Author:
Daniel Machiela
Search for other papers by Daniel Machiela in
Current site
Google Scholar
PubMed
Close
Open Access

The language of the Aramaic scrolls from the Qumran caves has grown into a topic of considerable scholarly debate. This discussion has been driven by multiple factors, but especially the following two: 1.) The relative dearth of Aramaic writings preserved from the Hellenistic and early Roman periods prior to the discoveries around the site of Qumran. With the Aramaic scrolls, we now have much better access to an otherwise poorly-attested phase of this language’s development. Tied up with this point are a number of areas of tangential interest, most notably the often related questions of “the language of Jesus,” the linguistic landscape of Second Temple period Judaism, and the historical development and use of the Jewish targums.1 Such interests have clearly impacted the amount of attention dedicated to the question of language in the Aramaic Dead Sea Scrolls. 2.) The Aramaic dialect of the scrolls has been an influential factor in dating them, as shall be seen below. These compositions neither contain dates indicating when they were written or copied, as the documentary texts from the Judean Desert sometimes do, nor do internal clues make them easy to situate in the history of Second Temple period Judaism. As a result, we are left to employ whatever methods and criteria are at hand for dating, and one of our best available options is the Aramaic dialect used by those who composed or copied the scrolls. Consequently, the language of these works may play a determinative role in how they figure into discussions of the historical, social, and religious landscape of early Judaism. To cite one prominent example, Kutscher’s dating of the language of the Genesis Apocryphon to around the turn of the Common Era, which was followed by Fitzmyer and others, influenced significantly how that text has factored into reconstructions of early Judaism and the broader corpus of Aramaic works from Qumran.

In what follows, I provide a survey of the history of research on what is often called Qumran Aramaic, going on to give my own account of the language of the Aramaic Qumran scrolls. The guiding questions behind this chapter are derived from those driving the book as a whole: To what extent can we speak of Qumran Aramaic as a coherent dialect, thereby supporting or undermining the notion of these texts forming an interrelated corpus? To what period can the dialect(s) of these scrolls be assigned, and with what level of certainty can we make that assignment?2 Finally, how compelling is the distinction – still commonly made by linguists – between Biblical Aramaic and Qumran Aramaic?

1 Yehezkel Kutscher and the Typological Method of Dating Aramaic Texts

No serious treatment of the language of the Qumran Aramaic literature can ignore the enormously influential, pioneering work of Yehezkel Kutscher on the language of the Genesis Apocryphon (1Q20).3 A doyen of ancient Aramaic and its dialects, Kutscher set the pattern for decades of researchers by placing different texts in a diachronic sequence based on the comparison of carefully chosen linguistic diagnostic traits. Kutscher and those who followed him assumed that these traits furnish the researcher with important clues for establishing the relative time and place of a text’s composition, and help to fill in a developmental sketch of the Aramaic language, from Official Aramaic (Reichsaramäisch) to the Jewish targums. Among the most important characteristics for Kutscher’s dating of the Genesis Apocryphon were:

  • דן/דנה: The demonstrative pronoun דן “this,” much more common than the longer form דנה in the Genesis Apocryphon, is considered later than the דנה of MT Daniel (or זנה of Official Aramaic).

  • Use of aleph instead of he as a prefix for some conjugations of the causative (אפעל/הפעל) and passive-reflexive (אתפעל/התפעל) verbal stems: The usual aleph of the Genesis Apocryphon is considered later than the he of MT Daniel and earlier dialects.

  • אן/הן: The spelling of the conditional particle אן “if,” used intermittently in the Genesis Apocryphon, is considered later than the consistent use of הן in MT Daniel.

  • Certain spellings in the Genesis Apocryphon (e.g., מרי, ‪ראישה‬, לואתי) are deemed to betray what Kutscher called a Middle Aramaic background, later than spellings found in MT Daniel.4

  • Some word forms are of special note, especially those ending with nun in the Genesis Apocryphon (e.g., תמן, ‪כמן‬; = תמה, ‪כמה‬ in MT Daniel) and ordinal numbers (e.g., תניאני in the Genesis Apocryphon; = תנינות in MT Daniel). The forms in the Genesis Apocryphon are considered by Kutscher to be later than those in MT Daniel.

  • Use of the קושט noun pattern, which resembles a standard Hebrew pattern and occurs regularly in the Genesis Apocryphon, is considered to be a later deviation from the more classical Aramaic קשוט.

  • Dissimilation of a geminate consonant through nasalisation (addition of nun; e.g., ‪ינדע‬ [יד״ע]) and retention of etymological nun (e.g., ‪ינפק‬ [נפ״ק]), common in both MT Daniel and the Genesis Apocryphon, is considered to be generally earlier than the assimilated forms (יפק, ‪ידע‬) typically found in later dialects.

  • Employment of lamed to mark the direct object (e.g., להון), again almost ubiquitous in MT Daniel and the Genesis Apocryphon (and in Official Aramaic), is considered to be earlier than use of the particle ית (e.g., יתהון).

  • The internal passive פְעִיל form occurs regularly in MT Daniel and the Genesis Apocryphon, but is “conspicuous by its absence” in later Middle Aramaic.

  • The particle of negation אל, followed by a shortened prefix-conjugation verb (e.g., אל תקוצו), is used in earlier dialects, including MT Daniel and the Genesis Apocryphon, but not in later “western” dialects.

  • ד/די: The relative pronoun די (זי in Official Aramaic) is generally earlier than the shortened, prefix form –ד, which marks the Middle Aramaic dialects.

  • First person plural suffixes tend to be –נא (e.g., עלנא, ‪אנחנא‬) in MT Daniel and the Genesis Apocryphon (also Targum Onkelos), as opposed to the later –ן ending.

