Navigation – Plan du site
English Translations

The Study of Mishnaic Hebrew

Some Historical Milestones
Sophie Kessler-Mesguich
p. 136-152

Texte intégral

1The first grammar entirely devoted to Mishnaic Hebrew was published in 1844. Although the linguistic study of this period of the language is relatively recent, it constitutes today a particularly active field of investigation, in particular in Israeli universities. By placing this article in the perspective of the history of ideas in linguistics, I aim to show how the scientific concept “language of the sages” (lešon azal) was constructed, and to analyze the different descriptive methods that have been applied to it from the Middle Ages onwards.

Mishnaic Hebrew in Medieval Grammarians and Lexicographers1

2In the fragments of ‘Egron by Sa’adya Ga’on (first printed in 902) one can find a hundred or so words taken from the Mishna and the Tosefta, or about 10 % of the total number of entries. Sa’adya felt that the language of the Bible and that of the Mishna (which for him was a colloquial language) formed a single unit. This attitude is better understood in the light of Sa’adya’s anti-Karaite stances: to understand the written law, it is as impossible to disregard the linguistic information provided by the Mishnaic texts as to disregard the interpretation provided by the oral law in general. However it would be a mistake to reduce this attitude to simple needs of religious polemics, since it permeates and drives all of Sa’adya’s linguistic works.

3The Ge’onim also compiled lists of difficult words in the Talmud, within the framework of their rabbinical commentaries and in response to questions they were asked — the Geniza has yielded many fragments. Here I will mention two of the most ambitious lexicographic works, in which post-Biblical Hebrew is the topic in its own right and not a simple auxiliary to exegesis. The first, which is lost today and is known only through mentions made by Nathan ben Yeiel of Rome in his ‘Arukh, is the lexicon of difficult words in the Talmud written by ema ben Paltoy, who was the Ga’on of Pumbedita from 872 to 891; the second is the Kitāb al-āwī, a true dictionary organized methodologically (according to an anagrammatic classification of roots) and based on a specific corpus, that of the post-biblical literature from the Mishna up to the works of the Ge’onim2. The existence of these two works shows that the vocabulary of the Mishnaic language was studied autonomously, independently of the Bible.

4Among the successors of Sa’adya, attitudes as regards Mishnaic Hebrew were predominantly positive. The questions they were asked were both exegetic (Can one use Mishnaic words to clarify the difficulties of scriptural text?) and linguistic (Is Mishnaic Hebrew, with its morphological innovations, a corrupted form of the prestigious language of the Bible?). For instance, to those who rejected Mishnaic Hebrew because of the presumed irregularity of its forms, Ibn Ğana responded by stressing that the Bible itself is not free of grammatical anomalies; furthermore, he shows that denominative verbs such as taram, “offer” (from teruma)3 were constructed by analogy on forms of the gebura type. Thus linguistic reasoning prompted him to accord Mishnaic Hebrew the same dignity as the language of the Bible, and as a consequence, to use it in his grammar and in his dictionary in a much broader and systematic fashion than Sa’ayda had done. Ibn Ğana is, parenthetically, one of the few authors to have had recourse to Mishnaic Hebrew not only for the lexicon but also as regards grammatical questions, such as for example the existence of nitpa’’al.

5The only authors who expressed a negative opinion on the Mishnaic language were Menaem ben Sarūq and his disciples. In their view, it was a faulty, incomplete language, inferior to Biblical Hebrew; hence, there was no reason to include it in a grammatical description. Menaem reserves the term “holy language” for the language of the Bible. In his eyes, a denominative of the taram type is a morphological monstrosity. However this position was marginal, including in the later tradition; even someone like Profiat Duran (early 15th century), a disciple of Ibn Ğana but conscious of the “degradation” of the sacred tongue over time, begins with the use of Aramaic in the Talmud and does not hesitate to take linguistic examples from the Mishna.

6One might expect that the Karaites’rejection of the oral law would lead them to take an identical position as regards the language of the Mishna and the Talmud; this however is not true for all authors. An analysis of loan words from Mishnaic Hebrew used by the Karaites shows that on this level, little differentiates them from their rabbinic adversaries. A recent dissertation4 has demonstrated that the Karaites knew the Mishnaic language and used it freely, in particular in their poetry, when the Bible did not contain the term they needed. Even in their most polemic writings, the Karaites rarely dealt explicitly with linguistic issues. The grammarian Abu’-l-Farağ (11th century) only refers once to a Mishnaic occurrence, but it would be hasty to draw a conclusion since most of his work (Kitāb al-Muštamil) is unpublished or lost. The attitude of David ben Abraham al-Fasi (second half of the 10th century) is one of the most positive: he mentions in his preface “the words which only appear once in the Bible and that are found in the language of the sages” (luga al-‘awayl) and, in the dictionary itself, refers to them more than eighty times, above all to clarify the meaning of rare words and Biblical hapax legomena.  Much more than for their explicit value judgments, it was by their actual use of linguistic forms belonging to Mishnaic Hebrew that one can judge the Karaites as being less hostile to rabbinic language than to their religious beliefs.

7Thus the medieval grammarians were far from having a negative opinion of Mishnaic Hebrew. However they never studied Mishnaic grammar for itself, for several reasons:

8(a) An ideological reason: only Biblical Hebrew deserves to be described. In fact, the first Jewish grammarians were highly influenced by the Arab grammarians, from whom they borrowed their theoretical concepts and their terminology; like them, the believed the language of their scriptural sacred text to be perfect. The Bible, for Jewish grammarians, thus naturally takes on the prestigious level occupied by the Koran for their Arab predecessors and contemporaries.

