (e.g.,
McClelland et al. 1995; Squire and Alvarez 1995) and
cognitive network (Zhang 2009). In the simulations,
the episodic memory is randomly activated that coincides with the key characteristic of dreaming. The only

1039

D

1040

D

Drill and Practice in Learning (and Beyond)

simulation of semantic learning in dreaming is
reported by Zhang (2009), in which the source of
dream learning is randomly activated segments of
past experience in a hippocampal-like storage. Importantly, the learned conceptual knowledge is demonstrated to be flexible. Flexibility is the key criterion of
semantic knowledge/memory.
Researchers now have less doubt about the memory
consolidation function of dreaming after the discovery
of hippocampal replay of waking activation patterns.
Semantic learning from randomly activated experience
is computationally demonstrated possible, but the
direct neurobiological evidence is yet to be discovered.
The capacity of dreaming is a gift from billion years of
evolution, and its impact likely affects every aspect of
human cognitive capacity. On the other hand, our
knowledge about human brain, among most complicated systems in the universe, is very limited; there
should be a long way before we can entirely understand
dreaming, its functions, and corresponding neuropsychological mechanisms.

Cross-References
▶ Amnesia and Learning
▶ Basal Ganglia Learning
▶ Computational Models of Human Learning
▶ Concept Learning
▶ Human Cognitive Architecture
▶ Linking Fear Learning to Memory Consolidation
▶ Memory Consolidation and Reconsolitation
▶ Neuropsychology of Learning
▶ Procedural Learning
▶ Song Learning and Sleep

References
Hobson, J. A., & McCarley, R. W. (1977). The brain as a dream-state
generator: An activation-synthesis hypothesis of the dream process. The American Journal of Psychiatry, 134, 1335–1348.
McClelland, J. L., McNaughton, B. L., & O’Reilly, R. C. (1995). Why
there are complementary learning systems in the hippocampus
and neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychological Review,
102, 419–457.
Pavlides, C., & Winson, J. (1989). Influences of hippocampal place
cell firing in the awake state on the activity of these cells during
subsequent sleep episodes. The Journal of Neuroscience, 9,
2907–2918.
Squire, L. R., & Alvarez, P. (1995). Retrograde amnesia and memory
consolidation: a neurobiological perspective. Current Opinion in
Neurobiology, 5, 169–177.

Tulving, E. (1972). Episodic and semantic memory. In E. Tulving &
W. Donaldson (Eds.), Organization of memory (pp. 381–403).
New York: Academic.
Zhang, Q. (2009). A computational account of dreaming: Learning
and memory consolidation. Cognitive Systems Research, 10,
91–101.

Drill and Practice in Learning
(and Beyond)
CHAP SAM LIM1, KEOW NGANG TANG1, LIEW KEE KOR2
1
School of Educational Studies, Universiti Sains
Malaysia, Minden, Penang, Malaysia
2
Department of Mathematical Sciences, Universiti
Teknologi MARA Malaysia, Merbok, Kedah
Darulaman, Malaysia

Synonyms
Manipulative practice; Repetitive learning; Rote learning; Routine exercise; Systematic repetition

Definition
The term drill and practice is defined as a method of
instruction characterized by systematic repetition of
concepts, examples, and practice problems. Drill and
practice is a disciplined and repetitious exercise, used as
a mean of teaching and perfecting a skill or procedure. As
an instructional strategy, it promotes the acquisition of
knowledge or skill through systematic training by multiple repetitions, rehearse, practice, and engages in
a rehearsal in order to learn or become proficient.
Similar to memorization, drill and practice involves
repetition of specific skills, such as spelling or multiplication. To develop or maintain one’s specific skills,
the subskills built through drill and practice should
become the building blocks for more meaningful
learning.

Theoretical Background
The history of drill and practice can be dated back to the
1940s when educational psychologists introduced
instructional tools such as programmed instruction
that employed behaviorism learning theory. Edward L.
Thorndike (1874–1949), John B. Watson (1878–1958),
and B. F. Skinner (1904–1990) were among several

Drill and Practice in Learning (and Beyond)

distinguished advocates of behaviorist theory. Behaviorism regards learning as a measurement of observable
changes in a learner’s behavior. In order to bring about
the behavioral changes, the learner practices repetition
of a new behavioral pattern until it becomes automatic
and habitual to the learner. A popular example is the
“Skinner’s Box” that used operant conditioning to train
rats to remember the way out of a maze. Operant conditioning rewards an act that approaches a new desired
behavior through the use of positive and negative
reinforcement.
Drill and practice is rooted in the theory of
behaviorism. It focuses on the repetition of stimulus–
response practice that leads to strengthening of habits
and consequently facilitates mastery of content learning. Similar to behaviorist’s ideology, drill and practice
operates on the concept of reinforcement. It trains
learners repeatedly to master new information or perform a new act through repetitive exercise and rewarding them with immediate feedback.
Relating to the mechanism of knowledge acquirement, cognitive psychologists believe that the fundamental units of knowledge or skill can often be broken
down into smaller units or subskills. Drill and practice
emphasizes on repetition, remedial action, and feedback to reinforce the mastering of each subskill as well
as to automate certain performance. Through drill and
practice, learners are trained repeatedly to achieve
automated performance in the lower-level subskills
before they proceed to the higher level of complex
skill. A good example is the training method employed
by the physical fitness trainer or the music teacher. In
their training, they use systematic and repetitive practice to train learners from basic to advance level until
they can effortlessly execute the more complicated
movements without much energy and hard thinking.
Merrill and Salisbury (1984) address the attainment of
this kind of level as automaticity whereby the ability to
do things requires less and less attention and it happens
naturally without occupying the mind with other
ongoing cognitive processes. In other words,
automaticity is a response that has become automatic
or habitual to a learner.
On the relationship of brain and learning, cognitive
scientists who used functional magnetic resonance
imaging (fMRI) or MRI scan found that there is an
actual shift in brain activation patterns when untrained
facts are learned. This observation can be justified by

D

the fact that when a learner can retrieve automatically
intermediate steps (such as an algorithm stored in the
working memory), then he/she can probably solve
a complex computation quicker without much cognitive interferences (Delazer et al. 2004). In other words,
automatic retrieval of facts cuts down the shift in brain
activation patterns and helps to reduce the load in the
working memory (i.e., cognitive load).
Furthermore, according to neuroscientists,
repeated experience is necessary in forming connections or synapses between the brain cells. Therefore,
drill and practice is important because the brain will
retain the learning experience longer when these connections are strengthened and reinforced through repetitions. For instance, most mathematics teachers
would encourage their students to memorize the multiplication table through drill and practice to improve
retention. Consequently, students can easily recall the
multiplication facts since it has become automated.
Drill and practice is a teaching technique conceivably practiced most frequently by teachers as compared
to many other pedagogic methods. It is commonly used
by school teachers to drill students on the past year
examination questions in pursuit of excellent
academic examination results. This teaching technique
is also effective especially in helping less-able students
to score in a fact-based examination.
An effective drill and practice training takes on the
following measures:
● Organize and structure the lessons/activities so as to

●
●
●
●

put emphasis on reinforcement of previously
learned concepts.
Differentiate the teaching materials according to
levels of difficulties.
Provide immediate feedback or answers to each
problem solved.
Set up a management system to keep track of
learners’ progress.
Allow learners to master the learning materials at
their own pace.

In addition, effective drill and practice training
needs to recognize the type of skill to be developed
and the appropriate strategies to develop these competencies. The training may focus on a specific subject
area (such as reading or mathematics); or a part of the
subject area (e.g., spelling or addition). Alternatively, it
may expand to improve different skills involving

1041

D

1042

D

Drill and Practice in Learning (and Beyond)

several areas of the curriculum, for example, music,
physical exercises, swimming, etc.
On the whole, drill and practice is used by teachers
as a reinforcement tool when they perceive that their
students need more practice to reinforce basic skills.
Drill and practice is commonly conducted in a variety
of methods that includes paper-and-pencil worksheets,
tutorials, as well as teacher’s questioning technique.
It is also most commonly employed by mathematics
and language teachers.

Important Scientific Research and
Open Questions
With the recent advancement in computer technology,
paper-based drill and practice has given way to
computer-based system of drill and practice. Computer-based system of drill and practice is gaining its
popularity due to having several features over the conventional system. These features include its ability to
respond to individual differences as well as to provide
students with immediate corrective and instructional
feedback. In addition, with the personalized and interactive nature of most software, this type of practice
allows programmed and extended practices.
Computer-based self-instruction uses drill and
practice to teach and rewards learners with each correct
response. It operates on structured reinforcement to
increase learner acquisition of basic skills. Drill and
practice software packages normally offer questionand-answer activities supplemented with appropriate
feedback. In this type of packages, learners are allowed
to select an appropriate level of difficulty for which
specific content materials are set to start. The software
packages may also provide drill through games. The
inclusions of a gaming scenario, as well as colorful and
animated graphics are all set to motivate learners to
respond to the questions quickly and accurately. Thus,
effective drill and practice software usually provides
immediate feedback, corrections of errors, explanation
of how to get the correct answers, as well as a tracking
management system on student progress.
Nevertheless, to what extent drill and practice of
basic skills can contribute to the achievement of literacy
and higher-order thinking skills? Does it apply to all
fields or should it be differentiated between academic
and physical training?
Drill and practice has been one of the key features in
school mathematics. Li (2006) argued that while

mathematics educators in the West regard drill and
practice as rote learning and merely imitatively behavioral manipulation, China and East Asian countries
consider routine or manipulative practice as an important learning style. In fact, Li (2006) linked drill and
practice to the cultural belief of “practice make perfect.”
He cited a statement in the Confucius’ Analects (论语)
that 学而习之,不于悦乎 which can be translated as
“it is a pleasure to learn and practice again and again.”
Hence, many teachers and students in China believe
that “through imitation and practice again and again,
people will become highly skilled” (Li 2006, p.130).
Furthermore, practice makes perfect which is 熟能生
巧 and can be interpreted at two levels. At the first level,
the students are merely doing repetitive manipulation
or routine exercise. At the second level, students will be
given the similar problems or exercises in various
forms. Hence, the meaning of the word, 熟 is beyond
“practice” or “do.” It means both familiarity as well as
proficiency.
Practice is thus vital to become proficient or to be
perfect in any skills, such as musical, physical or cognitive skills. Yet, there remains controversial debate
whether it is worthwhile to spend the time and effort
in drill and practice, as opposed to more conceptual,
explorative, innovative, and creative activities. Repetitive and extensive practice may promote automaticity
that makes learning effortless and less cognitively
demanding. However, will the automaticity reduce or
prevent individual’s creativity?

Cross-References
▶ Behaviorism and Behaviorist Learning Theories
▶ Feedback in Instructional Contexts
▶ Repeated Learning and Cultural Evolution
▶ Repetition and Imitation: Opportunities for Learning
▶ Skinner B.F. (1904–1990)

References
Delazer, M., Domahs, F., Bartha, L., Brenneis, C., Locky, A., & Trieb,
T. (2004). The acquisition of arithmetic knowledge – an fMRI
study. Cortex, 40, 166–167.
Li, Shiqi. (2006). Practice makes perfect: A key belief in China. In
F. K. S. Leung, K.-D. Graf, & F. J. Lopez-Real (Eds.), Mathematics
education in different cultural traditions – A comparative study of
East Asia and the West (The 13th ICMI Study, Vol. 9,
pp. 129–138).
Merrill, P. F., & Salisbury, D. (1984). Research on drill and practice
strategies. Journal of Computer Based Instruction, 11(1), 19–21.