Various other “early” or “late” elements of the Genesis Apocryphon, not connected with MT Daniel, were also discussed by Kutscher, leading to a rather mixed and confusing picture that could be skewed toward various dates depending on the relative importance allotted to different patterns of traits. Kutscher ultimately settled on dating the Genesis Apocryphon’s language to the first century BCE (to the first century CE), somewhat later than the language of MT Daniel, but prior to that of Targum Onkelos to the Pentateuch. He was followed by Joseph Fitzmyer, who dated the Apocryphon’s composition to the first century BCE or the first century CE, based mainly on the opinion of Kutscher.5

An important, though sometimes overlooked, response to Kutscher’s article was published in 1963 by H.H. Rowley, who restricted his investigation to the relationship between the Genesis Apocryphon and MT Daniel.6 Surveying fifty-six linguistic traits, many of which had also been treated by Kutscher, Rowley judged that “the language of the scroll is very close to that of the Aramaic parts of the book of Daniel, though slightly later.”7 Rowley’s study moved beyond Kutscher’s in certain respects. For example, he paid more attention to the marking of word-final ‘a’ and ‘e’ vowels either by א or ה, most notably to indicate the final ‘a’ vowel of the determined state and the feminine noun suffix. The presence or absence of א as a vowel marker in some other situations (e.g., as a medial vowel, or to indicate פ״י verb endings) was also noted, with א generally deemed to indicate a later stage of Aramaic. Another important consideration for Rowley was a text’s relation to several phonetic shifts from earlier to later written forms: ‪ק‬>ע (e.g., “earth” ארק/ארע; Heb. ‪ארץ‬), ‪ז‬>ד (e.g., “altar” מזבח/מדבח), ‪ש‬>ת (e.g., “to weigh out, pay” שקל/תקל), and ‪ש‬>ס (e.g., “ten” עשר/עסר). Not all of these changes occurred at the same time, or in the same ways, but they do give some general sense of when a text was written. After considering Kutscher’s date of the Genesis Apocryphon, Rowley objected that, “[o]n linguistic grounds there is nothing to preclude a date in the second century BC, since there is nothing that would require any long interval between the date of the Aramaic of Daniel and the language of the Genesis Apocryphon.”8 A comparison of Kutscher and Rowley demonstrates how examination of similar factors could lead to appreciably different results. Kutscher and others had frequently acknowledged the woefully limited material with which to work, and the refinement that would be possible if further Aramaic texts from the Second Temple period came to light. Happily, with the many added manuscripts from Qumran we are now in a better position to assess not only the date of the Genesis Apocryphon’s language, but also the consistency and character of “Qumran Aramaic” more generally.

As other Aramaic Qumran scrolls were published, some were put through the paces of Kutscher’s method, using his and Rowley’s categories for comparison, and assigning a date relative to the now-growing list of texts. The so-called Job Targum from Cave 11 (11Q10) was the first to receive such attention, initially by van der Ploeg and van der Woude, and then by Sokoloff.9 The former scholars suggested very tentatively that the translation should be placed, on linguistic grounds, between MT Daniel and the Genesis Apocryphon, in the second half of the second century BCE. Several years later Sokoloff offered a far more comprehensive comparison of traits, nearly amounting to a full-scale grammatical treatment of the scroll. The comparative indicators of Kutscher and Rowley were expanded yet further by Sokoloff, including attention to morphology of verbs, pronouns, and use of the determinate and indeterminate states of nouns. In the end, Sokoloff tallied the scroll’s linguistic traits according to the following categories: 1.) Like Biblical Aramaic [16]; 2.) Between Biblical Aramaic and the Genesis Apocryphon [6]; 3.) Like the Genesis Apocryphon [3]; 4.) Later than the Genesis Apocryphon [3]; 5.) Inconclusive [4]. Based on these numbers, he offered that “[t]he date of composition of Tg1 is thus placed sometime between BA (D) and GAp, and – if a date may be hazarded – probably sometime in the second century BCE.”10

A similar approach was adopted by Sokoloff in a later article dedicated to the Aramaic fragments of 1 Enoch. The study was, in part, also a review of J.T. Milik’s The Books of Enoch, and focused especially in on the linguistic and scribal traits of 4Q201 (Ena), which Sokoloff judged “the most interesting of all the published manuscripts.”11 He isolated as noteworthy the following eleven points of orthography and grammar:

  1. A heavy preference for ה rather than א to mark the final ‘a’ vowel of the determinate state, feminine nouns, and other words ending with this vowel.

  2. A heavy preference for ס rather than ש to mark the phoneme ‘s.’

  3. Preference for the א prefix of the causative and reflexive/passive verb forms, rather than ה.

  4. Defective spellings of 2nd pl. perf. verb forms and 2nd and 3rd pl. pronominal suffixes.

  5. Regular assimilation (or elision) of etymological א in both verbs and nouns.

  6. Non-assimilation of נ in the single example of a פ״ן verb (ינפק).

  7. Employment of 3rd f. pl. forms for the pronoun and verb.

  8. Spelling of the pl. dem. pronoun as אלין, not אלן.

  9. Spelling of the 3rd m. pl. independent pronoun as אנון, not המו(ן).

  10. Consistent use of the full form די, not –ד, for the relative pronoun.

  11. Original short ‘u’ vowels are not represented in the orthography.

Sokoloff compared 4Q201 (Ena) with 11Q10 (Job) in much the same way that he had earlier assessed 11Q10 (Job) relative to Biblical Aramaic and the Genesis Apocryphon (1Q20). He concluded that, “[t]he morphological features of 4QEna indicate that it should be placed somewhat after 11QtgJob.”12 All of the above resulted in the following proposed chronological sequence of Aramaic texts, from earlier to later: Daniel (as reflected in the MT), 11Q10 (Job), 4Q201 (Ena), and 1Q20 (apGen). The basic method of Kutscher, Rowley, and Sokoloff has been drawn upon by others for assessing the dates of additional compositions, an example being García Martínez’s treatment of the New Jerusalem text.13

