THE KARAITE TRADITION OF HEBREW GRMIMATICAL THOUGHT IN ITS CLASSICAL FORM VOLUME 1 STUDIES IN SEMITIC LANGUAGES AND LINGUISTICS EDITED BY T. MURAOKA AND C.H.M. VERSTEEGH VOLUME XXXVIIIl THE KARAITE TRADITION OF HEBREW GRAMMATICAL THOUGHT IN ITS CLASSICAL FORM THE KARAITE TRADITION OF HEBREW GRAMMATICAL THOUGHT IN ITS CLASSICAL FORM A Critical Edition and English Translation qf al-Kitab al-Kan n al-Luga al-cIbraniyya by 'Abfl al-Fared Hiiriln ibn al-Farqj BY GEOFFREY KHAN MARiAANGELES GALLEGO JUDITH OLSZOWY-SCHLANGER VOLUME 1 BRILL LEIDEN . BOSTON 2003 This book is printed on acid-free paper. Library of Congress Cataloging-in-Publication Data Abu al-Faraj Harun ibn al-Faraj, II th cent. The Karaite tradition of Hebrew grammatical thought in its classical form: a critical edition and English translation of al-Kitab al-kilfi: fi: al-Iuga al'Ibraniyya by 'Abi:i al-Faraj Hariin ibn al-Faraj / by Geoffrey Khan, Mafia Angeles Gallego,Judith Olszowy-Schlanger. [Kitab al-kafi fi al-Iughah al-'Ibraniyah. English &Judeo-Arabic] p. cm. - (Studies in Semitic languages and linguistics; v. 37) Includes bibliographical references and index. Text inJudeo-Arabic and English; commentary in English. ISBN 90-04-13272-4 I. Hebrew language--Grammar-Early works to 1800. 2. Karaites-Early Works to 1800. I. Khan, Geoffrey. IT. Gallego, Mafia Angeles. Ill. Olszowy-Schlanger,Judith. IV Series. PJ4557.A282003 492.4'82-dc21 2003050297 ISSN ISBN ISBN ISBN 0081-8461 9004 13272 4 (Set) 9004133119 (Hd. 1) 9004133127 (Vo!' 2) © Copyright 2003 by Koninklijke Brill nv, Leiden, The Netherlands All rights reserved. No part qf this publication mqy be reproduced, translated, stored in a retrieval system, or transmitted in any firm or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items fir internal or personal use is granted by KoninkliJke Brill provided that the appropriate fies are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers MA 01923, USA. Fees are subject to change. PRINTED IN THE NETHERLANDS CONTENTS PREFACE ....... INTRODUCTION 1.0. 2.0. ................................................. IX ......... Xl )AbO al-Faraj Hartln and the Karaite grammatical tradition ............ Xl Al-Kitab al-Kafifi al-Luga al-Abii al-Faraj Hartin's grammatical theory and its background will be published in a separate volume. I became acquainted with al-Kitiib al-Kiifi for the first time when examining the manuscripts in the Firkovitch collections in the National Library of Russia in St. Petersburg during a research trip in 1993 sponsored by the British Academy and the Russian Academy of Sciences. I am grateful to these institutions for granting me the opportunity to undertake this research and should like to express particular appreciation to Dr. Olga Vasilieva, the curator of Oriental manuscripts in the National Library of Russia, for her warm hospitality. Photographic reproductions PREFACE x of the manuscripts in the Firkovitch collections were kindly supplied by the Institute of Microfilmed Hebrew Manuscripts in Jerusalem. I should like to thank also the Leverhulme Foundation, from whom I received a research grant that allowed me to work in collaboration with a Research Associate for three years as well as covering other expenses such as the purchase of photographs of manuscript and travel. It has been my privilege to work on this edition together with Judith OlszowySchlanger and Maria Angeles Gallego. Judith worked on the project for nine months (January-September 1999) and Maria Angeles for two years and three months (October 1999-December 2001). Judith made a preliminary draft of the edition of a large proportion of the text. Maria Angeles revised the draft by making a systematic collation of manuscripts and adding further chapters to the edition. I checked all the manuscript readings and located further manuscripts that filled any remaining gaps in the text. The interpretation of the text and its translation into English were my responsibility. Maria Angeles wrote the sections on the codicological features of the manuscripts and the scribal practices in the introduction to the edition, with some contributions on paleography being made by Judith. The rest of the introduction was written by me. I prepared the final camera-ready version of the work that was submitted to the publishers. Finally, I should like to thank Friedrich Niessen for proofreading the volume and preparing the index of Biblical quotations. Geoffrey Khan