9(b) A technical reason: The Biblical text is the only one to have a robust reading tradition, thanks to systematic and complete vocalization. The absence of textual certainty still raises problems today in certain cases for researchers who base themselves on partially vocalized manuscripts or on oral reading traditions; a fortiori it must have seemed difficult to medieval writers to construct a grammatical theory from irregularly vocalized texts.

10(c) A reason of priority: recourse to Mishnaic text, for authors who were not specialists in Halakha but (and this includes many of them) who were scholars, was connected to the need to interpret the Biblical text. Mishnaic Hebrew is considered to be one tool among many to understand the divine word expressed in the scriptural text. Linguistic occurrences attested in the Mishna which made no contribution to the understanding of Biblical Hebrew were never examined – with the notable exception of the ‘Egron, destined for poets.

11This latter point makes it clearer why Mishnaic Hebrew was, at that time, dealt with almost exclusively from the lexicographic point of view. Occasionally a rare or difficult word from the Bible was used in Talmudic sources, in a context which made it identifiable or comprehensible; in this case the grammarians turned willingly to it. But no author envisaged making Mishnaic Hebrew the subject of a grammatical description. Those who, after Ibn Ğana, devoted works to post -Biblical Hebrew, exclusively wrote dictionaries; in the early 12th century, Nathan ben Yeiel of Rome (Arukh, Talmudic and Mishnaic dictionary); in the mid-13th century, Tanhum Yerushalmi (on the vocabulary of the Mishneh Torah); in the 16th century, Elie Levita (Tishbi, a lexicon of rabbinical words not explained by his predecessors). This was also the case for Christian Hebrew specialists. One of the first to stress the importance of the rabbinic texts and an interest in the post-Biblical language was the royal reader G. Genebrard (1537-1597), whereas other Hebrew speaking scholars of the 16th century judged it to be “barbaric” (S. Münster) and “mutilated” (A.R. Chevalier). The languages of the rabbinical literature (Hebrew and Aramaic) emerged in the following century in the lexicographical works of Buxtorf.

The Haskala Period

12From a linguistic point of view the debate on Mishnaic Hebrew centered on two issues: Can one use it to enrich the Hebrew language? Was this language spoken, or was it only an artificial rabbinic jargon? As of the end of the 18th century, speakers of Hebrew indeed claimed that Hebrew was spoken by the Tannaim and that the language of the Mishna should be the basis for a revival of the language. However, these claims only extended to the lexicon. The first Maskilim were confronted with the problem of the paucity of Biblical vocabulary; but to the extent that they were not concerned with speaking the language that they were re-creating, they did not need to deal with issues of phonology or morpho-syntax. In the area of syntax, it is clear that they referred to the Biblical norm.

13The role played by Mendele Moykher Sforim in the process of appropriation of Mishnaic Hebrew by the Maskilim during the second half of the 19th century deserves note. It is often said that he was the first writer to use a “synthetic” language; i.e. to have introduced a certain amount of Mishnaic Hebrew into Biblical Hebrew. But although it is true that he was one of the first to have made literary use of this “combined” language, he in no way invented it; in fact, he extended the trend of literary creation in Hebrew that existed from the first prayer texts that combined Biblical and post-Biblical elements. However, with a few rare exceptions, writers did not study Mishnaic grammar; it was not before the middle of the 19th century that the first grammar of the Mishna appeared, in German.

14The role of Mishnaic language in the revival of Hebrew was a topic of heated debate. As of 1892, Yosef Klausner (1874-1959) emerged as the vigorous defender of Mishnaic Hebrew as the only possible source. Placing himself resolutely in a linguistic perspective, he showed that Mishnaic Hebrew is the final stage of development of a living language. Biblical Hebrew was always a literary language and became a dead language. Mishnaic Hebrew in contrast was spoken, and only a historical accident halted its development. It would be logical, according to Klausner, to take up the language at it was when it ceased to be spoken. In accordance with these ideas he published a Short Grammar of Modern Hebrew (Tel Aviv, 1935), in which revival Hebrew is described in the light of the rabbinic language.5

15Ben Yehuda and the other pioneers of the language had a different opinion: Mishnaic Hebrew was not the prime source. Rather, all the sources, classic (Bible, Mishna, Talmud) and later (literature, sacred and lay poetry, commentary, responsa) could provide terms for the revival of Hebrew. As regards morphology and syntax, a few decisions were made by official bodies, but their scope was infinitely more restricted than for questions of vocabulary. The final outcome of this process left its mark in the title of the work published by A. Bendavid in 1951 (but written several years earlier): “Language of the Bible or language of the Mishna?” When a new edition came out in 1974, the “or” was turned into an “and”, and the question mark disappeared.

16Thus for the hundred years separating the beginnings of the Haskala and the revival of Hebrew as a spoken language, Mishnaic Hebrew found itself at the center of a linguistic and literary debate where in the end Biblical Hebrew maintained its prestige but the language of the Mishna emerged as a potential treasurehouse of terms and expressions likely to remedy the “paucity” of Biblical language. In the end it would be agreed that both strata are part of the classical heritage. At about the same time, a second scientific debate arose over the nature of Mishnaic Hebrew.

Mishnaic Grammar, from A. Geiger to M.H. Segal

17Geiger’s treatise, entitled Lehr-und Lesebuch zur Sprache der Mischnah was published in 1845 in Breslau. It had two parts, a manual and an anthology of annotated texts. The manual was relatively short: fifty-four pages, of which fifteen listed classical and recent texts which, either directly or indirectly, examine the language of Talmudic texts. The author states his purpose as of the first lines of the preface:

The pages that follow are an attempt to place post -Biblical Jewish literature, and above all the Mishna, in the framework of a scientific approach (wissenschaflicher Behandlung).