Drug Conditioning

Drive
▶ Academic Motivation
▶ Motivation and Learning: Modern Theories

D

other USs. These interoceptive drug stimuli may also
disambiguate when some other CS will or will not be
paired with a US, thus having a modulatory or occasion
setting role.

Theoretical Background

Drive-Reduction Theory
The idea that responses are reinforced when they
reduce a drive, where drives are aversive states such as
hunger or thirst that reflect an unmet need of the
organism.

Dropout
▶ Learning Retention
▶ Research of Learning Support and Retention

Drug Conditioning
RICK A. BEVINS1, THOMAS J. GOULD2
1
Department of Psychology, University of
Nebraska-Lincoln Burnett, 19A, Lincoln, NE, USA
2
Temple University, Philadelphia, PA, USA

Synonyms
Conditioned sensitization; Conditioned tolerance;
Pavlovian drug conditioning

Definition
Researchers in the field of drug conditioning investigate
Pavlovian (classical) conditioning processes involving
ingested or administered ligands (drugs). These ligands
may be studied as reinforcers or unconditioned stimulus (US) paired with conditional stimuli (CSs) such as
a discrete environmental cue, situation, or context.
Alternatively, the stimulus effects of drugs may be
studied as interoceptive contexts or cues paired with

1043

Ingested substances like cigarette smoke, coffee,
evening cocktail, pain relievers, etc., contain molecules
that can broadly affect nervous system processes that
translate into measurable changes in physiology
and behavior. If repeatedly ingested, some of these
effects may enter into associations with other stimuli
that are present through Pavlovian conditioning
processes. At the most fundamental level, Pavlovian
(or classical) conditioning refers to the establishment
of a relation between two stimuli – one termed CS
or conditional stimulus and the other termed US or
unconditioned stimulus. Conditioning is said to occur
when the CS comes to control or modify a response
that it previously had not. For good conditioning,
the CS and US must occur close in time, with CS
onset before US onset; closer spatial proximity between
the stimuli will typically facilitate conditioning.
As a familiar example, in the experiments of Ivan
Pavlov and his colleagues with dogs, the ringing of
a buzzer (the bell) or click of a metronome was
the CS, and the food was the US. With repeated
pairings of buzzer CS-food US, the buzzer came to
elicit salivation on its own (termed conditioned
response or CR).
Most often, drug conditioning refers to the drug as
the US and the exteroceptive stimuli present at the time
as the CSs. Take as an example smoking. From this
perspective, the cigarette, lighter, smell, and taste of
smoke, throat irritation, fellow smokers, and smoking
areas (car, living room, pub) are potential CSs that are
repeatedly paired with the physiological effects of nicotine – the primary addictive constituent of tobacco.
Through these pairing, the smoking-related CSs come
to evoke CRs that are often referred to as withdrawal,
cravings, urges, etc., that lead to drug seeking and
precipitate relapse. As will be detailed in the following
paragraphs, these CRs can be enhancement of
approach and other drug-like responses to the CS
(conditioned sensitization) or can be a CR that opposes
a drug effect (conditioned tolerance).
Much effort has been spent by researchers to understand the behavioral and physiological processes

D

1044

D

Drug Conditioning

underlying drug conditioning in which environmental
stimuli come to control approach and other drug-like
conditioned responses. The development of animal
models to study such effects in more detail reflects
the translational link between the clinical and human
laboratory observations that presentation of drugassociated paraphernalia to an addict will increase
drug urges, cravings, wanting, as well as changes in
physiological measures (Little et al. 2005). Two widely
used animal models include conditioned place preference and conditioned locomotor sensitization. In the
place conditioning procedure, the animal (usually a rat
or mouse) receives a drug paired with one environment
(the context CS). On a separate day, the animal is
exposed to a second distinct environment without
the drug. This procedure is typically repeated until
the context CS has been paired several times with the
drug US. In a later choice test, the animal is given
unrestricted access to both environments. If the
drug US has rewarding effects, then more time will be
spent in the drug-paired environment. This increase in
time spent in the paired environment is thought to
reflect an anticipatory approach CR evoked by the
environment CS and has been described as a model
of drug seeking.
In the conditioned locomotor sensitization task,
the animal receives an environment (the CS), in
which activity can be measured, paired repeatedly
with the drug US (note that there is not a second
environment as in the place conditioning task). Activity tends to increase across repeated administration
with stimulant drugs such as amphetamine, methamphetamine, cocaine, and nicotine (termed locomotor
sensitization). Conditioning to the context CS is
assessed in either a drug-free test day and/or a drugchallenge test. Rats that had the environment paired
with drug are more active than controls that receive
equal exposure to drug in the homecage (termed
conditioned locomotor sensitization or conditioned
hyperactivity). Research from these tasks, and others,
has spawned a class of theories of addiction based
on sensitization of behavior and neural processes
that leads to increased motivation for drug and
compulsive use that is mediated in part by these
drug conditioning processes (e.g., Robinson and
Berridge 2008).

Another theory of addiction, the opponent process
theory, may help to explain some drug conditioning
effects. In the opponent process model, the effects of
a drug change with repeated administration over time
because the body produces an adaptive response to
counter the effects of the drug and return the body to
homeostasis (Koob et al. 1997). This process may in
part explain why higher doses are needed to evoke
a response similar to the initial effects of the drug, or
in other words why tolerance develops. In addition, in
the absence of the drug, the body may be in a state
below homeostasis, and this may contribute to withdrawal symptoms and/or drug cravings. Initially, the
drug evokes this response, but this drug effect can
become conditioned to external stimuli such that the
stimuli can then evoke a conditioned change in homeostasis in anticipation of drug administration. This effect
can result in conditioned tolerance (Siegel et al. 2000).
With conditioned tolerance, a higher dose of a drug can
be tolerated when external stimuli associated with prior
administration of the drug are present. If such stimuli
are absent, administration of the drug may lead to an
overdose. Conditioned tolerance can occur with discrete stimuli such as a syringe and with compound
stimuli such as a context. The ability of these CSs to
evoke an adaptive response counter to the anticipated
effects of a drug may also explain in part why drugassociated stimuli can evoke drug craving and seeking
behavior when they are presented in the absence
of the drug.
Many drugs ingested by people also have perceptible stimulus effects. Anecdotally, this is reported as
jittery when one has had too much of a caffeinated
beverage or drowsy after taking some cold medications.
Thus, drug conditioning can also refer to these interoceptive stimuli functioning as a CS. In this role, the CS
becomes associated with other USs. This area of drug
conditioning has received much less attention than that
of a drug as the US. In the smoking example, the CS
effects of nicotine may be paired with alcohol, socializing, work breaks, relief from stress, etc. As in the
earlier example, the nicotine CS will come to evoke
CRs that reflects this conditioning history. If
these USs are hedonically positive, as in the examples
just given, then this conditioning may increase the
desire or motivation for nicotine, hence increasing

Drug Conditioning

the tenacity of the smoking habit. Drug stimuli may
also serve as modulators or occasion setters. In this
case, the drug state signals whether or not the CS
will be followed by a US. As an example, the nicotine
stimulus may occasion when the outside smoking
area at work (CS) is associated with relief from stress
of job (US).
Drug conditioning research on drugs as Pavlovian
stimuli (CSs or occasion setters) in humans is limited
to a few demonstrational studies. Thus, what is known
to date has relied on animal models – mostly using rats
(Bevins and Murray 2011). This research has shown
that drugs from different classes (stimulant, anxiolytic,
hallucinogen, etc.) can serve as a CS and/or as an
occasion setter. One of the most widely studied drug
CSs is nicotine. That research uses a discriminated
goal-tracking task in rats. In that task, rats are given
intermixed nicotine and saline sessions. There is intermittent access to liquid sucrose on sessions where nicotine was injected pre-session. No sucrose is available
on saline sessions. Nicotine comes to differentially control anticipatory approach and food-searching CR near
the site where sucrose is delivered (i.e., goal-tracking).
This drug conditioning follows many of the rules
established in more widely studied tasks using exteroceptive stimuli such as tones or lights as CSs. For
example, the magnitude of the nicotine-evoked CR is
affected by the quality and frequency of the US, the
salience of the stimulus (i.e., nicotine dose), and the
removal of the US (i.e., extinction).
The occasion setting version of this task merely
adds presentations of a discrete CS such as a light or
white noise to each session. If drug is a positive feature,
then on drug sessions the CS will be paired with the US;
CS presentation will be non-reinforced on saline sessions. In this case, the discrete CS (e.g., 15 s of light
illumination) will evoke a goal-tracking CR only when
the drug has been administered. A drug may also serve
as a negative feature. Here, the discrete CS is paired
with the US only on no-drug (saline) sessions. Rats
learn to withhold responding to the light on nicotine
sessions. Other drug conditioning research in this area
has used the conditioned taste aversion paradigm to
study drugs as occasion setters. In this discriminated
taste aversion task, a drug such as morphine will
disambiguate when a novel taste (CS) will be paired

D

with lithium chloride induced illness (US) in thirsty
rats. Thus, drinking the taste CS is withheld if the drug
is the positive feature signaling that illness will
follow consumption of the tastant. Alternatively, the
rat will readily consume the tastant if morphine
signals no illness.

Important Scientific Research and
Open Questions
The bulk of drug conditioning research has been in
the area of drug addiction. Basic research in this
area has informed treatment approaches. For instance,
cue-exposure therapies reflects the translation of
research on extinction learning to clinical attempts to
reduce reported cravings or urges evoked by drugrelated CSs. Extinction refers to presentation of
the CS without the US. Non-reinforced presentation
of the CS results in a decrease in conditioned
responding evoked by the CS. This decrease reflects
new learning that competes with or inhibits the earlier
learning. Cue-exposure therapy attempts to create
this competing learning by presenting the drugrelated stimuli (cigarette, ashtray, guided imagery
of smoking place) without exposure to the drug US
(nicotine). Such approaches have had limited success.
Future research examining ways of increasing the effectiveness of cue-exposure therapy and identifying alternative approaches will be important (cf. Conklin and
Tiffany 2002).
Drug conditioning research in other important
health areas is quite limited. There is some research
suggesting that drug conditioning processes are involved
in antipsychotic medication effectiveness, insulin treatment for diabetes, and adherence to chemotherapy
schedules. In this latter example, it appears that hospital/clinic-related stimuli (the CS) become associated
with the illness induced by chemotherapeutic agents
(the US). These treatment-site cues (smell of hospital,
sight of clinic, etc.) come to evoked anticipatory nausea
(see earlier discussion of conditioned taste aversion).
This CS-evoked nausea can lead to patients skipping
critical treatment appointments. From a drug conditioning perspective, antiemetic medications are helpful
because they blunt US magnitude (i.e., degree of nausea), thus decreasing conditioning to the treatment CSs.
There is little doubt about the importance of drug

1045

D

1046

D

Dual Enrollment

conditioning effects. However, it is still unclear how
ubiquitous they may be and to what extent they impact
a particular individual or health area. Finally, as detailed
earlier, there is very little drug conditioning work in
either human or nonhuman animals examining the
mechanism or import of their interoceptive stimulus
effects serving as CSs or occasion setters. What work
has been done suggests that increased research and consideration of such effects will enhance our understanding and treatment approaches to many health issues.

