Archive for the ‘Text Criticism’ category

Kutscher, EY. The Language and Linguistic Background of the Isaiah Scroll (I Q Isaa), Studies on the Texts of the Desert of Judah, 6. Leiden: Brill, 1974.

February 12, 2009

Among the original scrolls found at Qumran, the Isaiah scroll 1QIsaa may have been the most remarkable. This scroll, containing all 66 chapters of Isaiah (along with a fragmentary version 1QIsab) dates from the last few centuries before the Common Era, while the earliest known manuscripts of the Prophets were from 916 and 895 CE. Thus there was considerable enthusiasm that a text much earlier than the MT had been found. Kutscher, however, argues that a comprehensive linguistic analysis of the scroll shows that it is actually of a later text type than that reflected in the MT. Analyzing the linguistic anomalies in the scroll, Kutscher concludes that they reflect the Hebrew and Aramaic spoken in Palestine towards the end of the Second Commonwealth. From this he suggests that the Isaiah Scroll is actually a popularized text whose language has been emended for the semi-literate masses. Further, this language seems to reflect a situation of diglossia in the Hebrew of this time period, where a “standard” dialect used for liturgical reading coexisted with a “substandard” dialect reflecting more colloquial use.

The first clues that the text is of a later type than the MT come from the orthography. One indicator is the spelling of proper names. For example, Damascus is always spelled דרמשק in the scroll, in contrast to דמשק in the MT. External sources confirm that the MT spelling is the more ancient, while the scroll’s spelling can date no earlier than the last few centuries BCE. A second indicator of a late date is the increase of plene spelling with ו and to a lesser extent י.

In this period knowledge of Hebrew seems to have been decreasing as Aramaic became the colloquial language in Palestine. Knowledge of Classical Hebrew had already begun to wane at the beginning of the Second Temple period as reflected in the language of the later biblical books and the complaint by Nehemiah (13:24) that “their children spoke half in the speech of Ashdod, and could not speak in the Jews’ language…” In its place a new dialect was developing which by the 2nd century CE had become Mishnaic Hebrew.

Aramaic seems to have been the lingua franca of the Ancient Near East from an early point, however based on 2 Kg 18:26 (= Is 36:11) in the 8th century Aramaic was not yet understood by the average Judaean. Over time, Aramaic influence grew and ultimately it became the official language of the Persian Empire. In Syria and Palestine, Aramaic seems to have also become the language of the common people, displacing Hebrew and the other local dialects. This dialect is not Imperial Aramaic, however, but Middle Western Aramaic, which branched into Galilean Aramaic, Christian Palestinian Aramaic, and Samaritan Aramaic. Aramaic seems to have finally replaced Hebrew as the common tongue by the days of R. Judah the Prince (2nd century CE). Mishnaic Hebrew seems to continue as a spoken language during this period, but it had already become a “classical” language. The period of the Dead Sea Scrolls must reflect a time in which Hebrew and Aramaic were still competing for supremacy as the common tongue.

This situation seems to be reflected in the language of the scroll. Whether consciously or unconsciously, the scribe has updated the classical language to reflect the contemporary Hebrew dialect by replacing rare words and archaic forms. The plene spelling seems to be used to aid pronunciation, especially by differentiating Hebrew from Aramaic when confusion might arise. At the same time, heavy influence of Aramaic can still be detected. Thus the scroll seems to have been produced for an audience who still had some knowledge of Hebrew, but whose primary language was Aramaic.

The scribe seems to have often substituted more familiar roots for rare words. Kutscher acknowledges that in some cases the Scroll’s reading may be superior to the MT, but in general the comparative evidence suggests that the MT is most often the earlier. For instance, in Is 13:10 he emends הלל to אור ‘shine’. The root הלל is not known in either Aramaic or Mishnaic Hebrew with the meaning ‘to shine’. In Is 33:7 and 42:2 the root צעק is amended to זעק. This is somewhat surprising since both appear in the Bible with relatively the same frequency, but their distribution is not equal. While in the Pentateuch צעק is used almost exclusively, in Chronicles, Nehemiah, and Esther, it is זעק which is much more common. Thus זעק seems to have been much more common during the Second Temple Period.

The scribe also avoids archaic morphological forms. For instance in Is 22:2 the form הַחֹ֣מֹתַ֔יִם occurs with both a feminine and dual ending which the scribe emends to simply חומות. In Is 41:2 the nifal participle נִדָּ֖ף “driven away” is emended to נודף, the more common form in Mishnaic Hebrew for פ”נ verbs. The scribe often writes the hiphil infinitive without ה, as is common in Qumran Hebrew, such as לחיות for MT להחיות in Is 57:15. All the later dialects tend to substitute /o/ for /a/ in the imperfect forms of intransitive verbs, thus we have אפעולה for MT אֶפְעַ֖ל “I will work” (Is 43:13). This form also reflects the spread of the lengthened form of the imperfect from being strictly a cohortative to being used as a normal imperfect. This seems to have begun already in the later books such as Ezra, Daniel, and Nehemiah. Note, however, that this form did not continue into Mishnaic Hebrew.