18Geiger’s work is not a full description of Mishnaic Hebrew. Believing that “the language of the Mishna is a continuation of Biblical Hebrew,” he only studied divergences between the two grammatical systems. He describes Mishnaic Hebrew as an emanation of Biblical Hebrew that evolved under the influence of Aramaic; from this point of view, most of his remarks were echoed in later grammars. He was always quoted in German dissertations from the turn of the century. But the prime impact of his work stems from several lines on page 1:

Thus, Hebrew ceased to be a living language; nevertheless, it remained, like Latin in the Middle Ages, a religious language of scholars during the period of the Second Temple, and during the two hundred years that followed. When scholarship vanished from Palestine and when Babylon took on the primacy, Hebrew was entirely swept away by Aramaic. The Mishna was written in this scholarly language (hebräischen Gelehrtensprache).

19This view was shared by G. Dalman, the author of a grammar of Galilean Aramaic (1905). However, immediately, Graetz responded to Geiger (in a review in the journal Litteraturblatt des Orients) by challenging the concept of “Gelehrtensprache”. He was followed by numerous authors including Luzzato (1846), Weiss (1867), and at the end of the century, Bacher (in the article on the “Hebrew Language” in the Jewish Encyclopedia).

The grammar of L. Dukes (1846)

20Entitled Die Sprache der Mischnah, it is 127 pages long. The introduction stresses that Biblical Hebrew was a means of communication (Umgangsprache) among scholars as Latin was in the Middle Ages; the similarity between Aramaic and Hebrew doubtless enabled everyone, even the lay person, to understand something of Aramaic, but the texts in Hebrew were only accessible to scholars. The first part of the volume is a study of the lexicon; the second describes the grammar in the traditional order (verb, noun, particle), the order that Geiger also follows. The terminological differences are minimal, but it is worth noting that Dukes in general prefers the German terms to the Latin ones used by his predecessor: Zeiten instead of tempora und modi, Zeitwort instead of Verbum, Hauptwort instead of Nomen...Aside from a few formal differences, Dukes’ grammar bears much resemblance to that of Geiger, both in the description of language features, centered on the lexicon and morphology (in particular the morphology of the noun) as in the view of Mishnaic Hebrew as a “Gelerhrtensprache.”

The grammar of Y.H. Weiss (Vienne, 1867)

21The volume has a German title (Studien über die Sprache der Mischna) but it was written entirely in Hebrew; the author states in the preface that he wishes to reach the public who chooses to read scientific works written in Hebrew. The Hebrew title testifies to the desire to be part of a modern and scientific perspective: “The laws of the Mishnaic language, the study of its principles, its grammar and its features according to the bases of modern philology.” The preface by Jellinek goes in the same direction, and contrasts scientifically-oriented Talmudic study with the traditional pilpul approach that disregards history and reason. Linguistically speaking, Weiss’ work provides only a few new features: it remains however in the history of research as the first work in Hebrew entirely devoted to the language of the Mishna.

The grammar of C. Siegfried (1884)

22This 90 page text, published by H. Reuther in 1884, is a return to the German tradition of grammatical description. The appearance of the phrase “Neuhebräischen Sprache” is worthy of note, and enjoyed a certain amount of popularity6. The history of the Hebrew language from the first centuries presented in the introduction draws on Geiger’s and Delizsch’s conclusions. The description of the “grammatical system” of post-Biblical Hebrew is also a continuation of previous works: since the linguistic structure of post-Biblical Hebrew is the same as for Biblical Hebrew, it suffices to describe the changes it underwent. In contrast, although there is still not a separate part on syntax, it is worth noting the greater attention to constructions: the degrees of comparison of the adjective, particles and the verbal system provide, in addition to a morphological description, an opportunity for more in-depth analyses of syntax than in earlier works. The corpus on which the description is based is extensive: Mishna, Babylonian and Jerusalem Talmuds, late midrashim.

The grammar of K. Albrecht (1913)

23A disciple of Strack, K. Albrecht dedicated his book to him (Neuhebräische Grammatik auf Grund der Mišna). The first difference as compared to all his predecessors is that Albrecht mentions that he worked on the German edition of the Mishna (Jost, Berlin, 1832-34), but that he occasionally used manuscripts — Cambridge, Kaufmann and Münich codex. Second difference: Albrecht writes after the publication of Segal’s article (1908), which marked the end of the first period of Mishnaic studies (the “Geigerian” period). He discusses the controversy whether Mishnaic Hebrew is a living or dead language. The book is divided into two parts (phonology/ morphology). The consonants are grouped according to point of articulation. The morphology is presented systematically (particles, verbs, nouns). As in Siegfried’s – Albrecht’s grammar is quite similar – the syntactic analyses are integrated into the morphological description. There are numerous examples, translated into German, but exclusively drawn from the Mishna, which the author points to as an innovation as compared to his model. The typical root is represented by QTL: the terminology is still highly Latinized.

24Highly influenced by Strack’s Biblical grammar and Siegfried’s work on the Mishnaic language, Albrech’s work differs on two fundamental points: (i) he describes the language of the Mishna alone, and (ii) he takes the main manuscripts into account. He can be seen as the direct forerunner of Segal’s grammar.

The work of M.H. Segal (1927 and 1936)

25Curiously enough, the entry in the Encyclopedia Judaica on Segal (1876-1968) deals almost exclusively with his works in Bible criticism, which today have been forgotten. Segal’s celebrity is linked to his works on the Mishna: first of all an article which was published in 1908 in the Jewish Quarterly Review and which attacks Geiger by a rigorous demonstration showing that Mishnaic Hebrew had to be a vernacular; secondly his Grammar of Mishnaic Hebrew, published in English in Oxford in 1927; lastly, a Hebrew version of this grammar, Diqduq lešon ha-mišna, published in 1936 in Tel Aviv. As has just been pointed out, Segal drew heavily on Albrecht, but he also made major innovations:

26(a) He clarified the relationships between Biblical and Mishnaic Hebrew. All the previous writers had built up their grammars on the linguistic differences between the two periods, connected solely from a diachronic point of view. Segal took a different approach. If Mishnaic Hebrew was a spoken language, if it represented a colloquial level of language, it would be natural for a large proportion of its vocabulary to be absent from Biblical texts. The relationship between the two corpora is no longer simply diachronic (as was said explicitly or implicitly by the authors in the earlier periods) but the two forms of Hebrew necessarily coexisted; thus there were not two successive states of the same language but rather two different synchronies, one reserved for literary composition and the other for day-to-day exchanges. Segal thus introduced the fundamental concept of dialectal variants within Mishnaic Hebrew itself, which would be abundantly cited and employed in the next generation.