2 Responses to the Typological Method

A major critique of the “typological series” approach outlined above was issued in the early 1990s by two North American scholars, Michael Wise and Edward Cook, a development encapsulated in Cook’s statement that “the entire method of dating Qumran Aramaic texts linguistically needs rethinking.”14 Although Wise and Cook worked independently, their critiques share important similarities.15 Wise set his views within the broader frameworks of the use of Aramaic in Palestine before and during the period when the Aramaic Qumran texts were written, and the book culture of Greco-Roman antiquity.16 With respect to the former area, Wise bristled that scholars have not often enough recognized how Aramaic operated as a diglossic language in Second Temple period Palestine. According to Wise, there was a higher dialect (H) that was primarily written, was restricted to the educated classes, and resembled the older Official Aramaic of the Achaemenid period. This high dialect was accompanied by a lower one (L) that was primarily spoken, incorporated numerous changes into the older dialect (e.g., many of the abbreviated and altered forms listed above), and was much more widespread among the populace than the higher dialect. Wise applied his notion of Aramaic diglossia to the Qumran texts by proposing that the so-called “later” elements identified by Kutscher and others are simply instances where the lower dialect (L) had made its way into a group of writings that were written primarily in the higher dialect (H) by a group of well-trained scribes. If correct, this would be important for dating, since it would mean that two texts with slightly different constellations of linguistic features, like Daniel and the Genesis Apocryphon, could have been composed at the same time, with the former simply reflecting stricter adherence to the higher dialect, either at the stage of composition or subsequent copying. That is to say, the “later” elements are rendered much less effective for the type of dating attempted by Kutscher and others. A further complication, ignored by earlier scholars according to Wise, is that different manuscripts were copied for different purposes, a factor that may have influenced the language of any given copy. For instance, 6Q7 (papDan) is a poorly-written copy on low-grade papyrus, and Wise makes a connection between this manuscript’s physical features, presumed intended purpose as a “personal copy,” and relatively high number of intrusions from the lower dialect. The fact that this manuscript stands alongside copies of Daniel without these intrusions does not necessitate that 6Q7 (papDan) is chronologically later. Its scribe, potentially writing at the same or even an earlier time than the scribe of a more archaic-looking manuscript, simply may not have preserved the higher-dialect form of the text as well as the other scribe, who was perhaps writing a higher-quality copy. This led Wise to the important role of the copying scribe in textual change over time, a point emphasized with equal force by Cook.

Cook began his analysis by noting that, with the publication of 4Q201 (Ena), “it could be seen that the typological method was in trouble.”17 This trouble was caused by a tension between several linguistic features considered by Sokoloff to be late, and Milik’s relatively early dating of the scroll to the first half of the second century BCE on palaeographic grounds. Because of Sokoloff’s previous assignment of the language of 11Q10 (Job) to “sometime in the late second century BCE,”18 which he considered to be linguistically earlier than 4Q201 (Ena) but later than MT Daniel, 11Q10 (Job) would need to be moved to an earlier time, should Milik’s date be accepted. However, this ran up against the common assumption that MT Daniel’s Aramaic should be dated to around 165 BCE, with the chronological window between MT Daniel, 11Q10 (Job), and 4Q201 (Ena) now growing perilously small, if not collapsing altogether. Something clearly had to give, whether Sokoloff’s dating of the language of 11Q10 (Job), Milik’s palaeographic date, or the dating of MT Daniel’s Aramaic. While Cook suggested that MT Daniel’s Aramaic may be reasonably placed as early as the third century BCE in order to accommodate the typological dating system (with which I fully agree), he ultimately rejected this solution for two reasons: 1.) “the probability of orthographic and grammatical revision in the transmission of texts;” and 2.) the existence of “a variety of local orthographies” that confounds attempts at placing orthographic and grammatical phenomena in a straight-line chronological or typological development.19 Like Wise, these points led Cook to argue that “nothing is more certain than that individual scribes differed in their use of such features as matres lectionis, use of or ס, retention of historical spellings, and so on,” as can be seen in comparison between two manuscripts of the same work.20 For Cook, these factors compromise the typological approach so severely that our confidence in linguistic dating is almost completely eroded. In lieu of this system, he proposed that we should simply refer to all Aramaic material from the Judean Desert as “Jewish Palestinian Aramaic of the Hellenistic Roman Period.”21

More recently, Holger Gzella added his voice to this conversation, raising again the point that orthography is a very shaky criterion for dating a composition, since it can easily change over time with scribal transmission.22 More secure, argues Gzella, are morphological and morphosyntactic features, since they have greater resistance than orthography to change, while lexical development is sometimes difficult to untangle from orthography proper.23 Gzella shows through the Arsacid and later Samaria inscriptions, all of which are contemporary with or postdate the Qumran texts, that ostensibly early features can be preserved for a very long time. In fact, the Arsacid inscriptions display an almost flawless reproduction of the considerably earlier Achaemenid period standard.24 Rather than show that these texts antedate those from Qumran, Gzella rightly argues that they advocate for a more complex approach than has been used hitherto. Such an approach would complicate the picture by factoring in the possibility of different linguistic “registers,” literary traditions, written versus spoken forms of the language, and local dialects, as well as more carefully weighing the true import of various linguistic features. In the end, Gzella advocates a salvage project for the typological approach in lieu of the critiques of Cook and Wise, proposing that “one should utilize non-linear models of the development of Aramaic and its multi-dialectal scribal context with competing orthographies for improving the underlying typological method.”25

Another call for complicating the way we handle Qumran Aramaic and other Aramaic dialects has been issued by Aaron Koller, although he pushed forward the discussion in different directions than Gzella.26 Koller stressed especially the factors of genre and ideology in an attempt to break loose of models focused primarily on geography and chronology (e.g., a text or corpus is Eastern Middle Aramaic). A more nuanced approach, argues Koller, should recognize that the situation on the ground included many more factors, and often does not conform easily to a neat geographic-chronological model. With reference to the Qumran texts specifically, Koller follows Ursula Schattner-Rieser by laying emphasis on the heterogeneity of the Qumran Aramaic corpus, in contrast to the more homogeneous approach of, for example, Cook.27 This topic will be taken up further near the end of the present chapter, but suffice it to say that one’s opinion on the heterogeneity or homogeneity of the corpus rests largely on one’s definition of a dialect, and the factors chosen as the focus of comparison.28

3 Scribal Preferences and Textual Transmission I: The Evidence for Scribal Change

Taken together, the studies of Wise, Cook, Gzella, and Koller provide an insightful corrective to Kutscher’s method. While Wise and Cook despair of gleaning any firm conclusions from the typological approach, I agree with Gzella that parts of the method are of serious value to scholars studying these texts, so long as text-external factors like local dialects and individual scribal preferences are used to inform and condition the method. There is no going back to the placement of every Qumran Aramaic text in sequence on a single developmental trajectory; the situation is undoubtedly much less tidy than that.