Cross-References
▶ Amphetamine, Arousal and Learning
▶ Associative Learning
▶ Conditioning
▶ Context Conditioning
▶ Directed Associations and (Incentive) Learning
▶ Extinction Learning, Reconsolidation and the Internal Reinforcement Hypotheses
▶ Habit Learning in Animals
▶ Pavlov, Ivan P. (1849–1936)
▶ Pavlovian Conditioning
▶ Place Preference Learning

References
Bevins, R. A., & Murray, J. E. (2011). Internal stimuli generated
by abused substances: Role of Pavlovian conditioning and its
implications for drug addiction. In T. Schachtman & S. Reilly
(Eds.), Associative learning and conditioning: Human and animal
applications. New York: Oxford University Press.
Conklin, C. A., & Tiffany, S. T. (2002). Applying extinction research
and theory to cue-exposure addiction treatments. Addiction, 97,
155–167.
Koob, G. F., Caine, S. B., Parsons, L., Markou, A., & Weiss, F. (1997).
Opponent process model and psychostimulant addiction. Pharmacology, Biochemistry and Behavior, 57, 513–521.
Little, H. J., Stephens, D. N., Ripley, T. L., Borlikova, G., Duka, T.,
Schubert, M., Albrecht, D., Becker, H. C., Lopez, M. F., Weiss, F.,
Drummond, C., Peoples, M., & Cunningham, C. (2005). Alcohol
withdrawal and conditioning. Alcoholism, Clinical and Experimental Research, 29, 453–464.
Robinson, T. E., & Berridge, K. C. (2008). Review. The incentive
sensitization theory of addiction: Some current issues. Philosophical Transactions of the Royal Society B: Biological Sciences, 363,
3137–3146.
Siegel, S., Baptista, M. A., Kim, J. A., McDonald, R. V., & Weise-Kelly,
L. (2000). Pavlovian psychopharmacology: The associative basis
of tolerance. Experimental and Clinical Psychopharmacology, 8,
276–293.

Dual Enrollment
▶ Interactive Skills and Dual Learning Processes

Dual-Process Models of
Information Processing
HEATHER M. CLAYPOOL1, JAMIE O’MALLY2
JAMIE DECOSTER3
1
Department of Psychology, Miami University, Oxford,
OH, USA
2
Department of Psychology, University of Alabama,
Tuscaloosa, AL, USA
3
Institute for Social Science Research, University of
Alabama, Tuscaloosa, AL, USA

Synonyms
Two-process models; Dual-process theories

Definition
Dual-process models of information processing contend
that humans use two different processing styles. One is
a quick and automatic style that relies on well-learned
information and heuristic cues. The other is a qualitatively
different style that is slower, more deliberative, and relies
on rules and symbolic logic. Dual-process models have
been developed to explain specific psychological phenomena, such as persuasion, person perception, attribution,
and stereotyping. Other more generalized models have
been proposed to simultaneously explain processing
across a variety of domains. The labels applied to the
processing styles vary from theory to theory. The automatic style has been labeled “heuristic,” “peripheral,”
“experiential,” “impulsive,” and “associative,” whereas
the more deliberative style has been labeled “systematic,”
“central,” “rational,” “reflective,” and “rule-based.” In the
remainder of this entry, the former type of processing is
referred to as “automatic” and the latter as “controlled.”

Theoretical Background
Dual-process models of information processing are
plentiful in psychology, especially in the areas of social

Dual-Process Models of Information Processing

and cognitive psychology. The theoretical underpinnings of these models can be traced back as far as
William James and Sigmund Freud, who both postulated two types of reasoning, one that is associative and
another that is analytical and rational. In the modern
era, the building blocks of various dual-process theories vary depending on the area in question.
In social psychology, dual-process models
began in the area of attitude change and persuasion
with the publication of two similar theories at approximately the same time: the Elaboration Likelihood Model (Petty and Cacioppo 1981) and the
Heuristic-Systematic
Model
(Chaiken
1980).
According to these perspectives, attitudes can change
from exposure to a persuasive expert either because
someone mindlessly relies on a simple heuristic, like
“trust experts,” or because someone carefully considers
the specific arguments contained in a message. Fairly
superficial processing occurs in the first case,
whereas deep and careful processing occurs in the
second.
By the turn of the century, dual-process models
had become so popular in social psychology that
an entire book was published detailing them
(Chaiken and Trope 1999). Though the many dualprocess models use different terminology and
disagree on some important issues, there is a great
deal of conceptual similarity among them (Smith and
DeCoster 2000). First, there is a broad consensus
regarding the characteristics of the two processing
modes. Second, these models tend to agree that the
automatic mode operates preconsciously, with only
the outcome of the processing available to conscious
awareness, and that the controlled mode operates
consciously, such that individuals are aware of the
outcome of processing as well as the steps involved.
Finally, several of the models suggest that whereas
automatic processing can be performed under most
circumstances, controlled processing typically occurs
only when one has both the motivation and ability
to do so. These notions underscore the functional
benefit of having two processing styles. Given the number of cognitive operations humans must regularly
perform (such as categorizing stimuli, drawing
causal inferences, and making decisions), it would
be impossible for anyone to carefully and deliberately

D

execute every action. However, humans are sometimes motivated to thoroughly and consciously consider particular categorizations, inferences, and
decisions, so it is important that people are capable
of giving detailed attention to important issues. The
existence of two processing styles allows humans to
flexibly process automatically or deliberately as
necessary.

Important Scientific Research and
Open Questions
Despite the commonalities among dual-process models,
there are a number of debates in the literature related to
this topic. One question is whether the two processing
modes can operate simultaneously. Some models argue
that processing is either automatic or controlled, whereas
other models argue that both processing modes can
occur in parallel. Many models assume that the automatic mode always operates because of its quick and easy
nature. In these models, a person might additionally and
simultaneously use controlled processing when the task
is important enough and when the individual has the
ability to do so.
A second issue is whether the two processes
are linked to different cognitive systems. Smith and
DeCoster (2000) argue that the two processing
modes rely differentially on two memory systems with
different properties that operate simultaneously.
A “slow-learning” system picks up regularities in
the environment over time and thus forms the basis
of a knowledge store representing a vast array of
experiences. A “fast-learning” system allows for the
rapid binding of information into memory from
a single occurrence. According to Smith and DeCoster,
the automatic mode solely makes use of information
from the slow-learning system, whereas the controlled
mode makes use of information from both the fastand slow-learning systems. Other dual-process
models make no effort to link their processing modes
to different cognitive or neurological systems. Moreover, there is increasing skepticism regarding the
presence of distinct cognitive/neurological systems
(e.g., Keren and Schul 2009) which, by extension, renders the linkage of different processing styles to distinct
systems dubious. Thus, some researchers believe that
there may be two different types of information

1047

D

1048

D

Dual-Process Theories

processing, but do not ascribe to the notion that
the two processing modes are executed by different
cognitive systems.
In addition, some researchers question whether
there are even two distinct processing modes.
These models suggest that the numerous dualprocessing models are unnecessary, and that all
processing can be explained by greater or lesser engagement of a single process. For example, the “unimodel”
(Kruglanski et al. 1999) argues that all types of persuasion are qualitatively identical, involving “if-then” reasoning from “evidence,” but that the amount of
processing used to develop the evidence may vary
from instance to instance. Such single-process accounts
tout their abilities to explain data in a more parsimonious fashion than their dual-process counterparts.

SARAH A. FRASER1, KAREN Z. H. LI2
1
Department of Psychology, Centre de recherche de
l’institut universitaire de gériatrie de Montréal,
Université du Québec à Montréal, Montréal,
Québec, Canada
2
Department of Psychology, Center for Research in
Human Development, Concordia University,
Montréal, Québec, Canada

Cross-References

Definition

▶ Automatic Information Processing
▶ Controlled Information Processing
▶ Heuristics and Problem Solving
▶ Human Information Processing

Dual-task performance requires an individual to perform two tasks (i.e., Task A and Task B) simultaneously.
Typically this type of performance is contrasted with
single-task performance in which the individual only
has to perform one task at a time (Task A or B).
Motor learning occurs when an individual demonstrates relatively enduring improvements in their capability to perform a motor task after practice.

References
Chaiken, S. (1980). Heuristic versus systematic information
processing and the use of source versus message cues in
persuasion. Journal of Personality and Social Psychology, 39,
752–766.
Chaiken, S., & Trope, Y. (Eds.). (1999). Dual-process theories in social
psychology. New York: Guilford.
Keren, G., & Schul, Y. (2009). Two is not always better than one:
A critical evaluation of two-system theories. Perspectives on Psychological Science, 4, 533–550.
Kruglanski, A. W., Thompson, E. P., & Spiegel, S. (1999). Separate
or equal? Bimodal notions of persuasion and a singleprocess “unimodel”. In S. Chaiken & Y. Trope (Eds.), Dualprocess theories in social psychology (pp. 293–322). New York:
Guilford.
Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic
and contemporary approaches. Dubuque: Brown.
Smith, E. R., & DeCoster, J. (2000). Dual-process models in social and
cognitive psychology: Conceptual integration and links to
underlying memory systems. Personality and Social Psychology
Review, 4, 108–131.

Dual-Process Theories
▶ Dual-Process Models of Information Processing

Dual-Task Performance in
Motor Learning

Synonyms
Divided attention

Theoretical Background
Motor learning proceeds in stages. Historically, three
stages of motor learning were proposed (Fitts and
Posner 1967). The first stage was named the cognitive
stage, the second the associative stage, and the third the
autonomous stage. One of the reasons for naming the
first stage the cognitive stage is that cognitive processes
are highly involved in this stage of learning. In particular, attention to the instructions and to the demands
of the motor task to be learned is crucial during this
stage of learning. In contrast, the autonomous stage or
third stage of learning is considered a relatively automatic stage in which the motor task can be performed
with little or no need for attention. Finally, as would be
expected, the attentional requirements of the second
associative stage lie somewhere between the requirements for the first and third stages.
A paradigm commonly used to assess and distinguish the learning stages that have the highest and
lowest attentional requirements (first and third stages)
is the dual-task paradigm. In cognitive dual-task

Dual-Task Performance in Motor Learning

research, typically experimenters ask participants to
perform a motor task and an auditory task simultaneously (dual task) or separately (single task). Calculation of dual-task costs (dual versus single-task
performances) provides an indication of the extent to
which the two tasks compete for common processes.
An early model of attention assumes that there is
a single, limited attentional capacity (Kahneman
1973). Therefore, when attention is divided between
two tasks in the dual-task paradigm and both tasks
draw on this limited capacity there should be performance decrements in one or both tasks. Subsequent
dual-task experiments revealed a mixed pattern of
dual-task costs or interference, indicating that the
degree of interference can vary as a function of competition for input (sensory), representational, or output
(vocal, motoric) processes. Nevertheless, carefully chosen dual-task combinations with minimal levels of
interference have been used to determine the degree
of effortful cognitive processing associated with a given
task or stage of processing. The basic premise of this
approach is that tasks that are well practiced or automatized should not be vulnerable to interference from
a concurrent task, whereas tasks requiring effortful
processing should show reduced performance when
paired with a concurrent task.
In terms of motor learning, dual-task performance
(auditory + motor) is compared to single-task performance (motor only) in order to assess how much the
performance of the auditory task interfered with motor
task performance. In the cognitive stage, where attention is important to motor learning, divided attention
between a motor and an auditory task will cause the
greatest performance decrements (relative to other
stages). Later in the autonomous stage of learning
when the motor task has been well practiced and is
highly automatized, performance decrements in
a dual-task situation are minimized. As such, high
and low levels of interference measured with the dualtask paradigm are an index of the different learning
stages.