27 (b) Segal’s Grammar is the first that assigns an important role (and an autonomous section) to syntax. In the syntax section he examines not only the changes within the verb system but also sentence structure. An interesting difference is worth noting here: in the English version of the grammar, the description of verb tenses, the definite article, questions of gender and number, appear in the third part (syntax); in contrast, in the Hebrew version, only sentence structure syntax (noun, verb, simple, complex) appear in the third part. Issues of gender, number, and determination are dealt with in the part on the noun: the tense system is studied along with verb morphology. This approach links Segal’s grammar to the Hebrew grammatical tradition, which does not separate the description of forms from the study of their uses. However, beyond this difference in presentation, the fact that issues of syntax are treated separately is worth noting; here again, Segal’s grammar marks the beginnings of a new approach, even though the Hebrew terminology he uses has aged considerably since it has neither the weight of the traditional terminology (the medieval grammars hardly dealt with syntax at all, and they do not have a term for “noun phrase” or “object”) nor the rigor of modern terminology.

28Segal’s grammar, the only one accessible to linguists who do not read modern Hebrew, has not yet been replaced7: nevertheless it was subjected to severe criticism from the moment it was published. It was accused of not making a difference between the language of the Tannaim and that of the Amoraim, to be based on editions and not on manuscripts, to overly discount the Aramaic influence (in reaction to Geiger’s theory, hence on ideological and not scientific bases), to rely on old-fashioned linguistic and lexicographic concepts, and finally to ignore Semitic languages. Some of these criticisms were excessive and were branded as such.8 Nevertheless, Segal’s work has a certain coherence. His prime concern (twenty years after the publication of his grammar!) was to demonstrate that Mishnaic Hebrew was not a scholarly jargon, but rather a colloquial, spoken language. By presenting a unified picture of this language and by reducing the impact of Aramaic, he lent additional weight to his arguments; to the extent that the manuscripts reflect a much more diversified, much more fragmented state of the language, using them would have on the contrary detracted from his theory. In any case, one must admit that seventy- three years after the publication of the first edition, Segal’s work, although outdated, nevertheless provides a global vision of the language which contemporary research, although more advanced in its goals and methods, is unable to this day to provide.9

29Segal’s grammar marked in addition the high point and the end of European research on Mishnaic Hebrew. As of the 1930s, research on Mishnaic Hebrew was concentrated in Jerusalem and in particular at the Hebrew University. There is practically a direct link. The Jewish theological seminary of Breslau, founded in 1854, the first rabbinical seminary to introduce a modern program in higher Jewish education, became, under the direction of Z. Frankel – the author of Darkey ha-mišna and the founder of Monatsschrift für Geschichte und Wissenschaft des Judentums – a model for many other institutes of the same type all over Europe, and in particular for the Berlin Hochschule. When, in 1924, the Institute of Jewish Studies was founded (one associated with the beginnings of the Hebrew University), we find professors from Berlin and the heirs of Frankel, including J.N. Epstein and E. Urbach. Through the intermediary of Epstein, the first director of the Institute, the Berlin school, i.e. the great German philological tradition, continued to exist in the Jerusalem school.

The Jerusalem School

Jacob Nahum Epstein (1878-1952)

30Born in Brest-Litovsk, educated in Berlin, well-versed in Semitic languages, J.N. Epstein taught from 1923 as lecturer at the Berlin Academy of Jewish Sciences. In 1925, after invitations from Magnes, he became professor of Talmudic philology at the newly-established Hebrew University of Jerusalem. He himself demanded this title, refusing to have the chair be associated with rabbinic literature: in a certain way, he also invented the field, and hence deserves the title of “father of exact Talmudic science” endowed upon him by one of his colleagues. A rigorous and intransigent scholar, his books reflect his personality: monuments of twelve hundred pages in small fonts, which ferret out all the textual variants of the Mishna and other Tannaitic and Amoraitic texts. The absence of an index, doubtless deliberate on the part of a man who feared that the University of Jerusalem would turn into the seminars open to the public which were flourishing in America, made consultation of these works somewhat awkward, but this lack was remedied over the course of later editions and they today still constitute an inexhaustible source of information on the text of the Mishna.

31Epstein wrote three major works (plus a large number of articles): en 1948, Mabo’ le-nosa ha-mišna (“Introduction to the text of the Mishna”), published in Jerusalem and reprinted twice; in 1957, Prolegomena ad litteras Tannaiticas, Tel Aviv; in 1963, Prolegomena ad litteras Amoraiticas, Tel Aviv. If these three monumental texts have modest titles it is because they were destined to be forerunners of a critical edition of the Mishna, which Epstein pleaded for as early as 1925 and which had yet to see the day.

32These works contribute much on the lexicographical level but little, by contrast, on the grammatical level. Epstein primarily attempted to isolate the different units which make up the Mishna, to compare versions of the text, pinpoint existing variations. In this painstaking labor, linguistic analysis plays an important part but is never conducted for itself but rather serves the goal Epstein set down -- to describe all the variants of the Mishnaic text. Through his role in University teaching10, his intensive use of manuscript sources, and the rigor of his approach, Epstein paved the way for a systematic description of the language of the Mishna.