We have seen that an important factor, stressed by Wise, Cook, Gzella, and Koller, is the role of individual scribes and the extent to which we may assume a copy to represent faithfully a composition’s “original” text, whatever that may have been and by whichever processes it may have developed. The important distinction between the character of a work’s language at the point of original composition and its subsequent copies has often been left ambiguous in studies on Qumran Aramaic texts, though the factor of scribal intervention through the transmission process was raised already by Tisdall and Schäder in the early twentieth century for Daniel.29 For example, Kutscher did not specify whether he regarded the literary composition of the Genesis Apocryphon to date to the first century BCE (to the first century CE), or simply this copy of the work (i.e., 1Q20).30 This ambiguity resulted in different interpretations by the first editors of the scroll, Nahmad Avigad and Yigael Yadin, and Joseph Fitzmyer in his later commentary.31 Avigad and Yadin agreed with Kutscher’s date, but added that “[t]his does not, of course, fix the time of the original composition,” which they judged to be the second century BCE or earlier. Fitzmyer, on the other hand, collapsed the stages of original composition and copy, even considering that 1Q20 (apGen) may be the original autograph, a view widely rejected in later scholarship with good reason. In each case, the position adopted is bound to influence the subsequent historical treatment of the scroll.

As will become clear below, there can be no doubt that scribes had their own habits and idiosyncrasies when it came to copying a text, and that each copy could take on its own set of linguistic traits. Several examples from different copies preserving the same passage will suffice, for the moment, to introduce and illustrate this point:

  1. 4Q205 (End) 2i.26 ו̊ד̊כר די ען

    4Q206 (Ene) 4ii.12–13 ודכר] זי ע̇[ן

  2. 4Q209 (Enastb) 23.5 כוכבין ובדכן

    4Q210 (Enastc) 1iia+b+c.16 כ̊וכביא בדיל כ[ן

  3. 4Q112 (Dana) 5.3–4 ו]ומנהם פרזל מ[לכו פליגה]תהוא… […ואצבע]ת̊ רגליא מנהם פרז[ל

    MT Dan 2:41–42 ומנהון פרזל מלכו פליגה תהוא……ואצבעת רגליא מנהון פרזל

Each of these parallels contains a disagreement between presumably “early” and “late” linguistic features in two copies (or in Daniel’s case, the entire Masoretic tradition), thereby unambiguously catching the process of scribal updating or variation in motion.32 The first example, from two Enoch fragments, attests to the phonetic transition from ז to ד, representing the “hardening” (or dentalization) of an originally softer, more spirantized consonant. The same זי form is witnessed several other times in the Qumran texts (4Q206 [Ene] 4iii.16, 4Q213a [Levib] 3–4.5, 4Q530 [EnGiantsb] 2ii+6–12(?).1, 4Q536 [Birth of Noahc] 2i+3.4), at times in close proximity to the dentalized די form. Remarkably, in the same 4Q206 (Ene) that uses זי we also find what is presumably the latest form of this word, the prefixed –ד (twice in 1xxii.6). Both 4Q205 (End) and 4Q206 (Ene) are dated palaeographically to the first century BCE. In the second example, from the Astronomical Book of Enoch, we find several linguistic changes in close proximity, including the addition of a conjunction and the collapsing of two words into a contracted form. In general, these two manuscripts exhibit a relatively high number of orthographic and morphological disagreements in the available parallel passages. Again, both manuscripts are dated by Milik and Drawnel to the first century BCE. While it is true that the Daniel manuscripts from Qumran agree to a significant extent with the MT consonantal text (i.e., the ketiv, and less so with the qere traditions), 4Q112 (Dana) contains what seems to be a stark archaism in the Official Aramaic form of the 3mp pronominal suffix הֹם–, rather than expected form as found in MT Daniel and Qumran Aramaic more broadly: הון–.‪33‬ 4Q112 (Dana) also uses once the aleph-prefixed ithpeel spelling rather than the he-prefixed hithpeel of the MT, another apparent archaism.34 Another Daniel manuscript, 4Q113 (Danb), spells the MT pronoun אנת and 2ms suffix ך– instead as אנתה and כה–, both forms usually considered to represent later orthographic practice. As with the examples above, Cross and Ulrich judge both 4Q112 (Dana) and 4Q113 (Danb) to be first century BCE copies.

What are we to make of differences like these? In what ways should they change our approach to dating the language of Qumran Aramaic texts, in keeping with Gzella’s suggestions? If our goal is to assess the overall coherence of Qumran Aramaic as an Aramaic dialect, and secondarily its relationship to other dialects, it is clear that discrepancies between copies like those just offered must condition the scope of linguistic variation that we should expect and accept when assessing Qumran Aramaic as a dialect. This ties into the very important, though infrequently discussed, topic of how we define a “dialect.”35 What do we mean by this word? What are our expectations of a dialect? How tight must the coherence among disparate texts be for inclusion within a single dialect? How many (or what sorts of) linguistic features must be present before a text is deemed to fall outside of a given dialect? These are very difficult questions to answer, given the high complexity of coordinating manifold linguistic features across a considerable number of texts.