The greatest difference between Mishnaic and Biblical Hebrew is the simplification of the verbal system. Mishnaic Hebrew no longer uses the infinitive absolute, and the infinitive construct is only used when preceded by ל. Accordingly, there is a marked tendency in the scroll to add ל to infinitives. Further, in Mishnaic Hebrew the preterite יקטל form as well as the wayyiqtol and weqatal forms have disappeared. Instead the יקטל form serves as modal and subjunctive, the קטל as the perfect, and the participle is used for the present and future. The language of Chronicles seems to already reflect this development. Note that literary Hebrew, however, preserved the waw-consecutive as late as the Maccabean Era.

Accordingly, in the scroll we often see the later verbal system. For instance, in Is 11:8 the MT וְשִֽׁעֲשַׁ֥ע יוֹנֵ֖ק is emended to וישעשע יונק, the weqatal being replaced by waw + imperfect. In Is 12:2 וַֽיְהִי־לִ֖י is emended to היהא לי, the wayyiqtol being replaced by the perfect. Sometimes the so-called prophetic perfect is emended with an imperfect as in Is 11:9 where כִּֽי־מָלְאָ֣ה הָאָ֗רֶץ is changed to כי תמלאה הארץ. Is 6:4 contains an imperfect to indicate continued concomitant action, but the first part of the verse establishes the time reference as past, וַיָּנֻ֙עוּ֙ אַמּ֣וֹת הַסִּפִּ֔ים מִקּ֖וֹל הַקּוֹרֵ֑א וְהַבַּ֖יִת יִמָּלֵ֥א עָשָֽׁן “Then the foundations of the threshold shook at the voice of the one calling, while the house filled with smoke”. Here the scroll uses a nifal perfect והבית נמלא עשן.

While the language is updated, the scribe also seems to have attempted to avoid ambiguity between Hebrew and Aramaic features in order to aid the reader. For instance, we constantly find the spelling לוא rather than לא for the negative particle to emphasize the Hebrew // against Aramaic //. Similarly is the spelling יואמר to indicate Hebrew /yōmar/ against Aramaic /yēmar/, etc. We also find cases where the scribe seems to have avoided genuinely Hebrew roots because of their similarity to Aramaic roots. For example, in Hebrew both הן and הנה are presentative particles, but in Aramaic הן = אם ‘if’. Thus the scribe has consciously avoided the use of הן to limit confusion.

However, the influence of Aramaic on the language is impossible to avoid completely. Aramaic influence is particularly strong with nouns and pronouns. For example עלוהי for עליו “upon him”, and Aramaic גופן /guwpnā/ for Hebrew גפן /gεfεn/ “vine”. Influence is also seen in the verbs, for example מהסיר for Hebrew מסיר reflects Imperial Aramaic in which the ה of the hafel appeared in the participle.

Another example is the 2fs form קטלתי instead of קטלת. This form appears in the Bible as a Ketib and perhaps once in the Song of Deborah where it is usually taken as an archaism. Kutscher argues, however, that here the form is an Aramaism. In fact, it occurs primarily in Jeremiah and Ezekiel, two late books under Aramaic influence. This seems to be an interesting phenomenon that is somewhat regular, early forms that were common to Hebrew and Aramaic have fallen out of Hebrew, only to be reintroduced at a later stage from Aramaic, where they have been conserved. This seems to hold true for the 2fs pronominal suffix כי- as well.

Lastly, there are several features which suggest a situation of diglossia in the Hebrew of the time period. A higher “standard” dialect seems to have been maintained for the liturgical reading concurrent with a colloquial “substandard” dialect. For instance, the scroll has several cases of the imperfect pausal form when the position in the clause does not warrant it, such as אכרותה for וְאֶכְרֹת in Is 37:24, etc. Such “pausal” forms are also found in good Rabbinical manuscripts as well as the transliterations of Origen and Hieronymus, and Palestinian Christian Aramaic where it must be a result of Hebrew influence. This suggests that forms such as תקתולו were actually accented penultimately in the colloquial Hebrew of the Second Temple period. Thus, while the synagogue readings retained the ultimate accent, it was gradually replaced by penultimate in the colloquial speech.

Along these same lines are the form of the 2ms perfect and 2ms pronominal suffix. The transcriptions in Origen’s Secunda reflect the Hebrew forms קָטַלְתְּ and דִּבָרָךְ in contrast to the MT forms קָטַלְתָּ and דִּבָרֶךָ. Kahle argued that the former are the original Palestinian forms, the latter the creation of the Tiberian Masoretes under Arabic influence. However, Bergsträsser pointed out that we must differentiate between Biblical Hebrew and Mishnaic Hebrew in this regard, as the suffix /-āk/ is only found in Mishnaic texts. Kutscher suggests that the two types of Hebrew existed side by side. The standard form was /-kā/ which was carefully preserved in the liturgy, while the colloquial form was /-āk/ (apparently influenced by Aramaic). Like Qumran Hebrew in general, the scroll shows many cases of the כה- personal suffix and תה- verbal ending, demonstrating its age. It would not be surprising, however, that the Jews who read for Origen’s transcription did not use (and perhaps did not even know) the standard pronunciation.

A final example comes from the segholates, particularly the *qutl forms which become qitol in Tiberian Hebrew. The LXX transliterations reflect either qotel or qotol, which also seems to be the pronunciation in the scribe’s dialect. In contrast, the Secunda (3rd century CE) reflects the form qotl exclusively without an epenthetic vowel, while Hieronymus (4th -5th century CE) reflects only a form qotel. The chain qutl > qotol > qotl > qotel hardly seems possible. In fact, the Secunda form qotl is barely different than the original *qutl. Thus Kutscher suggests that again we are dealing with different dialects. The form in the Secunda is more ancient than that in the LXX, even though it is chronologically later. It is hard to tell which was the standard form, but most likely it was the LXX – Hieronymus – TH form qotel > qitol.