Important Scientific Research and
Open Questions
In addition to delineating the different stages of learning, the dual-task paradigm has also been useful in
clarifying what processes are involved in motor learning. In particular, many researchers have attempted to

D

demonstrate that implicit and explicit learning may
require different attentional processes. Generally, it
has been argued that implicit learning occurs without
awareness and conscious control, whereas explicit
learning is strategic and requires conscious processing.
By this definition, explicit learning would require more
attentional processes than implicit. In order to assess
the attentional involvement in these two types of learning, many researchers have used the dual-task paradigm to evaluate whether the dual-task interference is
less when the motor task is presented implicitly versus
explicitly. Typically, implicit and explicit versions of
a motor sequence task or serial reaction time task are
used. In this type of task, individuals are asked to
reproduce a sequence of motor events (i.e., 1-4-2-3-41-3-2) presented visually, on a piano-like keyboard. In
the implicit version participants are not told there is
a sequence to be learned (simply tap along with the
motor events as they appear), whereas in the explicit
version participants are told to try and memorize the
motor events and follow the specific sequence. While
the level of attentional involvement in explicit and
implicit motor learning is still under debate, results
do indicate that dual-task interference may be less in
the initial stages of learning when the task is presented
implicitly versus explicitly. However, even when the
task is presented implicitly, learning is impaired when
a second task is presented suggesting that some attentional processes are also required when learning is
implicit.
Although motor learning research is based on
behavioral outcomes (i.e., improvements in speed and
accuracy during the performance of a motor task)
recent work involving neuroimaging techniques has
helped to extend the behavioral work and clarify the
brain areas associated with motor learning. As in the
behavioral research, neuroimaging research has used
the dual-task paradigm to assess what brain areas are
more involved in the cognitive stage versus the autonomous stage. Results from neuroimaging converge
with the existing behavioral results demonstrating
a shift from very frontally mediated activation during
early learning to more parietal and subcortical activations in later learning. Since the prefrontal cortex is
strongly associated with attentional processes, these
results align well with the proposal that attentional
processes are more heavily recruited in the early stages
of learning.

1049

D

1050

D

Duncker, Karl (1903–1940)

In addition to types of learning and underlying
brain areas involved in motor learning, an important
question assessed with the dual-task paradigm is: how
do different groups learn a motor task under divided
attention conditions? For example, results comparing
healthy controls to individuals with Parkinson’s disease
(PD) have revealed that the dual-task interference is
greater in PD but providing cues during learning can
significantly diminish the interference and improve
motor learning. These findings and neuroimaging findings in specific populations (children, older adults,
individuals with brain damage) can help to better
inform what processes are involved in motor learning,
what processes might be most affected by disease and
injury, and what processes are most amenable to rehabilitation strategies.

Cross-References
▶ Aging Effects on Motor Learning
▶ Attention and Implicit Learning
▶ Explicit Versus Implicit Learning
▶ Implicit Sequence Learning
▶ Motor Learning

References
Fitts, P. M., & Posner, M. I. (1967). Human performance. Belmont:
Brooks/Cole.
Kahneman, D. (1973). Attention and effort. Englewood Cliffs: Prentice-Hall.

Duncker, Karl (1903–1940)
NORBERT M. SEEL
Department of Education, University of Freiburg,
Freiburg, Germany

Life Dates
Karl Duncker was born in Leipzig, Germany, on February 2, 1903. Duncker was one of the most prominent
Gestalt psychologists. He was a student and coworker
of Wertheimer, Koffka, and Köhler in Berlin. His parents were the communist politicians Herman and Käte
Duncker. As a consequence, Duncker’s thesis for habilitation was rejected in 1934 but was then published in

1935. In the same year, he found political asylum in
Cambridge, England, where he worked with Frederic
Bartlett. Finally, Duncker accepted an offer from
Wolfgang Köhler and emigrated to the United States
of America, where he worked at Swarthmore College in
Pennsylvania until his death shortly after his 37th
birthday. He committed suicide in 1940 after some
time of suffering from depression (Schnall 1999).

Theoretical Background
Duncker stands alongside the famous trio of Gestalt
psychologists (Wertheimer, Koffka, and Köhler)
and contributed essential and influential research
on productive thinking and creative problem solving.
Furthermore, he was highly interested in the psychology of ethics and in research on the phenomenology
of feelings and sensual perception. Duncker coined
the term functional fixedness, which refers to difficulties
in visual perception and in problem solving due to
the fact that one element of a task or situation has
a fixed function which inhibits the restructuring
of the task or situation necessary to find the solution
to the problem.

Contributions to the Field of Learning
In his seminal research on problem solving, Duncker
worked with relatively simple laboratory tasks that
appeared novel to the participants, such as the X-ray
problem. This problem confronts subjects with the task
of how they would destroy an inoperable stomach
tumor with X-rays without destroying the surrounding
healthy tissues. The subjects’ verbalizations led
Duncker to infer that their problem analysis was
based both on conflict analysis (i.e., spatial coincidence
of X-rays and healthy tissue) and on the analysis of
functional fixedness, which hindered them from finding the optimal solution. In the 1980s, the X-ray problem was often used within the realm of research on
analogical problem solving (see, e.g., Gick and Holyoak
1983). Another popular example from Duncker’s
research is the candle problem, which is described in
nearly every textbook on problem solving. The difficulty of this problem arises from the functional fixedness of the candle box. It is a container in the problem
situation but must be used as a shelf in the solution
situation. Further problems Duncker used to illustrate
the necessity of overcoming functional fixedness were
a task in which an electromagnet had to be used as part

Dynamic Modeling and Analogies

of a pendulum, one in which a tree branch had to be
used as a tool, and one in which a brick needed to be
used as a paper weight.
Duncker’s seminal work on problem solving and
functional fixedness fell into oblivion for some decades.
Then it exerted great influence on Simon and Newell
(Newell 1985; Simon 1999) in their development of the
General Problem Solver. Simon (1999) has described in
detail how strongly Duncker’s theories and research
influenced the emergence of cognitive science. However, Duncker not only experienced a renaissance in the
United States but also in Germany, where Dietrich
Dörner (1976) – probably the most prominent German
researcher in the field of problem solving – referred to
Duncker and rescued this great Gestalt psychologist
from obscurity.

D

Dynamic Assessment
▶ Dynamic Testing and Assessment

D

Dynamic Decision Making
▶ Complex Problem Solving

Dynamic Mapping
▶ Initial State Learning

Cross-References
▶ Complex Problem Solving
▶ Gestalt Psychology of Learning
▶ Problem Solving
▶ Problems: Definition, Types, and Evidences

References
Dörner, D. (1976). Problemlösen als Inf sverarbeitung. Stuttgart:
Kohlhammer.
Duncker, K. (1926). A qualitative (experimental and theoretical)
study of productive thinking solving of comprehensible problems. Pedagogical Seminary and Journal of Genetic Psychology, 33,
642–708.
Duncker, K. (1935). Zur Psychologie des produktiven Denkens. Berlin:
Springer.
Gick, M., & Holyoak, K. J. (1983). Schema induction and analogical
transfer. Cognitive Psychology, 15, 1–38.
Newell, A. (1985). Duncker on thinking: An inquiry into process in
cognition. In S. Koch & D. E. Leary (Eds.), A century of psychology
as science (pp. 392–419). New York: McGraw-Hill.
Schnall, S. (1999). Life as the problem: Karl Duncker’s context. From
past to future: The drama of Karl Duncker. Papers on the History
of Psychology, 1(2), 13–28.
Simon, H. A. (1999). Karl Duncker and cognitive science. From Past
to Future: The Drama of Karl Duncker. Papers on the History of
Psychology, 1(2), 1–11.

Duration Discrimination
▶ Temporal Learning in Humans and Other Animals

1051

Dynamic Modeling and
Analogies
NORBERT M. SEEL
Department of Education, University of Freiburg,
Freiburg, Germany

Synonyms
Analogy-based modeling; Dynamical analogies; Simulation model of analogy-making; Simulation of
dynamic systems; Systems dynamics

Definition
Dynamic modeling describes the behavior of
a distributed parameter system in terms of how one
qualitative state can turn into another. Systems can
be deterministic or stochastic (depending on the
types of elements that exist in the system), discrete or
continuous (depending on the nature of time and how
the system state changes in relation to time), and static
or dynamic (depending on whether or not the system
changes over time at all). This categorization of systems
affects the type of modeling: Models, like the systems
they represent, can be static or dynamic, discrete or
continuous, and deterministic or stochastic. Dynamic
models describe the behavior of a distributed parameter system in terms of how one qualitative state can

1052

D

Dynamic Modeling and Analogies

turn into another. Accordingly, dynamic models contain at least two variables at two different times that are
causally related:
Yt ¼ f ðYt1 ; :::; Ytk Þ
That means, at the time t vector Y functionally
depends on the state of Y at former times t-k. The
system’s dynamics consists in the fact that states or
interventions in former times affect the state of Y in
the point t.
Dynamic models have historically been based on
analogies which can be defined as likeness or similarity
in some respects between things that are otherwise
dissimilar. In logic analogies constitute a form of reasoning in which a similarity between two or more
things is inferred from a known similarity between
them in other respects. For instance, dynamic models
of continuous manufacturing systems were often based
on analogies with electrical systems. Or macroeconomic systems were based on analogies with circuit
models. This corresponds with the conception of analogical problem solving as a process of transferring
knowledge from a well-understood base domain to
a new target problem area.