Cambridge, January 2003. INTRODUCTION 1.0. ~Abii al-Faraj Hiiriin and the Karaite grammatical tradition >AbO. a1-Faraj Hiirtin ibn Faraj lived in Jerusalem in the first half of the eleventh century C.E., where he was attached to a Karaite college (diir li-IJilm) that had been established by his teacher >AbO. YaCqub Yusuf ibn Nul).. This consisted of a circle of Karaite scholars, who devoted themselves to various fields of learning in addition to that of grammar, including philosophy, law, exegesis and Bible translation. >AbO. al-Faraj's contemporaries at the college included the philosopher Yusuf al-Ba~Ir and Levi ben Yefet, who was the son of the exegete Yefet ben CElL After the death oflbn Nul)., >AbO. al-Faraj took over the leadership of the college.! >Abu al-Faraj Hiirtin wrote numerous works in Arabic on the Hebrew language of the Bible. The largest of these is a comprehensive work on Hebrew morphology and syntax consisting of eight parts entitled al-Kitiib al-Mustamil Cala al-~U~al wa-l-Fu~al fi al-Luga al-Cfbraniyya ('The Comprehensive Book of General Principles and Particular Rules of the Hebrew Language'), which was completed in 1026 C.E? This consisted of eight parts, which may have originally been produced as separate ! Our source for most of these biographical details is the chronicle of Karaite scholars by Ibn ai-HIli, who was writing in the fifteenth century; see G. Margoliouth (1897: 433, 438-39). For the background of Ibn Nui).'s college, see J. Mann (1935· 33-34). For the Jerusalem circle of Karaite Bible translators see Polliack (1997). 2 For a summary of the contents of the al-Kittib al-Mustamil see Bacher (1895a: 232-256), who publishes a few short extracts. Recent studies of aspects of the grammatical theory of the work have been published by Maman (1996a, 1996b, 1997, 2001,2002) and Basal (1998,1999,2001,2002). XlI INTRODUCTION books. The work al-Kitab al-Kafi fi al-Luga alJlbraniyya (,The Sufficient Book on the Hebrew Language') was a shortened version of al-Kitab al-Mustamil. The earliest known manuscript of this work has a colophon dated 1037 C.E.3 Al-Kitab al-Kafi had a much wider circulation than al-Kitab al-Mustamil, judging by the large number of extant manuscripts containing the work. 4 We have a few fragments of two additional works that appear to be epitomes of al-Kitab al-Kafi. One of these is referred to by )AbO al-Faraj simply as al-Mulsta~ar Version') and the other was entitled Kitab alJUqud fi ('The Short Ta~arif al-Luga a!Jlbraniyya (,Book of the Pearl-strings on the Grammatical Inflections of the Hebrew Language,).5 In the introduction to Kitab alJUqud it is stated that this work is more concise than al-Mulsta~ar.6 A further work of 3 II Firk. Evr. Arab. I 4601, fo1. 107r. A note in the margin of fo1. 110r indicates that the manuscript was the property of the author's two sons, Faraj and Yehudah. 4 See Skoss (1928, introduction 11-27), Gil (1983, vo1. I, section 938, and the references cited there). Short extracts from al-Kitiib al-Kiifi have been published previously by M. N. Zislin (1962 [1.l5], 1965 [1.22.1-13]), N. Allony (1983 [1.21.18], [1.27.30]), D. Becker (1991 [1.22.1-27]), G. Khan (1997 [ILl6.12-15, II.17]; 2001 [1.1.1, II.8.7, 11.20.18, 11.4.20, II.4.22]) and Basal (2002 [1.25.53]). The fragment published by Allony (1983: 246) is incorrectly attributed to al-Kitiib al-Kiifi. 5 Fragments of Kitiib alJUqud were published by H. Hirschfeld (1922-23: 1-7). N. Basal (1997) has published some leaves that he identifies as coming from al-Mu!sta~ar of'Abii al-Faraj. The passages on grammatical particles published by S. Poznanski (1896b: 200-213) are shortened versions of ILl8 and II.19 of al-Kitiib al-Kiifi and so must also be assumed to come from either al-Mu!sta~ar or Kitiib al-CUqud. Other manuscripts containing an abbreviated version al-Kitiib al-Kiifi include II Firk. Evr. Arab. 12578 (18 fols.), II Firk. Evr. Arab. 12753 (1 fo1.), II Firk. Evr. Arab. I 2933 (1 fo1.), II Firk. Evr. Arab. I 2951 (2 fols.), II Firk. Evr. Arab. I 3368 (20 fols.), II Firk. Evr. Arab. I 3382 (2 fols.), II Firk. Evr. Arab. 13387 (1 fo1.). 