The linguistic structure of the scroll clearly points to the end of the Second Temple Period. The most telling characteristic is the use of ה in the hifil participle. Since this is a feature of Imperial Aramaic that has fallen out of use in all the later dialects the scroll seems to date no later than the 1st Century BCE. Since the main purpose of the plene spelling seems to be differentiating the Hebrew pronunciation and forms for an audience who primarily spoke Aramaic, and since this seems to be the overall character of the scroll, Kutscher suggests that this is a “popular” version of Isaiah meant for the semi-literate masses. As such, it must date from a time when Hebrew was still more or less understood by said masses, which best fits the period before the destruction of the Second Temple. Interestingly, the Samaritan Pentateuch and Septuagint also seem to be “popular” texts.

How then can the MT, a much later text, be superior to the LXX, Sam P, and Isaiah Scroll? Kutscher argues that the same distinction is found between the Greek papyrii found in Egypt and the Medieval manuscripts. The difference is between popular texts used for home study and the standard text carefully preserved in the Temple and centers of learning. The MT is a descendent of the standard text in contrast to these popularized versions.


The Study of the Composition of the Pentateuch Part 2 – Challenges to the Consensus

October 29, 2008

1. Introduction. Start with Part 1 – The Documentary Hypothesis. The Graf-Wellhausen Documentary Hypothesis was the consensus view at the beginning of the 20th century and continued to dominate Biblical Studies until being seriously challenged in the 1970’s. It remains influential primarily among American scholars, especially those of the Albright/Cross-Harvard school. The first challenges came in the development of Tradition and Form Criticism, which shifted the object of study from the written sources themselves to the oral traditions behind them. 

2. Form and Tradition Criticism. Hermann Gunkel (1862-1932) was the first scholar to look beyond the written sources to the origin and purpose of the narratives themselves. He studied the stories of Genesis as individual units and emphasized their pre-written, oral history. Drawing on studies of European folklore, he saw Genesis as a collection of stories, Sagen, that had originated early in Israel’s history and been passed down over time. He noted that many of the stories seemed to have etiological and etymological origins, and their seemed to be common types of stories.

Form criticism seeks to identify these stereotypical patterns or forms within language, and then relate those forms back to a particular life setting, or Sitz im Leben. Tradition criticism studies the history of these oral traditions during their period of transmission. Form critical studies were also done by Hugo Gressman (1877-1927), who studied the Moses stories, and Albrecht Alt (1883-1956), who studied the laws distinguishing between case law and apodictic law.

Gerhard von Rad (1901-1971) built on Gunkel’s work, developing a comprehensive theory for the development of the Pentateuchal traditions. He suggested that there were two sets of traditions. One was the Exodus-Conquest (the entry into Egypt, slavery, Exodus, and taking of the promised land) which became the common heritage for all the tribes and was celebrated at the Feast of Weeks. These elements are represented in Dt 26:5-9. The second tradition was the giving of the law at Sinai. These two traditions were then united to form a unified theology of history, or Heilsgeschichte. He also suggested that the pre-Deuteronomical works dated to the 10th century, during a period of cultural enlightenment at the time of Solomon.

Martin Noth (1902-1968) further refined von Rad’s themes, arguing that there were originally separate traditions belonging to particular tribes or group of tribes. These traditions were pooled once the tribes united as a tribal league or “amphictyony”. He identified five such traditions: 1) The promise to the patriarchs; 2) The exodus from Egypt; 3) The wilderness wandering; 4) The revelation at Sinai; and 5) The entry into the land. This single work was called G for Grundlage or “foundation”. Thus J and E were separate works which drew on the common tradition of G, but also added their own material, and eventually were combined with J dominating. Noth also argued that Deuteronomy is a prelude to the Deuteronomistic History (Joshua-2 Kings), not an addendum to the Tetrateuch (Genesis-Numbers).  

3. Challenges to the Consensus. While Gunkel, von Rad, and Noth attempted to work within the framework of the Documentary Hypothesis, later scholars of the traditio-historical approach came into more tension with it.

4.1. Ivan Engell (1906-1964) and the Uppsala School. Several Scandinavian scholars argued for the predominance of oral tradition in the East (following Nyberg) and a late, exilic or post-exilic, date of composition for the Pentateuch. Most notable is Ivan Engnell (1906-1964), regarded as the founder of the “Uppsala School”. His most significant work was a tradition-history introduction to the Old Testament (1945). He was strongly opposed to the Wellhausen method and described JE and P not as written documents, but rather as strands of tradition. For Engell, the existence of doublets and repetitions reflected the process of oral transmission, not the existence of parallel sources. He also argued that the stylistic and theological characteristics which had been ascribed to the various sources were often arbitrary and artificial. While conceding that some forms were committed to writing early, he argued that the majority of the Old Testament was developed and composed at the oral level. He argued for a Tetrateuch which went through a long process of growth and expansion that was brought to an end by P in the post-exilic period. It should be noted that Engnell’s work was responded to by Mowinckel, who largely attacked his methodology, arguing for the need of both literary criticism and tradition-historical approaches. 