Theoretical Background
Dynamic modeling presupposes functional intentionality in constructing and using models: First, models
may serve envisioning (i.e., making visible the invisible)
and the simplifying of an investigation to particular and
relevant phenomena in a closed domain. Secondly, and
more relevant for dynamic modeling, is the use of
models for simulating transformations of states of
a complex system. These simulation models allow
a learner to explore a dynamic system in a controlled
way to understand how the system’s components interact, and how alternate decisions can affect desired outcomes. Finally, models are often constructed as
analogies in order to map a well-known explanation
(e.g., Rutherford’s atomic model) onto a phenomenon
to be explained (e.g., quantum mechanisms). Such
models are called analogy models which are heuristic
hypotheses about structural similarities of different
domains (Holland et al. 1986). The cognitive psychology literature speaks of analogy models when the conditions in an unknown domain are inferred and
mapped in analogy to known conditions in the same
or another domain (e.g., viewing mechanical

phenomena in an electrodynamic model or memory
processes in analogy to computers). Analogy models
may be understood as heuristic hypotheses of
a structural similarity between different domains.
From an epistemological point of view, there are
two general conceptions concerning the nature of
dynamic modeling. The first conception emphasizes
the representational character of modeling – the
model represents reality, it is a model of something;
the other conception considers the model as
a cognitive artifact which is constructed intentionally
in order to create subjective plausibility with regard to
the original (Wartofsky 1979). In the related literature,
this kind of models is called mental (or internal)
model. The idea that learners create, store, and manipulate mental models of dynamic systems with which
they interact has been central to the theory and practice
of system dynamics since its inception. According to
this view, people can understand and manage dynamic
systems by constructing a mental model that helps to
simulate the likely outcomes of the systems’ behavior.
Learning occurs by comparing the expected results of
operations on a system with the observed consequences
of transformations. In the case of gaps between expectations and observations, the outcomes are used to
update or revise the mental model. Dynamic modeling
requires the identification of the system’s components
and their interactions. Therefore, a general plan for the
development of the model’s overall structure (and
functions) is necessary. Mental models provide
a rationale for the development of such a master plan
to operate effectively with the complexity of dynamic
systems. Given the central role of mental models in
operating successfully with dynamic systems, it is not
surprising that one of the primary goals of system
dynamics interventions is to change mental models to
make them more accurate, more complex, and more
dynamic.
In the practice of dynamic modeling, various tools
and methods can be used to create analogical models.
For instance, a mechanical device can be used to represent mathematical calculations, or the flow of water
can be used to model economic systems; electronic
circuits can be used to represent both physiological
and ecological systems. Since the emergence of ▶ system dynamics this goal is accomplished by using particular computer-based tools for model building, such
as STELLA, Powersim, and Model-it. These tools allow

Dynamic Modeling and Analogies

their users to construct their own dynamic models of
complex systems (Clariana and Strobel 2008). Basically,
these tools can be variable-based or agent-based.
STELLA, Powersim, and Model-it are popular examples of variable-based modeling tools and operate with
a particular formalized language for qualitative modeling (see in more detail: Hannon and Ruth 2001).
The resulting models are called system dynamics
models whose variables have no attributes but change
dynamically based on the mathematical models and
equations that define the relationships within the system. Model-it, for instance, allows the learner to construct qualitative models of cause and effect
relationships. Through this technology, the user creates
objects with which he or she associates measurable,
variable quantities called factors and then defines relationships between those factors to show the effects of
one factor upon another. Model-it provides facilities
for testing a model and a “Factor Map” for visualizing
it as a whole. Students define objects, factors, and the
relationship between factors’ qualities. The student is
facilitated in this modeling process by a variety of scaffolds. These scaffolds include features which (a) allow
for multiple linked representations, (b) options which
hide additional complexity, (c) learner guidance
through subtasks, and (d) prompting for explanations
for constructed relationships.
The other broad class of model-building tools puts
agent-based modeling at the center of development and
research. An agent-based model consists in a collection
of autonomous agents which interact between themselves ruled by some simple characteristics that are
modeled from the observation or analysis of realworld entities in order to simulate their more complex
characteristics. Agents can be considered as computational entities that can conceptually incorporate in
a natural way some human mechanisms such as perception, action selection, autonomy, etc. Clariana and
Strobel (2008) have pointed out that in agent-based
modeling tools the agents possess different attributes
and their behavior changes in accordance with
a predefined set of rules to respond to the situational
demands of a simulated environment. Actually, the
virtual or simulated environment with its spatial
dimensions provides the space where the agents
move, interact, and pursue their goals. Thereby, the
agents mimic the behavior of real people in a real environment. Agent-based simulation tools have been

D

widely used in geography (da Silva et al. 2004), social
sciences (Tobias and Hofmann 2004), and construction
management (Watkins and Mukherjee 2008).
Both variable-based and agent-based dynamic
modeling often use analogies to understand novel situations in order to “bootstrap” new knowledge based
on previously learned knowledge. Actually, the formation of analogies is an effective way of dealing with
complexity (cf. Hannon and Ruth 2001). The essence
of analogy formation is to identify the structure of one
system and compare it with structure of another system
whereby the similarities and differences between the
two systems are identified. Of course, the similarities
between the systems generate a particular set of
insights into the systems’ structural commonalities
but, at the same time, the dissimilarities provoke
a complementary view on both systems and show the
bounds of the analogy. Inextricably linked with the
formation of analogies and analogy models are conclusions by analogy for which Dörner (1976) has defined
the following steps: (1) abstraction of selected attributes of the phenomenon in question (esp. as regards
content), (2) the search for a model, which constitutes
a concretization of the abstract phenomenon, (3) transfer of (structural) attributes of the model back to the
original phenomenon, and (4) test whether the
hypothesized attributes are actually present in the original phenomenon. This procedure allows realizing
what abstractions are important for the formation of
the analogy – and which substructures of both the
original and the model can be disregarded.

Important Scientific Research and
Open Questions
Dynamic modeling provides a new perspective called
learning by system modeling and an extension to
approaches of simulations: Students are building their
own models and engaging at a much deeper conceptual
level of understanding of the content, processes, and
problem solving of the domain. Research from the area
of mindtools suggests that students’ conceptual understanding and application of knowledge is much deeper
and advanced when they are involved in learning by
modeling (Jonassen 1999). In consequence, variablebased dynamic modeling is increasingly considered to
create innovative learning environments which are consistent with how people learn: Variables can be limited to
a manageable level and structure and direction for

1053

D

1054

D

Dynamic Network Analysis (DNA)

learning can be provided, real-world problems can be
addressed, and students can take control and responsibility for their own learning progress. Dynamic modeling motivates students by keeping them actively engaged
in the learning process through requiring that problemsolving and decision-making skills be used to make the
simulation run. As the simulation runs, it is modeling
a dynamic system in which the learner is involved and
plays an active role as decision maker and problem
solver. Dynamic modeling enables students to engage
in systems thinking and problem solving by explorations
and analogical reasoning.
However, this puts high cognitive and metacognitive
demands on students and requires the construction of
powerful mental models (Doyle and Ford 1998; Isaacson
and Fujita 2006). However, both practical experience in
the field of system dynamics and controlled laboratory
experiments on dynamic decision making have shown
that mental models of complex systems are typically
subject to a variety of flaws and limitations. For example, mental models often omit feedback loops, time
delays, and nonlinear relationships that are important
determinants of system behavior. In addition, the limited capacity of working memory makes it impossible
for people to mentally simulate the dynamic implications of all but the simplest mental models. According
to the system dynamics view, only by adopting the
feedback perspective and modeling discipline of system
dynamics and taking advantage of the computer’s ability to calculate the dynamic consequences of mental
models can these flaws and limitations be overcome.

Cross-References
▶ Analogical Models
▶ Analogical Reasoning
▶ Calibration
▶ Complex Problem Solving
▶ Computer Simulation Model
▶ Mental Models
▶ Model-Based Reasoning
▶ Modeling and Simulation
▶ Simulation-Based Learning

da Silva, C. A., de Beauclair Seixas, R., & Monteiro de Farias, O. L.
(2004). Geographical Information Systems and Dynamic Modeling via Agent Based Systems. In ACM-GIS’05, 13th ACM international symposium on advances in geographical information
systems, 4–5 Nov 2005, Bremen, Germany.
Dörner, D. (1976). Problemlösen als Informationsverarbeitung. Stuttgart: Kohlhammer [Problem solving as information processing].
Doyle, J. K., & Ford, D. N. (1998). Mental model concepts for system
dynamics research. System Dynamics Review, 14(1), 3–29.
Hannon, B., & Ruth, M. (2001). Dynamic modelling (2nd ed.).
New York: Springer.
Holland, J., Holyoak, K., Nisbett, R., & Thagard, P. (1986). Induction:
Processes of inference, learning, and discovery. Cambridge, MA/
London: MIT Press.
Isaacson, R., & Fujita, F. (2006). Metacognitive knowledge monitoring and self-regulated learning: Academic success and reflections
on learning. Journal of the Scholarship of Teaching and Learning,
6(1), 39–55.
Jonassen, D. H. (1999). Designing constructivist learning environments. In C. M. Reigeluth (Ed.), Instructional design theories and
models: Their current state of the art (2nd ed.). Mahwah: Lawrence Erlbaum.
Tobias, R., & Hofmann, C. (2004). Evaluation of free Java-libraries for
social-scientific agent-based simulation. Journal of Artificial Societies and Social Simulation, 7(1). http://jasss.soc.surrey.ac.uk/7/
1/6.html
Wartofsky, M. W. (1979). Models. Representation and the scientific
understanding. Dordrecht: Reidel.
Watkins, M., & Mukherjee, A. (2008). Using adaptive simulations to
develop cognitive situational models of human decision-making.
Technology, Instruction, Cognition and Learning, 6 (3–4),
177–192.

Dynamic Network Analysis
(DNA)
▶ Social Networks Analysis and the Learning Sciences

Dynamic Network Formation
▶ Networks, Learning Cognition, and Economics

References
Clariana, R. B., & Strobel, J. (2008). Modeling technologies. In J. M.
Spector, M. D. Merrill, J. van Merrienboer, & M. P. Driscoll (Eds.),
Handbook of research on educational communications and technology (3rd ed., pp. 329–344). New York: Lawrence Erlbaum.

Dynamic Selection
▶ Multistrategy Learning

Dynamic Testing and Assessment

Dynamic Testing
▶ Dynamic Testing and Assessment

Dynamic Testing and
Assessment
WILMA C. M. RESING1, JULIAN G. ELLIOTT2
ELENA L. GRIGORENKO3
1
Department of Psychology, Department of
Developmental and Educational Psychology, Leiden
University, Leiden, The Netherlands
2
School of Education, Durham University,
Durham, UK
3
Department of Psychology, Department of
Epidemiology & Public Health, Child Study Center,
Yale University, New Haven, USA

Synonyms
Dynamic assessment; Dynamic testing; Learning
potential tests; Learning tests; Testing-the-limits

Definition
Dynamic testing and dynamic assessment (DT/A)
are two overlapping umbrella concepts for forms of
testing and assessment in which instruction or intervention are integral aspects of the testing procedure.
For some, dynamic assessment is a construct that
can be used to describe a general approach. However,
in line with more recent thinking (e.g., Sternberg and
Grigorenko 2002), we differentiate between dynamic
testing whereby the whole test procedure is transparent,
objective, and repeatable and a wider conception
of dynamic assessment in which testing is paired
with clinical and individualized intervention.
Compared with conventional, static testing and assessment, DT/A provides information yielding greater
insight into the nature of the individual’s cognitive
functioning. This should result in more valid assessment outcomes that can both aid prediction, and
point to the most appropriate means of cognitive
intervention.

D

1055

Theoretical Background
Inspired by Lev Vygotsky in the Soviet Union and
Reuven Feuerstein in Israel, DT/A is a conglomerate
of very different testing formats and assessment aims
(e.g., Grigorenko 2009; Haywood and Lidz 2007; Sternberg and Grigorenko 2002). The nature of the different
procedures is closely related to the differing aims of
testing. These include:
● Predicting those who are most likely to make edu-

cational progress given appropriate forms of
intervention.
● Aiming to develop, improve or modify the individual’s cognitive skills or functions by providing more
enduring support or mediation.
● Examining the effects of graduated prompt techniques on changes in the testee’s strategy use.
Despite their different aims and orientations,
researchers in this field are united in their criticisms
of conventional static testing and assessment. It is
argued that static measures of cognitive functioning
often:
● Reflect knowledge and skills acquired in the past.