6 T" T:m~ 1;:)N 1ln:J" 1ln:JN TN 7'NO 7NO' ... iJl77N 'I) 'I)N:J7N mln;:)N nJ:J 'i' 1':J''''N 1ln;:)"'N 'I summarized the book al-Kiififi al-Luga ... and somebody asked me to make another short version that is more concise than the aforementioned short version' (Hirschfeld 1922-23: 5). INTRODUCTION xiii 'Abii al-Faraj Hartin, which has survived in various manuscript fragments, is an introductory treatise on grammar entitled Kitiib al-Mad/sal Yi/ii Cllm al-Diqduq fi Turuq al-Luga alJlbriiniyya ('Book of Introduction into the Discipline of Careful Investigation of the Ways of the Hebrew Language'). According to the preface of this text, 'AM al-Faraj wrote it after his completion of al-Kitiib al-Mustamil al-Kitiib al-Kiifi. and 7 All these grammatical works of'Abii al-Faraj Harun are systematically arranged studies of the Hebrew language as an independent discipline. He, indeed, sometimes goes beyond a description of specifically Hebrew grammar and discusses general principles of language. In some sections of his works he addresses philosophical issues such as the origin of language and its nature. 8 Also extant are manuscripts of a grammatical commentary on the Bible known as Ta/sir al-YAlfii~ ('Interpretation of Words') that is attributed in the colophons to 'AM al-Faraj. The presentation in this work follows the order of the Biblical texts. It includes an Arabic translation of the Biblical verses and grammatical notes. 9 'AM al-Faraj also applied himself to a style of Bible exegesis that was only marginally concerned with grammatical issues. One surviving example of this is a reworking by 7 A large section of this work has been preserved in Il Firk. Evr. Arab. I 4601, fo1. 110a ff. Other extant manuscripts include: Il Firk. Ev. Arab. I 2549 (5 fols.), Il Firk. Ev. Arab. 12559 (2 fols.), Il Firk. Ev. Arab. 12609 (2 fols.), II Firk. Ev. Arab. I 2720, II Firk. Ev. Arab. 12827 (fols. 11-14), Il Firk. Ev. Arab. 12813, Il Firk. Ev. Arab. Il 285 (2 fols.). It is mentioned in a medieval booklist under the abbreviated title as al-Mad/sal fi al-Diqduq (Il Firk. Evr. Arab. 12524). 8 For the views of the medieval Karaites on the origin and nature of language see Olszowy-Schlanger (1997) and (1998: 87-97). 9 For this text see Olszowy-Schlanger (2001). A fragment of the text was published by Poznanski (1908: 55-67); cf. also British Library Or. 2499 fols. 1-21 (Margoliouth, 1899 no. 276). INTRODUCTION XIV )Abu al-Faraj of a commentary on the Pentateuch by )AbU Ya(qub ibn Nul). 10 We are indebted to )Abu al-Faraj mU-un also for one of the most important treatises on the pronunciation and cantillation of the Tiberian Masoretic tradition of Biblical Hebrew. This was know as Hidiiyat al-QiirP (,Guide for the Reader'), which was written in Arabic by the Karaite grammarian )AbU al-Faraj Harun." It was intended by )AbU al-Faraj to complement his work on grammar, al-Kitiib al-Mustamil and its shorter versions, which contained little description of the pronunciation of the language. He produced the work in a longer and a shorter version. As is the case with al-Kitiib al-Kiij'i, the shorter version of Hidiiyat al-Qiiri~ had a much wider distribution than the longer one judging by the number of manuscripts that are extant. 12 )Abu al-Faraj Harun was not the first Karaite grammarian. Works on Hebrew grammar had been written by Karaites before him. The works of his Karaite predecessors are briefly described here. This will allow us to assess )Abu al-Faraj's originality as well as his indebtedness to the earlier Karaite grammatical tradition. )AbU al-Faraj's teacher, )AbU Ya(qub Yusuf ibn Nul), the founder of the Jerusalem college and one of the foremost Karaite scholars of his age,13 wrote an Arabic work on Hebrew grammar that is referred to in the 10 The text is preserved II Firk. Evr. Arab. I 1754 (798 fols.). 11 For a detailed study of this text see Eldar (1994). 12 Extracts from the longer version are published in Eldar (1981, 1994). Most of the text of the shorter version that is known to be extant is published in Eldar (1987). 13 One source includes him in a list of scholars whom it describes as the 'teachers of Jerusalem'; cf. Mann (1935: 31). The other scholars in the list are his contemporaries Yefet ben (Eh and )Abii ai-Sum ibn Zuta. He is referred to in the colophon of another manuscript (MS II Firk. Evr. Arab. I 1754, fol. 105r.) as 'the prince Joseph' (ha-sar Yosep). INTRODUCTION xv colophons either simply as the Diqduq or as Nukat Diqduq 'Points of Grammar' .14 It is datable to the second half of the tenth century. )Abu al-Faraj Hartin mentions the grammatical work of Ibn Nul). in his writings. 