A first step toward answering questions like these with respect to Qumran Aramaic is to get a good sense of the range of scribal variation existing among the scrolls. To achieve this end, one can look at variation exhibited between different Qumran Aramaic literary works, as Sokoloff did for 11Q10 (Job) and 4Q201 (Ena). However, as the differences between the Enoch manuscripts illustrate, this may give a “false read” if our goal is to compare literary compositions and not simply individual copies. In other words, Sokoloff does not demonstrate that the language of Aramaic Enoch is older than that of the Genesis Apocryphon, but rather that the language of 4Q201 (Ena) is older than that of 1Q20 (apGen). Discrepancies between parallel passages show beyond doubt that another, now lost copy of the Genesis Apocryphon could have contained features considered older than those in 1Q20 (apGen), which may in turn compel us to date the composition’s (though not that copy’s) language to an earlier period. At the same time, it should be borne in mind that some Aramaic works from the Qumran caves are composite in nature, bringing together stories potentially written at different times, and with slightly different linguistic features. An example of this is the Aramaic Enoch anthology underlying what would eventually become 1 (Ethiopic) Enoch. There is also the possibility of more sweeping revisions to individual works – what we might call different editions – as may have been the case for the Astronomical Book of Enoch. Of course, such revisions could have included changes to the Aramaic in which they were written.

A constructive way to begin assessing the range of variation we might expect across the Qumran Aramaic corpus is to catalogue the types of scribal changes exhibited between copies in parallel passages of the same literary work. With this range of scribal variation before us, we may then go on to compare the range of linguistic variation between separate literary compositions in the Qumran Aramaic corpus, with the aim of seeing whether the variation between compositions exceeds that exhibited between copies of the same work. If the types of scribal changes found between copies of the same work are roughly equivalent to those found between separate compositions, then any attempt at the relative dating of literary compositions based on language alone is seriously compromised, and should be avoided. At most, it should serve only as tertiary, supporting evidence of arguments for dating made on other grounds. Though the sample size of parallel passages shared by two or more copies of the same work at Qumran is regrettably small, the range of discrepancies in these parallel passages can provide a baseline for linguistic comparison among the disparate Qumran Aramaic texts. This baseline can then inform our method for typological dating, since it exposes at least some of the factors that should be neutralized for dating purposes, based on demonstrated scribal intervention in the manuscripts. For example, if we see that two copies of the same work vacillate between using the aphel and haphel spellings for the causative stem in parallel passages, with how much confidence can we use this factor for the relative dating of literary compositions from our corpus based on the available manuscripts? The consistent use of the aphel spelling in a given Qumran Aramaic work, then, should be given little weight for dating the language of composition, since we can see in other contexts at Qumran this feature being changed by scribes over the course of transmission. To be sure, this aphel spelling remains useful for assessing the habits of that particular scribe and copy, though it will be suggested below that we should be careful dating with too much precision even an individual copy based on such linguistic features.36

Below is a complete list of parallels in the Qumran Aramaic scrolls, followed by an initial catalogue of the types of linguistic changes present in parallel passages from different copies of Qumran Aramaic works.37 For the Qumran Daniel and Ezra manuscripts I will also include the consonantal (ketiv) text of the MT for the purpose comparison.38

Parallel passages in the Aramaic Qumran scrolls39

1Q71 (Dana) 2.4–8//MT Dan 2:4–6

1Q72 (Danb) 1.1–13//MT Dan 3:22–28//4Q115 (Dand) 2ii.2–5

1Q72 (Danb) 2.1–6//MT Dan 3:27–30

4Q112 (Dana) 1ii.1–7//MT Dan 2:9–11

4Q112 (Dana) 3i, 17.1–18//MT Dan 2:19–33

4Q112 (Dana) 3ii, 4–6.1–18//MT Dan 2:33–46

4Q112 (Dana) 7.1–9//MT Dan 2:47–3:2

4Q112 (Dana) 8.17–18//MT Dan 4:29–30

4Q112 (Dana) 9.14–18//MT Dan 5:5–7

4Q112 (Dana) 10–11.1–6//MT Dan 5:12–14//4Q113 (Danb) 1–4.7–8

4Q112 (Dana) 12.1–6//MT Dan 5:16–19

4Q112 (Dana) 13.1–4//MT Dan 7:5–7//4Q113 (Danb) 12–13.3–4

4Q112 (Dana) 14.5–9//MT Dan 7:25–28//4Q113 (Danb) 15.19–21

4Q113 (Danb) 1–4.1–4//MT Dan 5:10–11

4Q113 (Danb) 1–4.7–8//MT Dan 5:12//4Q112 (Dana) 10–11.2–3

4Q113 (Danb) 1–4.11–12//MT Dan 5:14–15

4Q113 (Danb) 1–4.14–15//MT Dan 5:16

4Q113 (Danb) 5–6.1–3//MT Dan 5:19–20

4Q113 (Danb) 5–6.6–7//MT Dan 5:21–22

4Q113 (Danb) 7i.12–22//MT Dan 6:8–13

4Q113 (Danb) 7ii, 8.2–20//MT Dan 6:13–22

4Q113 (Danb) 9–11.11–22//MT Dan 6:27–7:4

4Q113 (Danb) 12–13.3–5//MT Dan 7:5–6//4Q112 (Dana) 13.2–3

4Q113 (Danb) 14.2//MT Dan 7:11?