This may lead to underestimates of an individual’s
potential as not everyone has had equal opportunities to acquire optimal levels of knowledge and
skills.
● Do not provide information that can be easily
drawn upon when advising on appropriate instructional interventions.
● Overemphasize the products of learning rather than
learning processes. Because of this, scores from such
measures typically shed little light upon how individuals learn, or fail to learn.
● Are seen as less optimal predictors of future educational performance; dynamic scores considerably
add to this prediction.
Although Alfred Binet, the father of intelligence
testing, was convinced by 1916 that intelligence test
scores (including those from his own test), did not
give an appropriate picture of the ability to learn, the
theoretical underpinnings for DT/A were provided
many years later by, among others, Vygotsky and
Feuerstein.
Using the concept of the zone of proximal development (ZPD), Vygotsky distinguished the level of actual
development, at which a person is able to function

D

1056

D

Dynamic Testing and Assessment

unassisted, from the level of potential development,
which concerns problem solving that is benefited by
guidance from parents, teachers, or peers. Individuals
might show a discrepancy between their actual and
potential level of development; such a discrepancy
might be evident in multiple domains. The zone of
proximal development concerns that area that defines
the difference between an individual’s independent and
guided performance on any activity, closing this discrepancy. The more that individuals can raise their
performance, having received appropriate assistance,
the wider the ZPD is considered to be. Vygotsky considered the assessment of one’s potential, by means of
this zone of proximal development, to be the principal
focus of testing and education.
According to Feuerstein, conventional test scores
can underestimate the cognitive capabilities of those
who have had impoverished learning experiences. To
assess the individual’s cognitive potential, Feuerstein
and his colleagues developed the Learning Potential
Assessment Device (LPAD) (1979). The training procedures of this assessment procedure emphasize individually tailored intervention procedures for a variety
of cognitive tasks but also the development of general,
metacognitive skills, such as systematic working and
controlling one’s own impulses. The LPAD was
designed to make lasting changes to the existing cognitive structure of the individual; a point of view that
other researchers in DT/A dispute.
Depending on the theories behind particular formats of DT/A, the form of feedback is either fixed, the
same for all those tested, or individualized, applied in
a tailored fashion contingent upon each individual’s
ongoing performance in the test situation. With this
latter approach, the nature and amount of assistance
provided depend upon individual differences
manifested within the testing context, including those
elements of the test that the examiner decides to prioritize and act upon. Although the large body of work by
Feuerstein has had a seminal influence upon many
psychologists and educationalists, his approach has
failed to gain acceptance by most researchers and has
not led to widespread usage by clinicians. Many DT/A
researchers consider that there is lack of conceptual
clarity in the model, which makes operationalization
of parts of the model difficult and validation problematic (Elliott et al. 2010). They point to the absence
of a standardized approach, the weak test-retest and

inter-rater reliabilities and the poor quality of empirical
studies that have examined this approach. In response,
Feuerstein contends that an assessment procedure that
seeks to conform to all test requirements will lack
clinical sensitivity, will fail to bring out the best
from the testee, and is unlikely to provide a rich
understanding of a person’s cognitive functioning.
DT/A procedures that seek a more scientific
approach typically examine a person’s potential to
profit from training and/or determine on-the-spot
learning processes that may change during testing.
Here the form of assistance is standardized but can
also be adaptive as the nature of the help provided
will reflect individual differences in the testee’s
responses. The aims here are not to engineer permanent progression or an enduring change in cognitive
abilities but, rather, to aid detection of
● Progression over time as a consequence of

instruction
Individual differences in progression
The way in which differences emerge
Variability or constancy in the use of strategies
Differences in the ways children verbalize their solving processes
● Changes in the posttest score order of groups of
children (as a measure of potential), as a consequence of the assessment period
●
●
●
●

Most DT/A procedures include a pretest, a training
phase, and a posttest. During the training phase, the
focus is on the process of solving the task, the feedback
or prompting given, and, importantly, the behavior of
the child. It is also possible to include the training
phase in the pretest, or to replace it by a much longer
instructional phase in the classroom, but in this latter
case researchers prefer to describe this as cognitive
training or a form of response to intervention (RTI).
Within a fixed, noncontingent training format, people
can be tested either individually or in a group. Individualized, contingent instruction however, should always
be given individually.
While some researchers define the difference between
pre- and posttests as a measure of learning potential, it is
generally accepted that these “gain scores” pose
a number of significant measurement difficulties, in
particular, poor reliability. For this reason, it is generally
accepted that it is preferable to use posttest scores in
isolation or advanced statistical procedures to overcome

Dynamic Testing and Assessment

these problems. Another test format, sometimes called
testing-the-limits, provides contingent feedback within
the administration of the static test. Here, the testee
receives assistance as soon as a significant difficulty is
encountered during only one single testing period.
Again, intervention can be based upon a structured
approach, with a predetermined hierarchy of hints, or
individualized according to the particular understandings and perceptions of the tester. Typically, there is no
baseline measurement with this approach.
Other structured approaches provide a series of
prompts or hints in response to testee errors. An example of this is the graduated prompts approach which
involves the ongoing provision of the minimal amount
of assistance necessary to the child until the relevant
test items can be solved independently. The graduated
prompts approach differs somewhat from that typically
used by other dynamic test users as the final unaided
performance of the testee is not the principal concern.
Rather, the approach seeks to highlight the amount of
assistance necessary to achieve prespecified outcomes
and the capacity of the testee to transfer learned principles to novel situations. Thus the child’s learning
potential is not defined on the basis of their best task
performance but rather, is represented by the inverse of
the minimal number of hints necessary to reach
a specified amount of learning. These DT/A procedures
are fully standardized, are based on detailed task and
process analyses, and hint (or “prompt”) sequences are
hierarchically ordered. However, there is variability to
the extent that some approaches are adaptive, that is,
hints are provided which are differentially contingent
upon the individual’s responses. Testers who prefer to
use these approaches will typically prioritize the examination of changes in problem-solving processes and
the use of strategy patterns during testing. The complexity of such approaches is such that they lend themselves to computerized forms of testing.

Research Outcomes and Future
Questions
A number of overviews and meta-analyses on the effectiveness of DT/A have been conducted (Grigorenko
2009; Sternberg and Grigorenko 2002). Swanson and
Lussier (2001) concluded that DT/A substantially
improved testing performance when compared with
static test conditions. However, many DT/A studies
(approximately 60%) had to be excluded from their

D

analyses because of poor methodological and/or psychometric quality. As far as the predictive validity of DT/A
was concerned, results were mixed, and often dependent
on the format of dynamic tests used: the more standardized the testing procedure, the better the predictive
validity (e.g., Sternberg and Grigorenko 2002).
In relation to the use of dynamic measures for
educational and clinical practice, researchers continue
to search for approaches that can yield valuable data
that can inform intervention.
Future research could profitably examine why
a potentially very valuable form of testing continues
to be used infrequently. Elliott (2003) concluded that
many DT/A researchers have focused too exclusively on
the use of the approach for purposes of classification
and prediction. Diagnosing the nature of the child’s
underlying difficulties and using this to advise upon
intervention should, he argued, be the principal concern of those working in education settings.
Research might also further examine the relatedness of the concept of DT/A to that of RTI (e.g.,
Grigorenko 2009). RTI originated in attempts to find
the best way to educate children with learning difficulties by intervening as early as possible, examining their
responses, and following up with evidence-based
(group and individualized) forms of instruction. RTI
is different from DT/A in its aim: while DT/A focuses
upon the potential for learning of a child, RTI focuses
on prevention and identification: the sooner evidencebased intervention strategies are implemented the better. Although RTI and DT/A originated from different
backgrounds, a future research question regarding the
interaction or integration of both points of view seems
justified, because for both the underlying aim is to find
ways to realize children’s potential. Both involve
looking for children’s solving processes and adapting
instructional or interventional procedures to signs of
learning or change a child shows during testing.
It is hoped that dynamic measures can provide
educators with insights into the way in which individual differences in progression emerge, in the variability
or constancy in children’s use of, and change in, cognitive and metacognitive strategies, differences in the
ways they verbalize their solving processes, their reactions to failure and success, and how they react to the
provision of assistance. An important question for
future research is how such information can be used
to guide the work of teachers and other professionals.

1057

D

1058

D

Dynamic Visualization

To describe or prescribe the positive and weak
aspects of a child’s functioning requires not only highquality, standardized tests with sound norms but also
understanding of, and sensitivity to, the unique problem-solving behaviors of children during testing. Particularly important is the design and deployment of
increasingly sophisticated adaptive, scaffolded feedback
procedures. High-quality tests can be seen as instruments which enable us to measure objectively, reliably,
and transparently, an individual’s behavior. Findings
from these should lead to descriptions of individual
functioning and associated recommendations for intervention that can be understood by all who are seeking to
help the testee. There is still a long way to go, and
achieving successful approaches that can be employed
across multiple learning contexts will require ongoing
collaboration between researchers, clinicians, and
practitioners.

Dynamical Analogies
▶ Dynamic Modeling and Analogies

Dynamical System
▶ Approximate Learning of Dynamic Models/Systems

Dynamically Capable
Organization
▶ Learning Organization

Cross-References
▶ Assessment of Learning
▶ Diagnosis of Learning
▶ Feedback and Learning
▶ Intelligence and Learning

References
Elliott, J. G. (2003). Dynamic assessment in educational settings:
Realising potential. Educational Review, 55, 2003.
Elliott, J. G., Grigorenko, E. G., & Resing, W. C. M. (2010). Dynamic
assessment. In B. McGaw, P. Peterson & E. Baker (Ed.) The
International encyclopedia of education (3rd ed.), 3, 220–225.
Amsterdam: Elsevier.
Grigorenko, E. L. (2009). Dynamic assessment and response to intervention: Two sides of one coin. Journal of Learning Disabilities,
42, 111–132.
Haywood, H. C., & Lidz, C. S. (2007). Dynamic assessment in practice:
Clinical and educational applications. Cambridge: Cambridge
University Press.
Sternberg, R. J., & Grigorenko, E. L. (2002). Dynamic testing: The
nature and measurement of learning potential. New York: Cambridge University Press.
Swanson, H. L., & Lussier, C. M. (2001). A selective synthesis of the
experimental literature on dynamic assessment. Review of Educational Research, 71, 321–363.

Dynamic Visualization
▶ Animation and Learning

Dynamics of Exploration and
Exploitation
▶ Learning Adjustment Speeds and the Cycle of
Discovery

Dynamics of Memory: ContextDependent Updating
ALMUT HUPBACH
Department of Psychology, Lehigh University,
Bethlehem, PA, USA

Synonyms
Memory modification; Memory reconsolidation

Definition
Memory updating is a phenomenon whereby, under
certain conditions, new information can be incorporated into preexisting memories. Various aspects of the
phenomenon have been described in cognitive psychology, but no unifying theoretical account exists that
explains memory updating and its triggers. In

Dynamics of Memory: Context-Dependent Updating

neuroscience, it has been shown that the modification of
a specific memory is dependent upon its reactivation. It
is assumed that through reactivation, a particular memory is transferred from a passive and stable state to an
active but fragile state, at which time it can then be
modified. In order for the memory to be maintained,
it needs to undergo a restabilization (reconsolidation)
process (cf. Hardt and Nader 2009). Spatial context, i.e.,
the environmental surroundings in which learning
takes place, has been identified as a crucial factor moderating memory updating effects.