15 Ibn Nul)., in turn, was heir to a tradition of Hebrew grammar that had developed among the Karaites ofIraq and Iran. This was brought to Jerusalem in the migrations of Karaites from the East during the tenth century. Ibn Nul). himself was an immigrant to Palestine from Iraq. This grammatical tradition will be referred to as the early Karaite tradition of Hebrew grammatical thought. 16 )AM al-Faraj Hartin continued some of the elements of this tradition, but was innovative in many ways, both in method and content. )AM YaAbu al-Faraj Harun attributes some grammatical concepts to the teachings of earlier Karaite grammarians in Iraq.25 The traditions of this earlier Iraqi school described by >AM al-Faraj correspond closely to what we find in Ibn NUl).'s Diqduq. Ibn NUl). was an immigrant to Palestine from Iraq, where he was, it seems, a pupil of the Iraqi circle ofKaraite grammarians. 22 )Eskol ha-Koper, 167, letter sin, 173, letter :jadhe. 23 Ed. Skoss, New Haven, 1936-1945. 24 E.g. II Firk. Evr. Arab. I 4323, fol. 9a: hiigii huwa maghab bacq. al-culamiP raJ:timahu alliih 'This is the opinion of one of the sages, God have mercy upon him', where the blessing raJ:timahu alliih suggests that the man in question is deceased. 25 E.g. al-Kitiib al-Kiifi: allagf gahaba )itii giilika ... min al-diqdaqiyyfna qawm min al-anic exegesis in the eighth and ninth centuries. This early tradition differed from the tradition based on the teachings of SIbawayhi, which became the mainstream school in Arabic grammar after the ninth century. It is relevant to note that Arabic grammatical thought in its early stages was closely associated with Qur>anic exegesis and only later became a distinct discipline. 42 This would parallel the association between grammar and exegesis reflected by the diqduqe ha-miqra list and also the fact that the Diqduq of Ibn Nii]:l has the structure of a Biblical commentary rather than a systematic description of grammar. Some of the technical terms that appear in the list of diqduqe ha-miqra and in Ibn Nii]:l's Diqduq could be calques of terms from the Syriac or even the Hellenistic grammatical tradition. It should be noted, however, that one should not necessarily attempt to trace the background of certain terms and concepts to one particular tradition. It is more likely that many ofthe Hebrew terms have their origin in the early Islamic period (7th-8th centuries) when the Near East was a melting pot of ideas. It was at this period that Arabic grammatical thought developed by drawing elements from the Hellenistic andSyriac traditions. Some Hebrew terms may have been borrowed by the masoretic and Rabbinic traditions from the Syriac and Hellenistic schools even before the rise ofIslam. In the present state of research, therefore, it would appear that Hebrew grammatical thought began to develop at about the same period as vocalization systems were developed and that the early Karaite grammatical tradition was not isolated from mainstream Judaism. By the time of Ibn Nii]:l, at the end of the tenth century, after Saadya had published his works on grammar, the Karaite tradition appears to have become distinct from what was followed by the Rabbanites. Before the time of Saadya, however, there is no reason to believe that there were any significant differences between Karaites and Rabbanites in this field. 42 See Versteegh (1993). INTRODUCTION xxv A few fragmentary texts are extant that are closely associated with Ibn Nul)'s Diqduq and belong to the early Karaite grammatical tradition. We have already mentioned a Judaeo-Persian grammatical commentary, which parallels the Diqduq both in grammatical theory and in structure. In addition to this, two early Karaite grammatical texts have been preserved that are not in the form of Biblical commentaries but rather are systematic classifications of the morphological patterns of Biblical Hebrew verbs and nouns. 43 These classifications represent the core grammatical thought that developed within the early Karaite discipline of diqduq. Unlike the works ofIbn Nul) and the anonymous Judaeo-Persian author, they are not concerned with the association between grammatical analysis and the exegesis of the meaning of the Biblical text. The text containing a classification of verbs is attributed to a certain Sacld. It is likely that this should be identified with the grammarian SaCld SIran, who is referred to in one source as a pupil of) Abu YaCqub Yusuf ibn Baktawaih. 