4Q113 (Danb) 15.18–22//MT Dan 7:26–28//4Q112 (Dana) 14.6–8

4Q115 (Dand) 1.1–3+2i.1–5//MT Dan 3:8–12

4Q115 (Dand) 2ii.1–6//MT Dan 3:23–25//1Q72 (Dana) 1.3–5

4Q115 (Dand) 3–7.1–6//MT Dan 4:5–9

4Q115 (Dand) 3–7.11–19//MT Dan 4:12–16

4Q115 (Dand) 8–9.1–14//MT Dan 7:15–23

4Q117 (Ezra) 2.1–4//MT Ezra 4:9–11

4Q117 (Ezra) 3.1–9//MT Ezra 5:17–6:5

4Q201 (Ena) 1iii.1–2//4Q202 (Enb) 1ii.6–7

4Q201 (Ena) 1iii.10–15//4Q202 (Enb) 1ii.15–19

4Q201 (Ena) 1iv.1–7//4Q202 (Enb) 1iii.1–8

4Q201 (Ena) 1iv.10–11//4Q202 (Enb) 1iii.10–11

4Q201 (Ena) 1ii.1–3//4Q204 (Enc) 1i.20–22

4Q201 (Ena) 1ii.5–11//4Q204 (Enc) 1i.24–30

4Q201 (Ena) 1iii.5–11//4Q204 (Enc) 1ii.24–29

4Q202 (Enb) 1vi.9//4Q204 (Enc) 1vi.16

4Q204 (Enc) 1xii.28–30//4Q206 (Ene) 1xxvi.14–17

4Q204 (Enc) 4.1//4Q205 (End) 2ii.30

4Q205 (End) 2i.26–29//4Q206 (Ene) 4ii.12–16

4Q205 (End) 2ii.27–29//4Q206 (Ene) 4iii.19–21

4Q209 (Enastrb) 23.5–8//4Q210 (Enastrc) 1iia+b+c.15–18

1Q23 (EnGiantsa) 29.1–2//6Q8 (papGiants) 1.4–5

4Q206 (Ene) 3i.5–6//4Q533 [4Q556] (EnGiantse) [Prophecya] 4.1–2

4Q213 (Levia) 1ii.4, 6//4Q214b (Levif) 8.1–2

4Q213 (Levia) 2.5//4Q214a (Levie) 2–3ii.5

4Q214 (Levid) 1.5//4Q214b (Levif) 5–6i.3

4Q214a (Levie) 1.1//4Q214b (Levif) 2–3.5

4Q214b (Levif) 2–3.8//4Q214 (Levid) 2.3//1Q21 (Levi) 45.2

4Q243 (psDana) 13.1–4//4Q244 (psDanb) 12.1–3

4Q529 (Words of Michael) 1.13–14//4Q571 (Words of Michaela) 1.13–14

4Q552 (Four Kingdomsa) 1ii.1–11//4Q553 (Four Kingdomsb) 3+2ii+4.2–7

2Q24 (NJ) 4.10–16//11Q18 (NJ) 20.2–7

2Q24 (NJ) 1.1–4//4Q554 (NJa) 1ii.12–15//5Q15 (NJ) 1i.1–2

4Q554 (NJa) 1iii.11–21//5Q15 (NJ) 1i.15–ii.4

4Q554 (NJa) 1iii.21//4Q554a (NJb) 1.1//5Q15 (NJ) 1ii.4–5

4Q554a (NJb) 1.1–10//5Q15 (NJ) 1ii.4–13

4Q554 (NJa) 1ii.15–22//5Q15 (NJ) 1i.3–6

4Q531 (EnGiantsc) 1.5//4Q532 (EnGiantsd) 2.10(?)

4Q534 (Birth of Noaha) 7.2–6//4Q536 (Birth of Noahc) 2ii.11–13

4Q535 (Birth of Noahb) 3.4–6//4Q536 (Birth of Noahc) 1–3

4Q543 (VisAmrama) 1a, b, c.1–8//4Q545 (VisAmramc) 1ai.1–8//4Q546 (VisAmramd) 1.1–4

4Q543 (VisAmrama) 2a–b.1//4Q545 (VisAmramc) 1ai.14–19

4Q543 (VisAmrama) 3.1–3//4Q545 (VisAmramc) 1a–bii.17–18//4Q546 (VisAmramd) 2.3–4

4Q543 (VisAmrama) 4.2–4//4Q544 (VisAmramb) 1.7–8//4Q547 (VisAmrame) 1–2.6–7

4Q543 (VisAmrama) 5–9.1–7//4Q544 (VisAmramb) 1.11–14//4Q547 (VisAmrame) 1–2.11–13

4Q543 (VisAmrama) 14.1//4Q544 (VisAmramb) 3.1

4Q543 (VisAmrama) 15.2//4Q546 (VisAmramd) 6.1

4Q544 (VisAmramb) 1.1–4//4Q545 (VisAmramc) 1a–bii.13–19

4Q544 (VisAmramb) 1.7–10//4Q547 (VisAmrame) 1–2.6–10

4Q544 (VisAmramb) 3.2//4Q546 (VisAmramd) 4.1

4Q545 (VisAmramc) 6.3//4Q547 (VisAmrame) 3.3

4Q196 (papToba) 13.1–4//4Q197 (Tobb) 4i.10–14

4Q196 (papToba) 14i.4–8//4Q197 (Tobb) 4ii.9–12

4Q196 (papToba) 14ii.4–11//4Q197 (Tobb) 4iii.1–8

4Q196 (papToba) 18.15–16//4Q198 (Tobc) 1.1–8

3.1 Catalogue of Scribal Changes Witnessed in the Parallel Passages

FIG130001
FIG130001
FIG130001
FIG130001
FIG130001
FIG130001
FIG130001

The data catalogued above are obviously limited, not indicating of the full range and depth of scribal intervention we would find in the Aramaic scrolls from the Qumran caves were we to have the corpus fully preserved. We must also compensate for the perception that there is more variation in some works than others, since this is likely due to how many manuscripts happen to be preserved of any given composition. Were we to possess as many manuscripts of the Genesis Apocryphon or Words of Michael as we do, say, for Daniel or the Visions of Amram, it is reasonable to assume that we would find a comparable range of scribal variation across all these works. The list above serves mainly to illustrate some of the shifts taking place through scribal intervention between manuscripts of the same work, and in this way provides a good starting point for analyzing Qumran Aramaic. In addition to the changes catalogued above, we may add some linguistic variations witnessed between different manuscripts of the same work, though not preserved in the extant, directly-overlapping passages. These include:

  1. The scribe of 4Q201 (Ena) generally preferred ס to represent etymological /s/, versus ש in 4Q202 (Enb) and 4Q204 (Enc) (e.g., סגי vs. שגי), and indeed the large majority of the Aramaic Qumran scrolls.

  2. 4Q208 (Enastra) and 4Q209 (Enastrb) exhibit several significant differences in spelling and phrase construction, such as the assimilation of נ and ל as object marker in the analogous phrases וב]צ̇יר מנ נהורה לשב[יעינ (4Q208 [Enastra] 15.2) and ובציר מנהורה שביעין חמשא (4Q209 [Enastrb] 7ii.6). Spellings of the words יממא “day” and לילא “night” also vary between the copies.