Hanokh Yalon (1886-1970)

33Born in Galicia and influenced by the Haskala, he first taught in Lvov and then in Jerusalem in a seminar for professors (he never held a university position). He only wrote articles; in his criticism he was often pitiless. For instance, on two occasions, he argued that “if Segal had taken into account the modern scientific editions which are now available, whole chapters and many details would have needed to be rewritten.” He touched on all fields, and all periods of the language. He was commissioned to vocalize the Mishna for the Albeck edition (1952–1959); this work led to a volume entitled Mabo’ le-niqqud ha-mišna (“Introduction to the vocalization of the Mishna”), Jerusalem, 1964, in which he explains the principles he used and expounds on a certain number of difficulties requiring linguistic analysis.

34Like Epstein, Yalon appealed for the use of codex manuscripts as the basis for linguistic description. But above all, he was the first to ascribe importance to oral traditions (the way different communities read the texts). Prior to this, these traditions were considered to be of no interest, late, corrupt and error-laden because they were influenced by the vernaculars of these communities. By showing that these traditions preserved, on the contrary, extremely ancient features, by making parallels between the Yemenite tradition and the Babylonian tradition of vocalization, Yalon truly opened up a whole new research field; even though his own works were not based on the methods of modern linguistics he is considered to be real founder of linguistic research on Mishnaic Hebrew.

Saul Lieberman (1898-1983)

35Born in Belorussia, Saul Lieberman arrived in Jerusalem in 1928, and taught Talmud there in 1931. In 1940, invited by the Jewish Theological Seminary, he went to New York. A specialist in the Jerusalem Talmud and the Tosefta, the author of a critical edition and commentary, Lieberman also wrote a number of articles on the influence of the Greek language and civilization in Palestine. Although he was not a linguist, his work is of great importance on the lexicographic level (he dealt with roughly 1500 Greek and Latin terms in the Talmudic literature); in addition, his edition of the Tosefta helps approach this text on a solid footing. Like the two scholars mentioned above, S. Lieberman “laid the groundwork” for linguistic research based on indisputable textual evidence.

36The period between 1930 and 1950 was thus marked by three features: (a) the center of research on Mishnaic Hebrew shifted “from Berlin to Jerusalem”; (b) the initial foundations were laid for research aiming at establishing reliable texts, in critical editions presenting all the variants in the manuscripts, which were also published as reproductions; (c) greater attention was paid to the oral traditions of the communities. The mass of textual information thus collected provided researchers with new material, very different from what had been furnished by editions on which all the descriptions of Mishnaic Hebrew, including those by Segal, had been based. The archeological finds of the 1950s and 1960s provided still more data. The enormous efforts in philology between 1930 and 1950 could then be transmitted by linguists: the decisive thrust in this direction was provided by E.Y. Kutscher.

Eduard Yehezqel Kutscher (1912-1971)

37Kutscher only devoted the last ten years of his life to Mishnaic Hebrew. Before that he was primarily involved in Galilean Aramaic and had published his monumental study on the language of the Isaiah Scroll. His work on Mishnaic Hebrew took two directions: the lexicon and grammar. As regards the lexicon, Kutscher summarized his position in two articles11 in Hebrew, accompanied by detailed abstracts in English. In the first article, he demonstrates that the discovery of all the non-printed sources (epigraphic documents, manuscripts, Babylonian vocalizations, oral traditions) and advances in knowledge of other ancient Semitic dialects (Galilean Aramaic, Samaritan, Syriac, Punic) at least partially invalidated the lexicographic work of the 19th century (Levy, Jastrow, Krauss). The second article enumerates the conditions required to establish a new dictionary, and form its subdivisions: separate the language of the Tannaim from that of the Amoraim, differentiate what is Biblical from what is not (in particular in the case of homographs); forefront the differences between Hebrew in Palestine and Hebrew in Babylonia (the former was, in Kutscher’s opinion, superior to the latter), correctly evaluate Aramaic influence. Kutscher, who had (fairly unjustly) criticized Segal for describing an unreal language in the sense that he did not differentiate sources, falls at times into the opposite extreme, and outlines an investigation where everything must be strictly differentiated, and where no overall picture is possible until all the details have been accounted for.

38Here are the main points in his presentation:

39 (a) Mishnaic Hebrew is not a homogeneous unit. The Hebrew of the Tannaim prior to 200 and reflecting a living, spoken idiom must be differentiated from that of the Amoraim, who lived at a time when the language was no longer spoken and where the influence of Biblical Hebrew on the one hand, and vernaculars on the other, began to be felt. The Hebrew of the Palestinian Amoraim is not identical to the Hebrew of the Babylonian Amoraim; the language of prayer is a separate unit.

40(b) Within the language of the Tannaim, priority should be given to the Mishna, which provides the most reliable manuscripts.

41(c) The linguistic description should base itself exclusively on textual archetypes (‘ab-text in Hebrew, the Hebrew equivalent of the German Urtext), i.e. on very high quality evidence.12 Some ancient manuscripts (Palestinian in origin), the fragments of the Geniza, recognized as such, should be the topic of exhaustive monographs, which will serve as a basis for the new grammar of Mishnaic Hebrew. Kutscher himself put his method into practice, by proposing an in-depth study of the Kaufmann codex, and by asking his students to study other trustworthy sources.

42By appealing for disciples and research, Kutscher helped constitute a well circumscribed field of investigation, which now, a generation later, is solidly implanted in Israeli universities. On the other hand, by basing his work on clearly defined units from the spatio-temporal point of view, and by affirming the presence of dialects within Mishnaic Hebrew, he can be considered as the first to have released this language from the endless debate on its diachronic relationship to Biblical Hebrew and gave it a synchronic one. By doing so, he brought research on Mishnaic Hebrew into a modern scientific perspective.