4.2. Rolf Rendtorff (1925-). Rendtorff was a student and then colleague of von Rad. Similar to Engnell, he argued in The Problem of the Process of Transmission in the Pentateuch (1977) that traditional source criticism and the traditio-historical method were incompatible. He saw the Pentateuch as largely a Deuteronomistic composition, while P was not a source but supplementary. He focused on the smallest units of tradition which were built by stages into larger blocks of material, and eventually combined at a late stage (early post-exilic) into a “history” by someone who already had the Deuteronomistic history. Rendtorff’s student, Erhard Blum, developed his thesis further and in more detail. He saw the priestly work as an attempt to correct certain elements of the Deuteronomistic theology. Further, he saw the impetus for the Pentateuch as a Persian demand for an official “Jewish law” within the Persian empire.  

4.3. John Van Seters (1935-). The tradition-history approach represented by Engnell argued for a late date for the written form of the Pentateuch, but also for the general reliability of oral tradition. However, in his study of the patriarchal narratives, Abraham in History and Tradition (1975), Van Seters argued that these stories better fit a date of composition in the late monarchic or early exilic period. His argued that the supposed Nuzi parallels to the social setting of the Patriarchs were forced and selective, while much better parallels exist in 1st millennium cuneiform literature. Similar conclusions were reached by Thomas Thompson in his The Historicity of the Patriarchal Narratives (1974). Van Seters also argued that the analysis of the process of oral tradition is too speculative, since we only have the written form. Instead, he suggested a new Supplementary Hypothesis following a model of historiography rather than theology. It was common for ancient history writers, such as Herodotus, to use folk traditions in order to fill in their works. He sees three main sources which are not parallel, but directly dependent on one another: D, J,  and P. Here, Van Seters turns the traditional dating on its head, suggesting that the Tetrateuch is actually an addition to Deuteronomy, written first by J in the exilic period and then supplemented later by P in the post-exilic period.

4.4. Hans Heinrich Schmid (   ). In 1976 Schmid published Der sogennante Jahwist which also called into question the early date of J. Schmid argued against von Rad’s notion of a ‘Solomonic Enlightenment’ and tried to show that J was actually heavily dependent on prophetic traditions and the Deuteronomic school. Thus he concluded that J should be associated with both. His student, Martin Rose, argued further that D was prior to J, in agreement with Van Seters.  

4.5. R.N Whybray (1923-1997). In his The Making of the Pentateuch (1987), Whybray presented probably the most complete methodological critique of the Documentary Hypothesis. In sum, he argued that its adherents recreate the hypothetical documents by simultaneously relying on the criteria of inconsistency and consistency. Sections are assigned to a specific document  based on the assumption that each document has internal consistency of language, style, and theology. However, the documents are distinguished within the final written form based on the criteria of inconsistency. If the redactors are not concerned with inconsistency in their final product, then how can we assume that the authors/editors of the individual documents were? He argued that consistency is a modern western literary value, not necessarily an ancient Near Eastern one. He tentatively suggests that the Pentateuch was the work of a single author, an antiquarian historian similar to Van Seters, who used many folk tales of his time to create a history, perhaps as a prologue to the Deuteronomistic history.

5. Conclusions. A major methodological flaw of the Documentary Hypothesis was its tendency to multiply sources. Early scholars attempted to escape this by moving beyond the written sources to the oral traditions. This had the interesting side effect of pushing the date of composition later, into the exilic and post-exilic periods. Further, while early scholars had great confidence in the historical reliability of the oral traditions, the studies of Van Seters and Thomas Thompson began to undercut that reliability. Thus, while the Documentary Hypothesis was the consensus view at the beginning of the century, by the end its exact opposite was gaining popularity – late composition by a single author.

The Study of the Composition of the Pentateuch Part 1: The Documentary Hypothesis

October 29, 2008

1. Introduction. The composition of the Pentateuch is not my main area of study, nor is it the usual subject of this blog. However, it is relevant to the issues of typology and chronology of Biblical Hebrew, and I am interested more generally in how “authorship” worked in the ancient world, how scribes worked with texts, etc. Most importantly though, the topic is on my Hebrew Bible comp. So here is a summary of some of my notes. It is a little long, but perhaps it will help you brush up on your own studies. And please, this being a weaker area of mine, if you feel I have misrepresented a scholar’s view or left a very important scholar out, please add a comment.  

2. Pre-critical Views. The traditional view of Jews and Christians has been that Moses, the prophet par excellence, was the author of the Pentateuch. However, already in the Middle Ages, Abraham ibn Ezra (c. 1092-1167) had delicately asked whether some portions of the Pentateuch were perhaps not authored by Moses. For instance, Genesis 12:6 reads: 

“Abram passed through the land to the place at Shechem, to the oak of Moreh. At that time, the Canaanites were in the land.”

The gloss mentioning that the Canaanites were in the land seems to imply that at the time this phrase was added, the Canaanites were no longer in the land, and thus Moses was unlikely to be the author of these words. During the period of the Reformation, Andreas Carlstadt (1486-1541) noted that it was unlikely that Moses authored the postscript to Deuteronomy narrating his death and burial in the third person, and therefore also unlikely that he actually compiled the five books named after him (De canonicis scripturis, 1520).   