44 As has been remarked above, Yusuf ibn Baktawaih is probably an alternative name of Yusuf ibn Nul). The text exhibits many parallels with the grammatical work of Ibn Nul), in its grammatical theory, terminology and argumentation. It consists of a series of chapters, each of which is devoted to verbs with imperative bases of one particular pattern. A full inventory is given of the verbs in each category, problematic issues are discussed and a complete paradigm of a representative verb is presented. In its overall structure, the treatise differs from Ibn Nul)'s Diqduq, which, as we have seen, consists of grammatical notes on the Bible arranged in the order of the Biblical verses. It, nevertheless, exhibits a similarity to the Diqduq in its method of discussing problematic issues. As is the case in the Diqduq, these discussions frequently offer a variety of different opinions concerning the 43 These two texts, together with the Judaeo-Persian grammatical text, are published in Khan (2000b). 44 Poznanski (1896a: 699), Steinschneider (1902: 89), Mann (1935: 30). INTRODUCTION XXVI derivation of a form. This applies especially where there is a problematic issue (mas~ala) concerning the derivation. The purpose of this method was to attempt to reach the truth by exploring many possible paths. The practice of presenting various views on an issue appears also to have had a pedagogical purpose. It encouraged enquiry and engagement rather than passive acceptance of authority. The main extant manuscript of this work contains a version of the text that has been elaborated by a second author, who refers to himself as al-mufassir 'the commentator'. The second text is a treatise consisting of a series of chapters that classify the nouns in Biblical Hebrew according to their morphological pattern. Each chapter is devoted to nouns of one particular pattern. It includes an inventory of the nouns with this pattern, a discussion of various issues relating to the inflection of the nouns and a sample paradigm of a noun in all its inflections. This layout is similar to what is found in the treatise on verbs. It is indeed possible that the author of the two treatises was the same. The technical terminology in the two texts is very similar. One cannot even discount the possibility that the two treatises were originally one work. Some. of the manuscripts of the treatise on nouns contain an elaborated form of the original with additions inserted by a second grammarian. )AbO al-Faraj Harun brought about a radical change in the Karaite grammatical tradition and his works clearly stand apart from those of the earlier Karaite grammarians. The perspective of his grammatical works differs from that oflbn Nul).'s Diqduq, the primary purpose of which was the investigation of Scripture by grammatical analysis rather than the study of the language per se. The comprehensive approach of )Abu al-Faraj to grammar contrasts with Ibn Nul).'s practice of concentrating on the problematic issues (masiPil, nukat). Another divergence from the approach of Ibn Nul). is the categorical approach of )AbO al-Faraj. He does not present alternative opinions as regularly as Ibn Nul). and where INTRODUCTION xxvii he cites these, he generally advocates one particular point of view, often by means of complex logical argumentation. )Abu al-Faraj refers to the grammarians of earlier generations such as Ibn Nul). as al-diqduqiyyuna. He did not use the term diqduqiyyuna to designate all people engaged in the study of grammar. He makes an explicit terminological distinction between the Arabic grammarians (al-nul:ui) and the early Karaite Hebrew grammarians (al-diqdaqiyyuna ).45 Moreover, the way he uses the term diqduqiyyuna in his writings implies that they were a set of scholars distinct from himself and that he did not regard himself as one of their number. He attributes some opinions to the earlier Karaite grammarians, for instance, by phrases such as 'the statement of the diqduqiyyuna that .... ', without qualifying the term diqduqiyyuna by adjectives such as 'other' or 'earlier' .46 The implication of this is that) AM al-Faraj regarded himself as in some way independent of these earlier grammarians. He considered, it seems, that the nature of his grammatical investigation was different. 47 One way in which this differed was that )AM al-Faraj undertook a systematic investigation of the language, whereas some of the earlier Karaite grammarians, such as Ibn Nul)., directed their attention only to problematic grammatical details of Scripture. Attempts were made already by certain circles of Karaite grammarians before the time of) Abu al-Faraj to systematize grammatical knowledge. This consisted mainly in 45 Cf. al-Kitab al-Kafi §II.l6.12 and also Khan (1997). 