New Horizons

43Kutscher’s epistemological and methodological choices were based on two suppositions: (a) the Hebrew of the Tannaim, which reflects a language which was still spoken, should be differentiated from the language of the Amoraim, since they consciously or unconsciously transformed the Hebrew texts they transmitted by introducing modifications linked to their own vernacular (Aramaic) or to Biblical Hebrew, the prestigious model; (b) to restore the language of the Tannaim, or in any case to come as close as possible to what it must have been, only one manuscript is used, the one considered to be themost faithful. These choices, which resulted in giant strides in the description of the Hebrew of the Mishna, had the side effect of excluding research on a certain number of manuscripts and denying any originality to the Hebrew of the Amoraim, in particular the Babylonians. Systematic research for ‘authentic and original’ forms of the language of the Tannaim ended up creating a paradox: any non-biblical form which entered into the grammatical system of the language of the Tannaim as it was represented was considered to be original, even if (or a fortiori) it was completely isolated. By contrast, phenomena characteristic of the language of the Amoraim, such as for example the form of the demonstrative plural hallalu, were considered to be ancient and Tannaic rather than as resulting from an internal evolution of Hebrew since it was believed that such an evolution was impossible during the era of the Amoraim.

44As of the 1980s, this somewhat rigid view that only the Mishna (represented by the Kaufmann codex and several other unquestionably reliable sources) reflected the language of the Tannaim, was radically altered. The typology of different Mishnaic manuscripts put forward by D. Rosenthal split into two branches (Babylonian/ Palestinian)13 and the one put forward by M. Bar-Asher for the distinction between two types (Western/ Eastern) within the Palestinian branch14 helped “save” a certain number of them and escape the narrow framework defined by Kutscher. The theory of linguistic types consists of ceasing to interpret certain variants as late corrections and to consider that they continue ancient traditions, some of which (identified after an in-depth critical examination) could reach back to the period when Hebrew was spoken. In other words, the changes observed in the manuscripts in the Babylonian tradition or the Eastern type of the Palestinian tradition should not be systematically attributed to Aramaic influence, the reintroduction of Biblical forms, or a scribe’s error, but can, in certain cases, constitute the trace of ancient Palestinian features, forgotten dialectal variants of which only the best manuscripts of the Palestinian branch have preserved exemplars. The concomitance between certain Babylonian occurrences (vocalizations and textual variants found in the Babylonian Talmud) and the forms transmitted by the Samaritan tradition prove that Babylonian and Palestinian phenomena had a common origin, and that this origin was the Palestine of the first centuries, where Hebrew was spoken with multiple dialectal variants.

45This theory led to an expansion of the field of research. There is now work on the Tosefta, on the piyyut, and researchers realize that these texts, considered by Kutscher to be insufficiently reliable, have in fact themselves preserved ancient forms, and can be integrated into a general linguistic description of the language of the Tannaim.15

The growth of research on the Hebrew of the Amoraim

46At the current time, the perspective on the Hebrew of the Amoraim is in a process of change. Some researchers, such as M. Mishor and Y. Breuer16 highlight phenomena which, although neither related to Biblical literature or Aramaic influence, attest according to them to an internal evolution in the language of the Amoraim – and hence perhaps to a spoken usage. This position is not shared by all: for instance Y. Naveh feels that no epigraphic document confirms the existence of such usage at the time of the Amoraim17:

47Is it possible for a language to continue to evolve when it is no longer spoken? To resolve this contradiction, we need to view the Hebrew of the Amoraim as a ‘living literary language’ to use the phrase that G.B. Sarfatti borrowed from Lifschitz. It is not certain that this debate can ever one day be resolved; to the extent that we are dealing despite everything with written and transmitted texts, there will always be cases for which it will be impossible to decide with absolute certainty whether a given phenomenon of the Babylonian tradition is a late correction or on the contrary the trace of ancient and authentic usage that only chance prevented from appearing in the evidence of the Palestinian tradition.

The importance of the Samaritan tradition

48The Samaritan Hebrew and Aramaic tradition was studied by Z. Ben-Hayyim, who devoted five volumes to it.18 The phrase “Samaritan tradition” should not be understood too narrowly. First of all, some of its features are found in the writings of Qumran – which cannot be suspected of Samaritan leanings! Secondly, the Mishnaic language is closer to the Samaritan tradition than that of Tiberias – whereas what the Samaritans rejected was precisely the oral law. Given this dual paradox, one is forced to admit that the tradition which is known to us today through this single community (hence its name) was in the past far more widespread in Israel. The Samaritan tradition provides useful data for Mishnaic research, in particular for rare forms, attested only in several manuscripts. The quality of the latter is not always sufficient to guarantee the authenticity of the forms in question (which could also come from copyist’s errors, which are present in even the best copies); in this case at times the Samaritan tradition confirms data in the manuscript. For instance the forms of the present participle nitpa’’al with nun instead of mem, which intrigued researchers, have parallels in the Samaritan reading of the Pentateuch and can hence be considered to be authentic despite their rarity.

The incorporation of oral traditions

49As we saw earlier, we owe to H. Yalon the discovery that the reading traditions of different communities held treasure troves of ancient and authentic forms. There is no doubt that his teaching in a seminar for professors from Sephardi backgrounds attracted his attention — he was above all interested in the Yemenite tradition; work on the pronunciation of the Jews of this community dates back to 1915. In the next generation, the reality of immigration, which brought numerous communities to Israel with their own reading traditions, opened up new perspectives for researchers and made anthropological, sociological and naturally linguistic work possible, beginning with a generalized collection of data published piecemeal and which can now be used advantageously. Here again, the initial impetus came from the Hebrew University of Jerusalem, in the form of the Language Traditions Project, launched in 1976 by S. Morag. Monographs describing the various oral traditions have been published in the series ‘Eda we-lašon, and its more than twenty volumes (in twenty years) testifies to the vitality of this field of research.