3. The Rise of Science and the Enlightenment. With the advance of scientific enquiry in the sixteenth and seventeenth centuries, accompanied by the radical skepticism most closely associated with Descartes (1596-1650), these questions were asked anew and pushed farther. Scholars noted a variety of repetitions, apparent contradictions, and general “defects” which suggested that the books as a whole did not come from the hand of Moses, but perhaps some later author had made use of notes from Moses along with other sources. Thus Hobbes (1588-1679) in his Leviathan (1651) suggested that the Pentateuch as a whole is post-Mosaic, except for the portions which are specifically ascribed to him. Spinoza (1632-1677) also held that the Pentateuch was written by someone other than Moses, generations after his death (Theological-Political Treatise, 1670). Spinoza’s view of the Bible as a natural work which should be studied empirically and rationally would become programatic for scholars after him.

4. Source criticism. Spinoza had suggested that understanding the meaning of the Bible included understanding both its historical context and the nature of its transmission. Source criticism arose as a discipline to attempt to establish the sources of the Pentateuch (and other biblical texts). In its early days three main theories arose regarding the nature of the sources: the Fragmentary Hypothesis, the Supplementary Hypothesis, and the Documentary Hypothesis.

4.1. The Documentary Hypothesis. The idea was first proposed by Henning Bernhard Witter in 1711, and then independently and more thoroughly by Jean Astruc in 1753, that the two names used for God, the “generic” Elohim and the proper name YHWH, may be a key to distinguishing the two main sources used by Moses in the book of Genesis (note that Astruc was interested in demonstrating Mosaic authorship). Witter had noticed that Gen 1-2:3 exclusively refers to God as Elohim, while Gen 2:4-3:24 consistently uses YHWH Elohim, and that these two passages appear to be parallel accounts of the creation. Astruc concluded that Moses had two sources of Genesis, one that used the name Elohim and one that used YHWH, along with further fragmentary sources. He imagined that these were originally laid out in four parallel columns, but were later combined into one book producing the uneveness of the text. This laid the foundations for the Documentary Hypothesis, the idea that the Pentateuch was composed by combining and editing several narrative strands, each of which were complete and independent.

Witter and Astruc had only considered the book of Genesis in their studies, but Johann G. Eichhorn (1752-1827) applied their method through Leviticus in his introduction to the Old Testament (Einleitung in das Alte Testament, 1780-83). Then, W. M. L. de Wette (1780-1849) suggested that the writer of Deuteronomy was different from the rest of the Pentateuch based on both style and content. He further suggested, based on 2 Kings 22, that Deuteronomy dated to the days of King Josiah at the end of the seventh century BCE, an important conclusion because it finally anchored one of the sources in Israel’s history. This also expanded the Pentateuchal sources to three: E (for the source that used the name Elohim), J (for the source that used the name YHWH, Y being J in German), and D (for Deuteronomy). Lastly, Hermann Hupfeld (1796-1866) concluded that there were actually two sources that used the name Elohim, dividing them into E1 and E2. The new source would later be named P (for a Priestly writer) and was originally seen as the earliest source.

The standard version of the Documentary Hypothesis is most commonly associated with Julius Wellhausen (1844-1918) and his Prolegomena zur Geschichte Israels (1883, the updated 2nd Edition, published originally as Geschichte Israels, 1 in 1878). Wellhausen drew heavily upon his predecessors, while adding to their evidence.

Wellhausen also more explicitly connected the composition of the Pentateuch to the development of Israel’s religion. Following K.H. Graf, he suggested that P is actually the last source rather than the first. His argument assumed that the Pentateuch had been composed in a sequence which reflected the evolution of the Israelite religion from primitive beginnings to sophisticated and elaborate central cult, an idea introduced by Wilhelm Vatke (1806-1882). As mentioned earlier, D was connected to the discovery of the law book by Josiah which provided a historical reference point. He argued that P reflected a more developed cultic setting than J and E, which represent a religion similar to the stories found in Judges, Samuel, and Kings. Thus the legislation of P could not have been in effect until after the exile, while J and E were situated in the early monarchy. J was seen as the earliest source, reflecting a southern provenance from the time of David and Solomon (c. 950), which provided the basic story line for Genesis and Exodus. In contrast, E was a northern work from the height of the Northern Kingdom (c. 850). This sequence yields the familiar JEDP. As the P source is last, Wellhausen assumed that the Priestly author or someone from his circle served as final redactor.

4.2. The Fragment Hypothesis. While the Documentary Hypothesis became the standard theory, it was not the only theory of source criticism. The Fragment Hypothesis originated with Alexander Geddes (1737-1802) and was adopted by J. S. Vater (1771-1826). This approach suggested that the Pentateuch was composed from many fragments of varying lengths, rather than whole documents. Rather than begin with Genesis, these scholars began with the law codes. Vater suggested that the process of formation began with the law codes in Deuteronomy which dated to the time of David and Solomon, but were rediscovered in the time of Josiah.

4.3. Supplementary Hypothesis. Heinrich Ewald (1803-1875) was the main critic of the Fragment Hypothesis, arguing that it failed to account for the great unity of Genesis. Ewald is attributed with the origin of the Supplementary Hypothesis. He argued that the core of the Pentateuch is Elohistic, which gives it unity of theme and structure. This was supplemented by material from numerous other sources, which are responsible for the diversity of vocabulary and style.

5. Conclusions. Wellhausen’s presentation of the Documentary Hypothesis became the consensus position at the turn of the century. However, the methodology of source criticism tended to highlight differences over unity. As the sources were themselves subjected to scrutiny they began to multiply (ie E1, E2, E3…), basically devolving into a new Fragmentary Hypothesis. Thus, while there was agreement on the basics of the theory, there was less consensus on the details, leading to the dissatisfaction of some with the usefulness of the method.