  3. 4Q530 (EnGiantsb) uses defective spellings such as אלן “these” and כל “all,” while the scribe of 4Q531 (EnGiantsc) preferred the full spellings אלין and כול.

  4. 4Q543 (VisAmrama) and 4Q548 (VisAmramf) use only כל, but 4Q544 (VisAmramb) and 4Q547 (VisAmramc) primarily ‪כול‬. 4Q546 (VisAmramd) is mixed in its spelling of this word.

  5. 4Q113 (Danb) regularly uses the fuller spellings כול and כה–, unlike 4Q112 (Dana) and MT Daniel.

  6. 4Q113 (Danb) 7i.15 and 4Q115 (Dand) 3–7.17 use the consecutive (and grammatically-correct) spelling כלקבל (cf. 4Q204 [Enc] כלקובל), versus MT Daniel כל קבל.

  7. 4Q214 (Levid) 4.3 uses the demonstrative pronoun דנה, rather than the דן of 4Q213a (Levib) and 4Q213b (Levic). The same difference obtains between 4Q531 (EnGiantsc) (דנה) and 4Q530 (EnGiantsb) (דן).

  8. The New Jerusalem manuscripts show a number of disagreements in orthography. In general, 4Q554a (NJb) is written more defectively than the other copies, as with the preposition גו (גוא in the other manuscripts). Most manuscripts have כל, except for 2Q24 (NJ) and 11Q18 (NJ), which use ‪כול‬. 11Q18 (NJ) has לקובל and 4Q554 (NJa) לקבל.

  9. 4Q198 (Tobc) twice has the causal stem with a ה prefix, while the אפעל predominates in 4Q196 (papToba) and 4Q197 (Tobb) (though there are two instances of הפעל in 4Q196 [papToba]).

  10. In the New Jerusalem text, 5Q15 (NJ) (1ii.7) spells “their width” as פותאהון, but 2Q24 (NJ) (7.1) has rather פותיהון.

Taken together, the evidence of scribal change collected above illustrates the following principles:

  1. A limited number of orthographic conventions changed freely between copies of the same literary work, as indeed they sometimes vary in a single manuscript (e.g., 1Q20 [apGen], 4Q201 [Ena]). Chief among these changes are:

    1. א and ה being interchanged in various situations, including at the beginning of the suffix conjugation causative and passive-reflexive verbal stems. The evidence suggests that over time א took general precedence in all of these situations among the Qumran scrolls. At times, this interchange at the end of nouns may have caused confusion between the definite state and a 3ms suffix.

    2. א being used interchangeably to represent full vowels. Again, it seems that the basic diachronic trend was a growing usage of א in this way.

    3. Scribes fluctuating between defective and full spelling preferences, with the bulk of our manuscripts exhibiting a diachronic trend toward the full system. This especially included the use of ו for long /o/ and /u/ vowels, but also י for long /e/ and /i/.

    4. Representation of the final, open vowel of the 2ms suffix conjugation verb (קטלת), 2ms pronoun (אנת), and 2ms suffix (ך–) through augmentation with the letter ה. The augmented forms with ה are typically taken to be later in their development.

    5. Varied representation of gemination or nasalization with the letter נ (e.g., מדע vs. מנדע, ‪מתכסין‬ vs. מתכנסין).

    6. Replacing ש with ס in certain words (e.g., עשר, ‪שגיא‬, and בשר).

    7. Other small orthographic or phonological adjustments based on pronunciation or perceived grammatical propriety.

  2. At least some scribes felt comfortable with the contraction, abbreviation, or symbolic representation of words or short phrases, which typically did nothing to alter their meaning. Examples of this are the contraction of בדיל כן to בדכן in 4Q209 (Enastrb), לא איתי to לת in 4Q201 (Ena), the abbreviation of אמין to א in 4Q554 (NJa), and the replacement of אלהא with four dots in 4Q196 (papToba). Occasionally, the meaning did change, as with הכא די in 4Q212 (Eng), rather than הא כדי in MT Daniel.

  3. There was clearly some tolerance for, or oversight of, minor adjustments in grammatical construction and phrase formation (morphosyntax), in some cases bringing better clarity to the text. This included the following:

    1. The addition, subtraction, or substitution of one to a few words with something that the scribe apparently felt was more suitable or stylistically attractive. For example, ושאלתה מן שמך in 4Q552 (Four Kingdomsa) versus ושאלתה ואמרת לה מן שמך in 4Q553 (Four Kingdomsb).

    2. The transposition of words.

    3. Adding or deleting a suffix.

    4. Adjusting the conjugation of a verb, for various reasons either apparent to us (e.g., agreement with a perceived subject) or no longer obvious.

    5. Changing the number or definiteness of a noun or adjective, at times in order to achieve agreement with the governing verb.

    6. The addition, subtraction or substitution of a preposition or conjunction. An example is 4Q113 (Danb) using בפם, whereas MT Daniel has על פם.

  4. Adjustments were occasionally made to word morphology. It stands to reason that some of these changes reflect the conventions better known to, or preferred by, the scribe writing the copy. These adjustments included:

    1. Changes in verb morphology, such as the addition of מ to the derived stem infinitive in 4Q544 (VisAmramb), graphic representation of the geminate stem in 4Q206 (Ene), and variation between the causative and passive-reflexive conjugation in 4Q112 (Dana) and MT Daniel.

    2. Representation of weakening or strengthening guttural letters, as with the difference between חעכון and העכן in 4Q543 (VisAmrama) and 4Q544 (VisAmramb).

    3. Shifts in the written form of nouns from the קטול to קוטל pattern, or vice versa.

  5. In some cases a scribe has replaced one word with another that fits the same context. There are few cases of this, and the words are typically very close or the same in meaning. The majority of these changes occur between the Qumran Daniel manuscripts and MT Daniel, which is likely due to the fact that we have more parallel text preserved there than elsewhere.