50Mishnaic Hebrew is today considered to be an autonomous state of the Hebrew language, the outcome of a dialect spoken as of the time of the First Temple, and whose distant echoes are found in certain Biblical forms. This dialect, which gradually distanced itself from the Biblical language, and was frozen as a literary language, was spoken in Palestine, alongside Aramaic and to a lesser extent Greek, up until the time of the Bar Kochba revolt. The disappearance of the intellectual centers of Judea ensured the victory of Aramaic as a spoken language, but at the same time Mishnaic Hebrew became a written language used to establish the Tannaitic corpora. These continued to be transmitted and taught by scholars who spoke Galilean Aramaic or Babylonian and for whom Hebrew was a language of prayer and study; however, alongside artificial changes due to the influence of the language of prestige (Biblical Hebrew) and the vernacular (Aramaic) one can find in the language of the Amoraim evolutions which do not appear to derive from either of these two phenomena; some use the term of “living literary language” for this second period of Mishnaic Hebrew.

51When Mishnaic Hebrew was spoken, it possessed numerous dialectal variants, many of which were partially eradicated from the manuscripts and systematically corrected in the editions. Current research, which focuses on the oldest and most reliable manuscripts, on epigraphic documents and on the oral traditions of the different communities, attempts to reconstitute Mishnaic Hebrew in its diversity. Naturally, the nature of the sources themselves creates numerous obstacles: the manuscripts are not all vocalized, the oldest of these were written seven centuries after the writing of the Mishna, and it is certain that the long oral transmission of the latter, even after the text had been set down in writing, led to certain modifications which made the characteristics of the spoken language of the Tannaim disappear forever. However, research has already generated promising results: monographs on a given manuscript, on a given tractate from the Mishna or the Talmud, syntheses on specific points of morphology or syntax, study of vocabulary, classification of reading traditions, recording and use of data provided by oral transmission, help highlight the factors of unity of Mishnaic Hebrew at the same time as those which contributed to its diversity. The systematic diachronic studies of the last century have given way to a synchronic vision of things: naturally the relationship of Mishnaic Hebrew to Biblical Hebrew is still defined, but the main thrust is to reconstitute the linguistic system as it is reflected in the different sources.

52The history of research on Mishnaic Hebrew is connected to the way in which it is viewed. With a few exceptions, the medieval grammarians, including the Karaites, were ready to use the Mishnaic corpus, but only to explain scriptural text. The science of grammar was considered, in its beginnings, as a tool for Biblical exegesis. In this perspective, it did not seem necessary to describe Mishnaic Hebrew. The first embryonic descriptions of Mishnaic Hebrew – in particular from a lexical standpoint – are only found in Christian authors, beginning in the 17th century, who were interested in reading rabbinical texts. In the framework of a history of the Mishna text, and in the perspective of the science of Judaism, Geiger provided the first autonomous description of Mishnaic Hebrew, but he refused to assign it the status of a spoken language. Dukes compared the linguistic situation of the Jews of the first centuries to the one he was familiar with. For over half a century, respectable but not very original works further explored the way paved by Geiger, and described Mishnaic Hebrew by placing themselves primarily from a diachronic point of view (similarities and differences with Biblical Hebrew). Hoping to demonstrate the authenticity of the language of Talmudic Judaism, Segal gave it a unified picture from the oldest texts (the midrash halakha) up to the most recent (texts by the Babylonian Amoraim). In doing so, he was assisted by the fact that he referred exclusively to editions, which in general “aligned” the grammar of Mishnaic Hebrew on the Masoretic vocalization of the Bible. In the next generation, the problem of the actual existence of Mishnaic Hebrew as a colloquial, spoken language was permanently resolved thanks to archeological finds, and it became possible to approach Mishnaic Hebrew in its diversity. In reaction doubtless to Segal’s  synthetic view, who based his description on all the possible sources through the use of editions, the point of departure of this new approach is the study of a single source (the Mishna) via a single manuscript exemplar. This scientific approach, which has made it possible to considerably clarify the data and to construct a coherent picture (if not a complete one) of Mishnaic Hebrew, has resulted in a diversification of research, which has developed both in a synchronic perspective (monographs on the manuscripts, study of the relationships between the contemporary Aramaic languages) and diachronic (relationships to Biblical Hebrew, evolution of the language of the Tannaim to that of the Amoraim, etc.). The existence of dialects has been put forward, but with great caution since the process of oral and then written transmission could have altered the image of Mishnaic Hebrew permanently; furthermore, the language of non-literary documents does not exactly match that of corpora destined to be taught and transmitted.

53This research is almost exclusively carried out in Israeli universities, and in particular the Hebrew University of Jerusalem. The study of classical Hebrew cannot be dissociated from the revival of the Hebrew language: Segal wrote his article in 1908, Ben-Yehuda began the publication of his great dictionary in 1910, and the importance of Talmudic sources in his project is well known. Elevated to the dignity of a source on an equal footing with Biblical Hebrew, Mishnaic Hebrew is perceived as one of the components of modern day Hebrew. At the time of the Wissenschaft des Judentums, Hebrew was a dead language, and the post-Biblical language was  at times called Neuhebräisch: it may have been the historical projection of this situation which supported the idea of a Gelehrtensprache, a non-spoken language of scholars. Today, in contrast, Mishnaic Hebrew is studied primarily by Hebrew speakers, and this research is closely tied to the major scientific projects of the Academy of the Hebrew Language. Thus it is not rare to find, in the journal of linguistics Le∆onenu, the same authors signing a scientific article on Mishnaic Hebrew and a chronicle. In addition, the philological works of the pioneering scholars trained in the Wissenschaft des Judentums, such as Epstein, linguistic studies have emerged in particular in the area of syntax, neglected by the first grammarians. The recourse to recent linguistic theories helps describe in a more satisfactory way certain features that resisted traditional descriptions, for instance certain surprising uses of tenses which can be defined on the basis of speech acts, and from the situation of enunciation.19 Finally recent analyses dealing with a range of linguistic features: stylistic, socio linguistic, pragmatic, discourse analysis, etc. are worthy of note.