Lieberman, Stephen J., “Response (to Blau),” Pages 20-28 in Jewish Languages, Theme and Variations. Edited by Herbert Paper. Cambridge, Mass: Association for Jewish Studies, 1978.

October 28, 2008

Lieberman makes a few salient points in response to Blau’s paper on the historical periods of the Hebrew language. First, he objects to Blau skipping from Middle Hebrew directly to Modern Hebrew, implicitly privileging spoken over written language. While spoken language indeed precedes written, the written can often achieve an independent status that is important in its own right. Thus, for example, it seems that the language of the Responsa of Maimonides should have a place in the development of Hebrew.

Second, he appeals to the use of sociolinguistics to better draw the boundaries between chronological periods, in analogy to dialect geography. Thus dialect boundaries are drawn based on the loose clustering of isoglosses (Lieberman aptly remarks that such boundaries should not be considered walls, but “seives”). These clusters are the result of cultural and historical factors, such as language loyalty. Further, we should be aware that there are a variety of levels of language available to an individual for different social situations. Thus changes in language due to historical factors are often turned into cultural or sociolinguistic differences.

Often languages are differentiated not on linguistic but literary grounds. That is, major written works become the standard for a given period. This is reflected in the standard division of Pre-Biblical Hebrew, Biblical Hebrew, Dead Sea Scrolls Hebrew, Mishnaic Hebrew, etc. These works have a great influence on subsequent linguistic development, but they may also mask the greater linguistic diversity of a period.

Lieberman thus suggests a scheme in which Early Hebrew represents pre-Exilic Hebrew (ie pre-6th century, and leaving aside the earliest stages of Hebrew which are not documented directly). Judean or Post-Exilic Hebrew extends from the time of the Babylonian Exile until the time when Hebrew ceased to be spoken (c. 200 CE). Scholastic Hebrew covers the period until the revival of the spoken language in the 19th century, which he labels Israeli Hebrew.

Within Early Hebrew, he prefers to divide between a prose and a poetry dialect which shows archaic features, rather than distinguish a separate period of Archaic Hebrew. Thus it is a sociolinguistic rather than a historical distinction. Further, he suggests that Early Hebrew should also be divided into a southern Judean dialect, which became prominent, and a northern Ephraimite dialect. 

In the next period, Judean Hebrew, he notes many sources of interference. Some of these are external to Hebrew, such as the influence of Aramaic, Persian, and later Greek and Latin. There also seems to be a situation of diglossia between a vernacular and a literary dialect which is reflected in later biblical books like Chronicles and Esther, as well as some compositions from the Judean Desert. The vernacular dialect is reflected in compositions which were originally oral within the Mishna, the Halakhic Midrashim, and other Tannaitic works.

In the Scholastic Period there were likewise two varieties of Hebrew in use: a dialect influenced by Early Hebrew, and the vernacular dialect as it was known from the Oral Law. Of course, there was again interference between the two, including the tendency to correct the vernacular to the more prestigious dialect.

Blau, Joshua, “The historical periods of the Hebrew language”. Pages 1-13 in Jewish Languages, Theme and Variations. Edited by Herbert Paper. Cambridge, Mass: Association for Jewish Studies, 1978.

October 28, 2008

In this short conference paper, Blau gives a summary of the historical periods of the Hebrew language with special attention to those features which have become the primary constituents of Modern Hebrew.

1. Pre-Biblical Hebrew (roughly 20th – 12th century BCE) is not well understood. Our only evidence is indirect in Akkadian and Egyptian documents. Even then, it is difficult to distinguish true “Hebrew” from “Canaanite”. 

2. Biblical Hebrew is attested predominately in the Bible, but also in some inscriptions and transcriptions (ie the Hexapla). The inscriptional evidence is limited by the use of a consonantal script, but we can make some inferences. For example, the spelling ין (presumably yēn) instead of יין (yayin) for ‘wine’ in northern ostraca suggests that monophthongization of /ay/ was more widespread in northern dialects than in Judea.

2.1. Archaic Biblical Hebrew is preserved mainly in poetry. It is marked by several features:

– long forms of prepositions (אלי elē, עלי alē, עדי adē)

– less frequent use of the definite article and the object marker את

– less frequent use of relative אשר 

– the ending / on nouns in the construct

– the pronominal suffix מו- -mō ‘their/them’

– the use of the construct form before prepositions

– the use of the shortened imperfect (preterite) as in Dt 32:8:

בְּהַנְחֵ֤ל עֶלְיוֹן֙ גּוֹיִ֔ם בְּהַפְרִיד֖וֹ בְּנֵ֣י אָדָ֑ם
יַצֵּב֙ גְּבֻלֹ֣ת עַמִּ֔ים לְמִסְפַּ֖ר בְּנֵ֥י יִשְׂרָאֵֽל

“When the Most High gave to the nations their inheritance, when he divided mankind, 
he fixed the borders of the peoples according to the number of the sons of Israel.”

2.2. Pre-exilic Classical Hebrew prose is the standard language of Biblical Hebrew.

2.3. Post-exilic prose shares features with Middle Hebrew and Aramaic. 

– preference for אני ‘I’ instead of אנכי.