Theoretical Background
Memory is not a static repository of learned facts and
experienced events. Rather, memory is dynamic and
reconstructive in nature. While relatively stable and
fixed core knowledge seems essential for successful
everyday functioning and high-level concepts such as
self-identity, the ability to forget outdated information
and to update memories in the light of new relevant
information is equally important. The study of memory modification and updating has a long history in
cognitive psychology, and includes such phenomena as
the influence of story schemes on the recall of folk tales,
false memory creation, the susceptibility of memories
to misinformation, and the misremembering of one’s
own previous answers when provided with the correct
solutions (hindsight bias). The ability of memory to
change seems not confined to a short time period after
initial learning (as traditionally assumed by the consolidation account), but can happen long after the memory was acquired. Findings in neuroscience
demonstrate that, in order to change old memories,
they must first be reactivated, i.e., brought from
a passive inactive state to an active state. The actual
triggers responsible for reactivation and subsequent
updating have rarely been studied in humans. In contrast, there has been great interest in the processes of
memory reactivation and subsequent need for
restabilization (reconsolidation) in the animal neuroscience literature. Recently, researchers have attempted
to bridge the gap between cognitive psychology and
neuroscience, highlighting parallel findings in both
fields. Hardt et al. (2010) propose that reconsolidation
processes could underlie many of the memory modification effects reported in cognitive psychology.
Memory, however, is not a unitary system.
According to one view, it can be divided into two

D

broad categories: explicit and implicit memories.
Explicit memories are consciously accessible, whereas
implicit memories can be expressed without conscious
awareness. Explicit memories include semantic and
episodic memories. Semantic memories contain
learned facts and world knowledge for which one can
commonly not recall the specific circumstances accompanying their acquisition. In contrast, episodic memory concerns the ability to remember events that can be
traced back in time and space. Thus, episodic memories
have a spatio-temporal signature. While the molecular
mechanisms accompanying updating are most likely
similar for the different memory subsystems, the circumstances that trigger memory reactivation are probably quite different. For instance, in fear conditioning
in rats (a form of implicit memory), a tone that was
concurrently presented with a foot shock during training later serves as a reminder, reactivating the fear
memory. The tone is intrinsically linked to the particular fear-conditioning paradigm; for other forms of
tasks and memory subsystems, other reminders will
be effective. The present chapter is concerned with
episodic memories, i.e., with an instance of explicit
memory. As stated above, episodic memories include
information about the place where the episode
occurred. Some argue that spatial context is
a superordinate memory cue: it provides a stable scaffold that allows for the integration of the various elements of an experience into a coherent event. That role
makes spatial context an excellent candidate for an
effective reminder. Returning to a place will most likely
reactivate the specific experience associated with that
place. However, analog to the fan effect or cue overload
principle, context should lose its predictive and
reactivating value when it is associated with many
different episodes. Thus, a highly familiar context
might not trigger the reactivation und updating of
memories.

Important Scientific Research and
Open Questions
Although there is extensive literature on the context
dependency of episodic memories, i.e., the effect that
memory performance benefits from contextual stability in comparison to context change, until recently, it
was unknown whether spatial context in itself would be
sufficient in returning memories associated with that
context to an active state, allowing for updating.

1059

D

D

Dynamics of Memory: Context-Dependent Updating

Hupbach et al. (2007) developed a paradigm to
study memory updating in humans that is based on
the animal reconsolidation work. In their paradigm,
participants learn a set of objects. Forty-eight hours
later, reminders are provided to some participants, but
not to others, and a second set of unrelated objects is
studied. Again 48 hrs later, memory for Set 1 is tested.
The results show that although a reminder does not
affect memory for Set 1, participants who receive
a reminder incorrectly intermix objects from the second
set when recalling Set 1, demonstrating that a reminder
can reopen a memory and allow new information to be
incorporated. It is important to note that the updating
effect is unidirectional, i.e., only affecting Set 1. When
asked to recall Set 2 in Session 3, intrusions from Set 1
into Set 2 are rare. Extending previous reconsolidation
findings in the animal literature, the study shows that
updating can be a constructive process, one that supports the incorporation of new information into old
memories. But what exactly triggers memory updating?
The reminder used by Hupbach et al. (2007) was multifaceted, consisting of the following three components:
1. Spatial context. Participants in the reminder group
learned the second list in Session 2 in the same
room in which they had learned Set 1 in Session 1.
Thus, the spatial context could have served as
a reminder reactivating Set 1. Participants in the
no-reminder group learned List 2 in a different
room.

2. Experimenter. For participants in the reminder
group, the experimenter was the same in Sessions
1 and 2. Therefore, the experimenter could have
served as a reminder. For participants in the noreminder group, a different experimenter administered the procedure on Day 2.
3. Reminder question. Participants in the reminder
group were asked to describe what the experimental
procedure of learning Set 1 right before learning the
second list. For participants in the no-reminder
group, the experimenter did not ask what had happened in Session 1.
In order to disentangle these three components,
Hupbach et al. (2008) manipulated them independently: one group studied Set 1 and 2 in the same
spatial context, another group worked with the same
experimenter in Session 1 and 2, and a third group was
asked a reminder question in Session 2 before learning
Set 2. Interestingly, only the spatial context triggered
updating: when Set 1 and Set 2 were learned in the same
room, Set 2 objects intruded into Set 1. The other two
groups showed very little intrusions (see Fig. 1).
This shows that memories associated with a spatial
context are automatically reactivated when participants return to this context. Memories can then be
updated by incorporating new information. When
brought to a novel context, new learning is attached
to a different scaffold, i.e., an entirely different episode
is created.

50
Mean percentage of objects recalled

1060

Recall (Set 1)
Intrusions (Set 2)

40
39.6
30

20

34.6
31.8

20.6

10
4.5

0
Context

Experimenter

6.0
Question

Dynamics of Memory: Context-Dependent Updating. Fig. 1 Mean percentage of objects correctly and falsely recalled
in the different reminder groups. Error bars represent standard errors of the means. Note: Subjects were asked to recall
objects from Set 1. Objects that were falsely recalled from Set 2 are labeled as intrusions

Dyscalculia

This research raises several important questions.
For example, what aspect about the spatial context
triggers memory reactivation and permits subsequent
updating? It seems to be the actual presence in the
spatial context, since neither briefly revisiting the
spatial context, nor recalling it in a new environment, is sufficient for inducing this form of memory
malleability. Future studies have to reveal whether
the context effect is tied to the repetition of the
perceivers perspective of the context (egocentric
coding in which locations are represented with
respect to the particular viewpoint of a perceiver),
or whether the perceiver-independent overall layout
of the context is sufficient (allocentric coding in
which locations are represented in reference to
other locations and are thus independent of the
perceiver). Given the importance of the hippocampus for both the allocentric coding system and episodic memories, one can hypothesize that allocentric
coding carries the effect.
Are there boundaries on the ability of spatial context to act as an agent for episodic memory change?
Familiarity of the spatial context constitutes one such
boundary. In both studies mentioned above, participants were tested in a room they had never visited
before, hence the context was unfamiliar. Familiar contexts are associated with a variety of different experiences. Given the risk of overall modification, it would
be counterproductive to reactivate all experiences associated with a particular context each time it is revisited.
Hupbach et al. (in press) explored the role of context
familiarity for memory updating in a sample of 5-yearold children. Using the paradigm outlined above, they
tested children either in an unfamiliar setting, or they
visited the children at home in their familiar environment. When the context was unfamiliar it triggered
memory updating, thus replicating the findings with
adults. Critically, when tested at home, the spatial context did not cause memory updating, supporting the
hypothesis that a familiar context does not
serve as a reminder for a specific episode. Interestingly,
the experimenter and the reminder question - the
two reminder components that did not contribute to
the reminder effect in an unfamiliar context – became
effective in a familiar context. One can conclude
that memory updating following reactivation is
a relatively ubiquitous phenomenon, but reactivation
is initiated by different cues depending upon the

D

specific situation. While an unfamiliar spatial
context appears to “overshadow” other cues, those
cues can initiate memory updating in familiar spatial
contexts.
Determining which reminders are effective in which
situations will not only be of theoretical importance
but will also have important practical implications. In
some situations, such as in eyewitness testimonies, the
goal is to prevent a modification of the memory for the
witnessed event. In educational contexts, however,
memory updating is desirable, and defining the optimal conditions under which new information is incorporated into previous knowledge could be of great
value to educators.

Cross-References
▶ Adaptive Memory and Learning
▶ Memory Codes (and Neural Plasticity in Learning)
▶ Memory Consolidation and Reconsolidation
▶ Memory Persistence

References
Hardt, O., & Nader, K. (2009). A single standard for memory: The
case for reconsolidation. Nature Review Neuroscience, 10,
224–234.
Hardt, O., Einarsson, E., & Nader, K. (2010). A bridge over troubled
water: Reconsolidation as a link between cognitive and neuroscientific memory research traditions. Annual Review of Psychology,
61, 141–167.
Hupbach, A., Gomez, R., Hardt, O., & Nadel, L. (2007).
Reconsolidation of episodic memories: A subtle reminder triggers integration of new information. Learning & Memory, 14,
47–53.
Hupbach, A., Hardt, O., Gomez, R., & Nadel, L. (2008). The dynamics
of memory: Context-dependent updating. Learning & Memory,
15, 574–579.
Hupbach, A., Gomez, R., & Nadel, L. (in press). Episodic memory
updating: The role of context familiarity. Psychological Bulletin &
Review.

Dyscalculia
▶ Mathematics Learning Disability
▶ Socioemotional and Academic Adjustment Among
Children with Learning Disorders

1061

D

1062

D

Dyscalculia in Young Children: Cognitive and Neurological Bases

Arithmetic learning difficulties; Developmental
dyscalculia; Math learning disabilities

because they do not depend on formal education and
can be assessed relatively early in life. Core number
abilities are thought to be innate and scaffold other
aspects of numerical cognition. The non-symbolic
magnitude representation system is often characterized
in terms of a mental or a spatial number line (MNL).
An intact MNL representation is hypothesized to support the development of cardinal and ordinal number
concepts, and deficiencies in it will likely affect the
development of the latter concepts (Butterworth
2005). A defensible definition of DD awaits a better
understanding of the long-term significance of individual differences in core number competences.