54The forty years that have gone by since the inaugural lesson by Kutscher (1963) have thus been productive ones. It is hoped that a volume bringing together the conclusions of the vast amount of work carried out over these four decades will be published.

Haut de page


1 See on this topic the (unpublished) dissertation by N. Netzer, Mishnaic Hebrew in the works of medieval Hebrew grammarians, Jerusalem, 1983 (in Hebrew).
2 A. Maman, in S. Sharvit (ed.), Studies in Ancient and Modern Hebrew in honour of M.Z. Kaddari, Bar Ilan Univ. Press, 1999, pp. 235-249. This is a general presentation of the work; numerous details can be found in another article by the same author, which appeared in Tarbiz 69/3, pp. 341-421.
3 The term teruma is a noun on the tequla model, whose root is R.W.M. The gerund taram was produced from it, which yields a secondary root (T.R.M.) by integrating the preformative with the root. Abraham ibn Ezra and Maimonides were opposed to Menahem on this point and show that it was a natural evolution of the language and not an erroneous construction. This type of analogical derivation is extremely common in modern Hebrew.
4 O. Tirosh Becker, Rabbinic Hebrew handed down in Karaite Literature, Jerusalem [Ph.D. awarded in December, 1999].
5 See A. Saenz-Badillos, A History of the Hebrew Language, 1993, Cambridge University Press, p. 273.
6 In Joüon’s Biblical grammar (1923), neo-Hebrew is the synonym for post-Biblical Hebrew.
7 By a work of equal magnitude, i.e., one which encompasses all the linguistic features. However today there are important monographs on verb morphology (G. Haneman, The Morphology of Mishnaic Hebrew according to the Tradition of the Parma Manuscript, Tel Aviv, 1980; M. Bar-Asher, L’hébreu mishnique, études linguistiques, Louvain-Paris, Peeters, 1999) and on syntax (M. Azar, The Syntax of Mishnaic Hebrew, Jerusalem, 1995), as well as dozens of articles.
8 M. Bar-Asher & S. Fassberg (eds.). Studies in Mishnaic Hebrew, Jerusalem (Scripta Hierosolymitana 37), 1998, p. 11.
9 According to M. Bar-Asher, this type of grammar cannot be written until thorough investigations of the various manuscript sources and oral traditions have been conducted. Things have moved however in that direction since he wrote in 1980 that we are still far from writing a complete grammar of Mishnaic Hebrew and in 1999 that “The task of writing a new grammar of Mishnaic Hebrew is not within easy reach at the moment... Perhaps it would be best if we content ourselves in this generation with writing an introduction which would summarize what has been done to date.” He adds in a note that “This project has already been undertaken.”
10 In 1962, Kutscher stated: “It seems to me that today in these fields (and I include among them the study of Mishnaic Hebrew), there is not one scholar who was not directly or indirectly influenced by him. He, Yalon and Lieberman educated a whole generation in the philology of Mishnaic Hebrew.” (Lešonénu 26, p. 151). This article is, elsewhere, critical of a (posthumous) grammar of the Aramaic of TB, which Epstein was unable to complete.
11 Published in Archive of the New Dictionary of Rabbinical Literature, vol. 1, 1972, pp. 3-28 and 29-82.
12 Recourse to the ‘ab-text helps at least partially to resolve the textual problems connected to the fact that no corpus of the oral law has ever been set down in a canonical form, in contrast to the Bible, where the Masoretic text (vocalized) is authoritative.
13 D. Rosenthal, Misnat Aboda Zara, A Critical Edition with Introduction. Jerusalem, 1980.
14 M. Bar-Asher, Tarbiz 53, 1984, pp. 187-220 (in Hebrew).
15 For instance Y. Yahalom, who works on ancient piyyut, notes that these are texts composed three hundred years after the writing of the Mishna came to an end, and hence are much closer to it, chronologically than to the most ancient manuscript!
16 M. Mishor, “New Edition of a Hebrew Letter, Oxford, MS Heb. D.69 (P)” Lešonénu 53, 1989, pp. 215-264; Y. Breuer, Language Studies II-III, pp. 127-153 (in Hebrew, English translation in the work cited above note 8).
17 Y. Naveh, Lešonénu 56/4 (1992), pp. 301-316 and 57/1 (1993), pp. 17-38 (in Hebrew).
18 The Literary and Oral Tradition of Hebrew and Aramaic amongst Samaritans, Jerusalem, Academy of the Hebrew Language, 1955-1977 (in Hebrew, the last volume in English).
19 This was done by M. Mishor in his dissertation on the tense system (The Tense System in Tannaitic Hebrew), typed dissertation, Jerusalem, 1983, in Hebrew.
Haut de page

Pour citer cet article

Référence papier

Sophie Kessler-Mesguich, « The Study of Mishnaic Hebrew », Bulletin du Centre de recherche français à Jérusalem, 12 | 2003, 136-152.

Référence électronique

Sophie Kessler-Mesguich, « The Study of Mishnaic Hebrew », Bulletin du Centre de recherche français à Jérusalem [En ligne], 12 | 2003, mis en ligne le 30 août 2008, Consulté le 30 juillet 2014. URL :

Haut de page


Sophie Kessler-Mesguich

University of Paris 8

Articles du même auteur

Haut de page

Droits d’auteur

© Bulletin du Centre de recherche français à Jérusalem

Haut de page