– the attachment of pronominal suffixes to the object marker את instead of directly to the verb

– more frequent use of the participle in the verbal system

Note, however, that in a broader perspective, the differences in language within the Bible are quite slight. Blau attributes this to two things. First, Biblical Hebrew had a status as a standardized literary language. Second, in the transmission of the text, later hands have leveled out the language.  

The Masoretic text has three levels: the consonantal text, the vowel letters (matres lectionis), and the diacritical marks for vocalization and cantillation. Even the consonantal text underwent updating by the scribes such as the replacement of śīn by samek and the replacement of antiquated forms (for instance, the older תַּחְתֵּ֑נִי in 2 Sam 22:37, 40, 48 but תַחְתָּ֑י in the corresponding verses from Ps 18).

In the vocalization of the text, it seems that the Masoretes have tried to eliminate the older qal passive where possible. For instance, שרף ‘to burn’ normally occurs in the qal. However, when context demands a passive meaning, as a perfect it is vocalized as a pual שֹׂרָ֑ף while as an imperfect it is vocalized as nifal יִשָּׂרֵֽף.

Thus it is interesting that in morphology and phonology, Biblical Hebrew represents a late stage of the language (when compared to corresponding structures in Modern Arabic dialects in reference to Classical Arabic), but in syntax Biblical Hebrew is quite archaic (for instance, Blau points to the comparatively rare use of subordinated clauses). The implication is that the phonology and morphology were able to develop while the syntax was tied to the consonantal text.

3. Middle Hebrew (or Mishnaic Hebrew) seems to have developed from the vernacular of Judea after it was resettled in the 6th and 5th centuries BCE. In the rest of Erets Israel it seems that Aramaic was the vernacular at this time. After the Bar Kokheba revolt (132-135 CE), the rabbis moved to Galilee, bringing MH with them. However, as a spoken language it died out within one to two generations. Thus, the language of the Tannaim was based on spoken language, while that of the Amoraim is from a period when MH is no longer living and influenced more by Aramaic and Biblical Hebrew.

4. Modern Hebrew draws on both Middle Hebrew and Biblical Hebrew. Most of the phonological deviations in Middle Hebrew from Biblical Hebrew were leveled by copyists and printers in the Middle Ages so that they had little influence on Modern Hebrew. However, Middle Hebrew has influenced the morphology and syntax of Modern Hebrew such as in the lack of so-called waw-consecutive forms, the lack of the infinitive absolute and use of the infinitive construct only with prepositions, predominately -ל. Some Middle Hebrew features seem to prevail in Modern Hebrew because they are simpler, such as the use of של for the genitive which reduces the number of special construct forms that must be remembered. Both Middle and Biblical Hebrew vocabulary is used, sometimes with synonyms split for more specialized meanings. For instance, Biblical Hebrew יֶלֶד is maintained for ‘child’ and Middle Hebrew תינוק tīnōq, originally ‘child’ as well, is used for ‘baby’. 

This also highlights an interesting feature of Modern Hebrew. As a spoken language develops ‘naturally’, the various layers of the literary dialect become stratified chronologically. There is no reaching back into the older strata to derive new forms. However, because there is an almost eighteen hundred year gap between Modern and Ancient Hebrew as a spoken language, the older layers are stratified side-by-side and are all available for derivation. Thus none of the old forms are ever really dead. Of course, they are fused into a new unity and extended by new derivational patterns.

Young, Ian, “Late Biblical Hebrew and Hebrew Inscriptions” Pages 276-311 in Biblical Hebrew: Studies in Chronology and Typology. Edited by Ian Young. London: T&T Clark, 2003.

September 8, 2008

The extra-biblical evidence for the typology and chronology of biblical Hebrew comes largely from two sources – the Dead Sea Scrolls and the Hebrew inscriptions. Hurvitz and others have argued that the Dead Sea Scrolls show numerous features in common with Late Biblical Hebrew (LBH), while the inscriptions share much in common with Standard Biblical Hebrew (SBH), supporting the chronological distinction between the two and establishing the pre-exilic date of composition for Genesis-2 Kings. Young, however, argues that the fact that SBH has a closer relationship to the Hebrew inscriptions than LBH does not prove that it could not have continued to be written into the post-exilic period. Further, he argues that the links between SBH and the Hebrew inscriptions are not as strong as sometimes suggested.

First, the corpus of inscriptions is limited in several ways. The amount of linguistic material represented is less than one percent of the size of the Hebrew Bible, and the focus of the inscriptions  are altogether different than the biblical material. Thus for the great majority of features contrasted between SBH and LBH, the inscriptions provide no evidence at all. Further, the majority of inscriptions of significant length come from the last half century of the kingdom of Judah (c. 652-586). This would place them during the time of Jeremiah and Ezekiel. Jeremiah is generally considered SBH, but there are early signs of LBH. Ezekiel is generally considered to reflect the transition period between SBH and LBH. Thus we should expect these inscriptions to exhibit a mix of SBH and LBH features. Lastly, Young wonders to what degree the inscriptions, which presumably reflect the administrative language, are relevant to the discussion of biblical Hebrew, a literary language?     

The relevance of the inscriptions is further complicated by the fact that rarely does a feature X occur exclusively in SBH, only to be replaced entirely by another feature Y in LBH. Rooker’s work on Ezekiel identified 37 items that were characteristic of LBH, and only two are complete replacements of the SBH feature. In 10 cases the feature X in SBH is carried into LBH, but augmented by a synonymous feature Y. In the remaining 25 cases, both features X and Y are attested in SBH texts, but in LBH Y becomes proportionately more significant. Thus with such a small corpus, the appearance of X in the inscriptions may not necessarily be an indicator of a link to SBH, nor Y to LBH.