Definition

Theoretical Background

Dyscalculia (usually referred to as Developmental
Dyscalculia – DD) is a specific learning deficit associated
with difficulties understanding numerical and arithmetic
concepts. DSM IV suggests prevalence rates of 2% for
DD; however, more recent estimates suggest 6.5% or
above (Butterworth 2010). Children with DD have difficulty acquiring number concepts, exhibit confusion
over math symbols, lack an intuitive grasp of numbers,
and have problems learning and remembering number
facts. DD should be distinguished from acalculia which
is acquired later in life, often as a result of neurological
insult; however, DD is also thought to reflect neurological dysfunction. It may or may not be comorbid with
other specific or general intellectual difficulties (e.g.,
dyslexia, IQ). Rubinsten and Henik (2008) suggest at
least two DD subtypes can be defined – pure DD and
a comorbid DD/dyslexia form – both of which may
have different etiologies.
With some exceptions, diagnosis of DD, and ipso
facto its definition, depends on computation test performance, which means that a formal diagnosis is
delayed until after the beginning of formal education.
Moreover, the fact that a diagnosis is frequently based
on an arbitrary cut-point on standardized test performance (e.g., below the tenth percentile) is of definitional concern.
Little is known of DD’s origins or its manifestation
in the infancy or preschool periods. There is much
speculation about whether pre-symbolic abilities are
early indices of DD. Some researchers claim that DD’s
origins lie in core number deficits (i.e., deficits in nonsymbolic approximate magnitude and/or small quantity representations). These abilities are of interest

The term “developmental dyscalculia (DD)” was first
invoked by Kosc in the 1970s to characterize a range of
children’s arithmetic difficulties, but not their underlying cause. In the 1980s and 1990s, DD was conceptualized in terms of limitations in general cognitive
functions (working memory, semantic associations
etc.); however, since the mid-1990s DD researchers
have tended to appeal to domain-specific, neurological
explanations for differences in numerical cognition
(Nieder and Dehaene 2009). Some researchers question
domain-specific claims, and continue to argue instead
that DD reflects general processing deficits. Nevertheless, linking neurological areas activated by behavioral
measures has resulted in theoretical advances.
The possibility that arithmetic difficulties have an
organic basis was first suggested by Gerstmann in the
1930s who claimed that acalculia, finger agnosia, left–
right disorientation, and agraphia reflect a common
neurological insult (brain lesions in the angular and
supramarginal gyri near the temporal and parietal lobe
junction). Although the status of Developmental
Gerstmann’s syndrome is controversial, the same difficulties may be associated with DD. Functional neuroimaging confirms that specific brain areas are activated
for processing numerosities and continuous quantity
(Butterworth 2010). These areas are neuroanatomically
distinct from regions serving general executive functions (Nieder and Dehaene 2009). The horizontal
intra-parietal sulcus (HIPS) is activated by non-verbal
quantity representations, analogous to a spatial map or
MNL (Wilson and Dehaene 2007). The bilateral IPS is
involved in symbolic and non-symbolic magnitude
comparison and areas of the rTPJ are activated in

Dyscalculia in Young Children:
Cognitive and Neurological
Bases
ROBERT A. REEVE, JUDI HUMBERSTONE
Developmental Psychology, Psychological Sciences,
University of Melbourne, Melbourne, VIC, Australia

Synonyms

Dyscalculia in Young Children: Cognitive and Neurological Bases

small, precise number enumeration (subitizing). The
rIPS is activated by simple calculation tasks; however,
a much larger neural network is activated for more
complex calculation that involves the frontal lobe, especially on the left, and the left angular gyrus.
On the basis of twin studies, DD appears heritable.
Analyses of atypical genetic groups suggest a possible
locus on a part of the X chromosome, though this does
not mean that all cases of dyscalculia are necessarily
inherited or associated with the X chromosome.

Important Scientific Research and
Open Questions
Research Related to Behavioral
Outcomes
Behavioral indices of DD in young children include
poor (1) number sense, (2) visuospatial abilities,
(3) arithmetic fact retrieval, and (4) computation ability. Young children with a poor number sense tend to
be unable to apprehend small numbers of dots (n < 3)
without counting (i.e., are unable to subitize), and their
non-symbolic and symbolic magnitude comparison
RT signatures often differ from those of non-DD children. Both of these abilities are thought to be associated
with the acquisition of counting and computation
skills. Landerl et al. (2004), for example, found that
compared to non-DD children, those with DD (1)
exhibited atypical dot enumeration and number magnitude comparison abilities, (2) had similar working
memory, language, and IQ, and (3) were substantially
less accurate on simple addition and subtraction tests.
Whether non-symbolic magnitude representations
scaffold symbolic magnitude representations is
a matter of theoretical importance. Rousselle and
Noel (2007) compared the symbolic and non-symbolic
magnitude judgment abilities of DD children with and
without comorbid reading disabilities, and non-DD
children. The DD children performed more poorly on
the symbolic magnitude judgment task, but not the
non-symbolic magnitude task. However, Price et al.
(2007) found that DD children have an impaired
non-symbolic magnitude system. DD and non-DD
children judged which of two simultaneously presented
sets of squares was larger. The comparison sets were
either similar (one to three squares different) or dissimilar (five to eight squares different). Compared to
non-DD children, those with DD were less accurate,

D

had slower RT’s and steeper distance effects. Significantly, Price et al. found an association between nonsymbolic magnitude judgment and neuroanatomical
activation (fMRI). They also found that non-symbolic
magnitude abilities predicted arithmetic abilities (see
De Smedt et al. 2009 for similar findings). These findings suggest that, compared to non-DD children, those
with DD exhibit different non-symbolic magnitude
judgment RT signatures, neurological activation patterns, and poorer arithmetic abilities.
Relatively little is known about the emergence or
development of non-symbolic magnitude representations. We know that magnitude judgment abilities are
evident early in life and become more refined in infancy.
Izard et al. (2009), for example, showed that 2-day-olds
react systematically and respond to quantity differences
across different modalities and formats (sequential vs
simultaneous). And magnitude judgment ability
becomes more refined in the first year of life (Mussolin
et al. 2010). However, the long-term significance of
differences in early non-symbolic magnitude representation ability is unknown. Insofar as the symbolic
number system (Arabic symbols) is supported cognitively by the non-symbolic magnitude system, deficits
in the latter are likely to affect deficits in the former and
ipso facto the development of numerical cognition.
Magnitude representation ability may be manifest
in different ways. The observation that similar brain
areas are activated in calculation and manual hand
tasks (Butterworth 2010; Gracia-Bafalluy and Noël
2008), and that finger agnosia is associated with DD,
has led to the claim that finger representations may
serve as a link between non-symbolic and symbolic
magnitude representation (Fayol and Seron 2005).
Interest in core number abilities has tended to overshadow other equally important work on the development of children’s numerical cognition. Contemporary
interest in the origins and development of number
owes much to Gelman and Gallistel’s (1978) work.
They suggested that a preverbal counting mechanism
guides the acquisition of verbal counting and that,
learning to count involves, in part, mapping preverbal
numerical magnitudes onto the verbal and written
number symbols and the mappings from these symbols
to preverbal magnitudes. The relevant capacities for
arithmetic must be able to represent numerosities of
sets independently of the properties of the objects in
the set. Capacities for number must be able to establish

1063

D

1064

D

Dyscalculia in Young Children: Cognitive and Neurological Bases

the numerical equivalence of two sets through one-toone correspondence, and to distinguish transformations that do and do not affect numerosity. These
requirements constitute benchmarks against which to
evaluate theoretical accounts of the foundational
capacities supporting the development of arithmetic
(Butterworth 2010).
Recently, Hannula and Lehtinen (2005) have found
individual differences in preschoolers’ tendency to
spontaneously focus on number events (SFON), and
that these differences predicted counting development.
What underlies this tendency is currently unknown;
however, the possibility that differences in the propensity to orient to number events might be an index of
DD is of practical and theoretical significance.

Research Investigating Brain
Dysfunction
Price et al. (2007) findings suggest that neural differences exist in DD children. Reduced gray matter has
been observed in the right and left IPS in DD children
(Butterworth 2010). Activation differences in nonsymbolic number comparison in children have also
been observed in the right IPS, and symbolic abnormalities in the left IPS. The reason for these apparently
conflicting findings is not clear. One possible reason is
that the organization of numerical activity changes
with age, shifting from right dominance to left dominance as representations of numerosity becomes associated with language. Further, it is likely that residual
specialization in the two parietal lobes, with the right
specializing in subitizing and estimation, and the left in
symbolic processing and calculation undergo development (Butterworth 2010). Longitudinal studies that
combine neuroimaging with careful tests of basic numerical capacities may reveal different developmental trajectories depending on the locus of the neural abnormality
(Ansari 2010).

Open Questions/Issues
Research into the origins and development of DD is in
its infancy, and many issues have yet to be resolved.
Here we list four interrelated issues. First, what is the
relationship between differences in non-symbolic magnitude representation and the acquisition of symbolic
magnitude representation? Insofar as precise symbolic
number and computation abilities are scaffolded by
a preexisting, approximate, non-symbolic magnitude

system, it is important to understand how difficulties in
the latter might affect difficulties in the former. Second,
what is the nature of the changing developmental relationship between neurological mechanisms and
numerical cognition? Third, do different developmental pathways underlie the acquisition of mathematical
competence? Fourth, what is the relationship between
a diagnosis of DD and remedial or intervention practices? Enormous progress has been made in identifying
early indices of DD; however, we know next to nothing
about effective, theoretically defensible, intervention
practices.

Cross-References
▶ Development and Learning
▶ Developmental Cognitive Neuroscience and Learning
▶ Individual Differences in Learning
▶ Mathematical Learning Disability

References
Ansari, D. (2010). Neurocognitive approaches to developmental disorders of numerical and mathematical cognition: The perils of
neglecting the role of development. Learning and Individual
Differences, 20, 123–129.
Butterworth, B. (2005). Developmental dyscalculia. In J. I. D. Campbell (Ed.), Handbook of mathematical cognition (pp. 455–467).
New York: Psychology Press.
Butterworth, B. (2010). Foundational numerical capacities and the
origins of dyscalculia. Trends in Cognitive Sciences, 14, 534–541.
De Smedt, B. D., Verschaffel, L., & Ghesquière, P. (2009). The predictive
value of numerical magnitude comparison for individual differences in mathematics achievement. Journal of Experimental Child
Psychology, 103, 469–479.
Fayol, M., & Seron, X. (2005). About numerical representations:
Insights from neuropsychological, experimental, and developmental studies. In J. I. D. Campbell (Ed.), Handbook of mathematical cognition. New York: Psychology Press.
Gelman, R., & Gallistel, C. R. (1978). The child’s understanding of
number. Cambridge: Harvard University Press.
Gracia-Bafalluy, M., & Noël, M.-P. (2008). Does finger training increase
young children’s numerical performance? Cortex, 44, 368–375.
Hannula, M. M., & Lehtinen, E. (2005). Spontaneous focusing on
numerosity and mathematical skills in young children. Learning
and Instruction, 15, 237–256.
Izard, V., Sann, C., Spelke, E. S., & Streri, A. (2009). Newborn infants
perceive abstract numbers. PNAS, 106, 106–109.
Landerl, K., Bevan, A., & Butterworth, B. (2004). Developmental
dyscalculia and basic numerical capacities. A study of 8–9 year
old students. Cognition, 93, 99–125.
Mussolin, C., Mejias, S., & Noel, M. (2010). Symbolic and
nonsymbolic number comparison in children with and without
dyscalculia. Cognition, 115, 10–25.

Dyslexia
Nieder, A., & Dehaene, S. (2009). Representation of number in the
brain. Annual Review of Neuroscience, 32, 185–208.
Price, G. R., Holloway, I., Rasanen, P., Vesterinen, M., & Ansari, D.
(2007). Impaired parietal magnitude processing in developmental dyscalculia. Current Biology, 17, 1042–1043.
Rousselle, L., & Noel, M. P. (2007). Basic numerical skills in children
with mathematics learning disabilities: A comparison of symbolic vs non-symbolic number magnitude processing. Cognition,
102, 361–395.
Rubinsten, O., & Henik, A. (2008). Developmental dyscalculia: Heterogeneity may not mean different mechanisms. Trends in Cognitive Sciences, 13(2), 92–99.
Wilson, A. J., & Dehaene, S. (2007). Number sense and developmental dyscalculia. In D. Coch, G. Dawson, & K. Fischer (Eds.),
Human behavior, learning and the developing brain: Atypical
development. New York: Guilford Press.

D

1065

Dysgraphia
▶ Language-Based Learning Disabilities
▶ Socioemotional and Academic Adjustment Among
Children with Learning Disorders

D

Dyslexia
▶ Language-Based Learning Disabilities
▶ Socioemotional and Academic Adjustment Among
Children with Learning Disorders