46 E.g. al-kalam fima yagkuruhu al-diqduqiyyuna fi al-'awamir 'Discussion of the statement of the diqduq scholars concerning imperatives' (al-Kitab al-Kafi, Part n, chapter 17); cf. also Khan (1997: 318). 47 On some occasions, however, 'Abii a1-Faraj uses the term didquq in the later sense of 'grammar' as an independent discipline, as he practiced it in his own works; cf. the title of the introduction to al-Kitab al-Kafi:: P'~' P"P"~ 0'37:1 hl'~ ~~ 'El :I~:I n"J~':l37'~ jJ;'''~ 'Chapter on the purpose of the study of grammar and the ways of the Hebrew language.' xxviii INTRODUCTION the classification of verbs and nouns according to their patterns and inflections. These treatises, however, lacked the scope of the grammatical works oPAM al-Faraj. A number of elements from the earlier Karaite grammatical tradition were incorporated into the works of)Abii al-Faraj. He makes a clear break with this tradition, however, by applying to Hebrew the theory of grammar that had been adopted by most Arabic grammarians of his time. This was the approach of the so-called Ba~ran school of Arabic grammarians, which had become the mainstream, classical tradition by the 10th century, but had had little impact on the earlier Karaite grammarians. The dependence of )AM al-Faraj on the Ba~ran tradition is seen in the scope of his works, in his grammatical theory and in his Arabic technical terminology. It is possible to identify close parallels between many passages of )Abii Faraj' s grammatical works and the extant texts of the Ba~ran Arabic grammarians upon which he was dependent, especially the compendia of grammar such as Kitiib aPU~alfi al-NaJ:tw ofIbn al-Sarraj, which had a wide distribution in the medieval Near East. 48 Indeed manuscripts are extant that contain transcriptions of such Arabic grammatical compendia into Hebrew characters for a Jewish readership.49 In many cases the discussion by )Abii Faraj of a grammatical issue has the form of a contrastive study between Arabic and Hebrew. He first describes a grammatical phenomenon in Arabic and then examines the parallels and differences in Hebrew. The grammatical theory that )Abii Faraj adopted from the Ba~ran tradition of Arabic grammar sometimes differed fundamentally from that of the early Karaite grammatical tradition. One example of this is his espousal of the view of the Ba~ran school that the derivational base of verbs is the infinitive rather than the imperative form. As we have seen, 48 ef. Basal (1998, 2001, 2002). 49 E.g. II Firk. Ev. Arab. II 290. INTRODUCTION XXIX the derivation of verbs from the imperative was a central feature of the early Karaite grammatical theory. Most of )AM Faraj's technical terminology is that of the Arabic Ba1?ran school. This differed in many cases from the terminology of the earlier Karaite tradition, much of which was Hebrew. In his work Kitiib al-Mad/!al ~ilii CIlm al-Diqduq fi Turuq al-Luga alJlbriiniyya (,Book of Introduction into the Discipline of Careful Investigation of the Ways of the Hebrew Language') )AM al-Faraj felt obliged to devote the opening chapter to a clarification of the terminology that was used by the earlier Karaite grammarians, presumably SInce his own grammatical terminology was very different. 50 The breach between )AM al-Faraj and the early Karaite tradition is shown by the fact that in some cases he appears to be unfamiliar with the precise usage of the early terminology. This is seen, for example, in his discussion of the terms "~,~ i~~'~ ('first imperative') and i~~'~ ')~n'~ ('second imperative') in chapter 11.17 of al-Kitiib al-Kiifi. The interpretation of these terms that he offers does not correspond to their usage in the extant early Karaite texts. This is all the more surprising since they occur extensively in the Diqduq ofIbn Nul)., who, according to the chronicler Ibn aI-Hili:, was the teacher of'Abu al-Faraj.51 Despite the fact that the methodology and grammatical theory of) Abu al-Faraj differed considerably from those of the early Karaite tradition, the declared purpose of his grammatical studies was the same as that of his Karaite predecessors. He explains in the introduction to al-Kitiib al-Kiifi that this was to ensure that Scripture was interpreted and read correctly. 50 II Firk. Evr. Arab. I 4601, fol. 1l0v.: Fa!!l fi gikr al-