In the bulk of the paper, Young treats 23 features that seem to link the inscriptions to SBH as well as 27 features that may be characteristic of LBH. In the former case, he concludes that the links between the inscriptions and SBH are weak at best. The three strongest are the use of the infinitive absolute as imperative (Arad 1.2; 2.1; 7.2; 11.2), the locative use of זה in מזה (“from this place” = “from there”, Lachish 3.18), and the use of בעוד for “while still” (Siloam Tunnel 2).   However, these are not strong enough to suggest a “self-evident identity between the two corpora  (ie inscriptional Hebrew and SBH).  

As for the links with LBH, Young divides these into two categories – those which have oppositions in SBH and those which do not. Scholars have always admitted that individual LBH features may be found in SBH works, it is the accumulation of such features that marks the language as LBH. No such concentration of LBH features appears in any given inscription. 

Young also devotes a considerable amount of time to features in the inscriptional corpus which are rare or absent from biblical Hebrew. Here there is a significant amount of material which calls into question the simple equation of SBH with the inscriptions. Instead, Young suggests that the inscriptions should be seen as an independent corpus of Hebrew that sometimes shares features with SBH, sometimes with LBH, and sometimes with other types of Hebrew such as Ancient Biblical Hebrew (ABH) and Mishnaic Hebrew (MH).

In conclusion, while it is plausible that the language during the monarchic period was something similar to SBH, it does not prove that SBH could not also have been written in the post-exilic period. Indeed, the orthography of the extant biblical manuscripts implies that no manuscript exists unchanged from before the Persian period. This does not imply that they were composed during the Persian period, but it does raise the question of whether scribal intervention was limited to orthography, or if all features were subject to revision during the Second Temple period. Young notes that features such as whether מן is attached to or separate from a following noun with the definite article is just the sort of thing to be changed during scribal transmission.

Ehrensvärd, Martin, “Linguistic Dating of Biblical Texts,” Pages 164-188 in Biblical Hebrew: Studies in Chronology and Typology. Edited by Ian Young. London: T&T Clark, 2003.

August 22, 2008

In this essay, Ehrensvärd questions the notion that Late Biblical Hebrew (LBH) is merely a deteriorated and Aramaized version of Early Biblical Hebrew (EBH), due to the fact that post-exilic authors no longer knew how to write EBH. Rather, he argues that use of EBH may have continued into the post-exilic period, as suggested by its use in some of the post-exilic prophetic books, and that LBH may thus represent a stylistic choice by post-exilic authors. Thus, LBH should be treated as a separate dialect from, and that coexists with, EBH rather than a development from EBH.

He begins by summarizing some of the distinctions that scholars have noted between EBH and LBH. There are small but consistent differences between the two which indeed distinguish them as dialects. Further, Ehrensvärd concedes that the distinguishing features of EBH are comparable to the language of the pre-exilic inscriptions, while those of LBH have much in common with post-biblical texts such as the Dead Sea Scrolls. However, he also points out that there are also many differences between EBH and the pre-exilic inscriptions, as well as between LBH and Qumran Hebrew (QH).

The vocabulary of LBH is marked by the presence of a great number of Aramaic loanwords. Further, while such loans also occur in EBH, about 15 of the loans in LBH have come from Persian through Aramaic. None of these words of Persian origin are found in EBH. The morphology of LBH is also distinguished by the presence of some Aramaic-like forms.

It is the same also with syntax, LBH shows similarities to Aramaic such as the occurrence of the double plural construction in the construct chain, גבורי חילים ‘valiant men’, and the quivis construction (the repetition of a nominal as a distributed plural ‘all, each and every’) preceded by כל as in כל יום ויום ‘every day’. There are also differences in the verbal system: the temporal construction ב + inf const + PS is found much more often without ויהי, the past use of yiqtol is less frequent, qatal is used more commonly for the past tense and less commonly in its other functions, and periprhrastic use of היה + participle for cursivity is more common.

However, Ehrensvärd points out that in all these features, the difference between EBH and LBH is one of frequency. Each distinctive feature of LBH already exists in EBH and vice-versa, just to a lesser degree. The two possible exceptions are the use of the participle as a narrative tense only in LBH, as suggested by Mark S. Smith, and the use of the infinitive absolute as a command form only in EBH. Ehrensvärd disputes the first, suggesting that in these cases the participle is being used in a cursive sense (cursive aspect expresses a universal truth or an ongoing event or action). Thus he concludes that EBH is a typologically earlier form of the language than LBH, but that they are two separate dialects of the language. That is, LBH is not a failed attempt by post-exilic scribes to write in EBH.

Finally, Ehrensvärd points to Isaiah 40-66, Joel, Haggai, Zechariah, and Malachi as examples of possible post-exilic works that are written in EBH (though only in Haggai and Zechariah 1-8 is there a scholarly consensus as to a post-exilic date). The frequency of LBH-like features in these texts is small and comparable to what can be expected from a normal EBH text, and there are no indisputably LBH features. Since these are all prophetic books the use of EBH may be seen as a continuation of the classical prophetic style, but even so he suggests that it demonstrates that post-exilic writers could produce EBH texts.