  6. Minor spelling mistakes occurred from time to time, for a variety of reasons. In some cases, the word may have no longer been current, or a little-used loanword. At other times the scribe seems simply to have erred based on graphic or aural confusion (e.g., בקדה for פקדה in 4Q196 [papToba]). When we look outside of the preserved parallel passages, we can see that some scribes were more susceptible to such slips, as in the case of 4Q542 (TQahat). That scribe might have taken a lesson from the scribe of 4Q537 (TJacob?), who was nearly flawless in his task.

  7. Only very occasionally do we witness larger-scale alterations in the available parallel passages. The only two preserved instances of this are found between 4Q112 (Dana) and MT Daniel, and 4Q209 (Enastrb) and 4Q210 (Enastrc) in the Enochic corpus. Neither parallel is fully preserved. In the case of Daniel, the change is actually quite minor, apparently restricted to the substitution and/or transposition of up to three words. The difference between 4Q209 (Enastrb) and 4Q210 (Enastrc) appears to be greater, with what Milik (followed by Tigchelaar and García Martínez) called “a long passage” omitted in the latter copy or added to the former. In fact, if Milik’s reconstruction is correct, the difference would be approximately one line or less depending upon line length. Milik suggested the possibility that this was a case of homoeoteleuton, but there is no evidence to support his claim.

We are able to see from this survey that the changes made between copies are, on the one hand, fairly regular and to be expected, while on the other hand they are restricted to a set group of minor changes to orthography, word or phrase morphology (morphosyntax), and only very occasionally word choice or more extensive phrasing. Though it was not included above, a similar picture emerges from comparison of the Qumran manuscripts of the Aramaic Levi Document and the considerably later copy from the Cairo Geniza.40 It is clear that at least some copying scribes used their own systems and preferences. Some scribes did so with consistency (e.g., the scribe of 4Q113[Danb]), while others were much more erratic (e.g., the scribes of 4Q201 [Ena] and 4Q208 [Enastra]). Importantly, the changes seen in the parallel passages above mirror, to a great extent, the divergences witnessed across different Qumran Aramaic compositions. A couple of the more salient points not covered by scribal intervention in the parallel passages discussed above are:

  1. Interchange between the equivalent independent personal pronouns המון and אנון for the direct object. המון is the earlier form used regularly in Official Aramaic, while אנון gains ascendancy in Jewish Aramaic (with similar forms in other dialects) over the course of the Second Temple period. המון is found only in 4Q242 (PrNab), 4Q112 (Dana), and 11Q10 (Job) at Qumran.

  2. Attestation of some of the phonetic shifts seemingly taking place during this period, such as between ק and ע, ‪ע‬ and א, or ש and ת. Since we do witness scribal alterations from ז to ד and ח to ה among manuscripts of the same work, it may be that the other shifts are simply not preserved in the parallels left to us. As Schattner-Rieser has noted, many of these shifts are already taking place in the fifth century BCE Aramaic texts from Elephantine and fourth century BCE Wadi Daliyeh documents.41

The brevity of this list shows that the orthographic and morphological features which vary between individual works and are not present between parallel passages in copies of the same work, are remarkably limited. Of course, our restricted comparative evidence does not allow us to assess satisfactorily the extent of the differences between two separate works. For instance, if 11Q10 (Job) and 4Q242 (PrNab) use primarily the הפעל verb and the pronoun המון, while 1Q20 (apGen) and 4Q531 (EnGiantsc) have overwhelmingly the later אפעל and אנון, should we consider the large extent of the disagreement to undermine the possibility that scribal intervention could adequately account for these differences? To put it another way, could a scribe in the process of copying – or several scribes over the course of several successive copies – intervene to the extent that a text with mostly the הפעל and המון was converted into one with largely or entirely the אפעל and אנון? The answer remains elusive.

4 Scribal Preferences and Textual Transmission II: Archaisms, Archaizing, and Lessons from Elephantine and Arsacid Elymais

For many of the scribal changes described above, the two or more linguistic options potentially chosen by a scribe can be identified as earlier or later in the diachronic development of Aramaic. These identifications rely on our knowledge of earlier dialects such as Early and Official Aramaic on the one hand, and later ones such as Targumic and Jewish Palestinian Aramaic on the other. They also form the foundation of the typological method established by Kutscher and others. It is clear that the later scribal features provide a terminus ante quem for when a Qumran Aramaic work was written, even if some problems remain with the precision of dating such elements. However, a methodological debate has developed over how the earlier elements – sometimes referred to as archaisms – should be understood, a debate that bears heavily on the task of linguistically dating these texts. The two sides of the debate are illustrated nicely by a collegial exchange between Ursula Schattner-Rieser and Steven Fassberg at a conference on Qumran Aramaic texts, at Aix-en-Provence in 2008.42 Schattner-Rieser assumed that archaisms in these texts provide a linguistic glimpse of elements from earlier stages in the life of this literature. While many of the archaisms have been lost through the process of scribal change and updating, the remaining early linguistic features more accurately reflect the date of composition than the later, updated ones. Fassberg, on the other hand, wonders if these so-called early features are rather an imperfect attempt at archaizing by those who wrote the Qumran Aramaic texts, in order to give their writings a patina of antiquity and authority. To support this idea, Fassberg adduced the examples of the Hebrew Qumran scrolls, the late books of the Hebrew Bible, and Ben Sira, all of which are widely regarded to archaize by mimicking earlier stages of Hebrew drawn from the ancestral writings of Israel. In this case, the later linguistic elements of the Qumran Aramaic texts are more indicative of their compositional setting, and the archaisms a faux embellishment by authors and scribes.

This debate is difficult to resolve, since either side can make a plausible argument based on the scribal changes witnessed in the Qumran scrolls. Those who favor the explanation of Schattner-Rieser can appeal to changes between copies as evidence that scribes were updating language in minor, but appreciable ways over the course of transmission. Each scribe would have done this in a different manner, and to a different extent. However, from Fassberg’s perspective the fact that the manuscripts are mostly dated to the first century BCE may simply suggest that some scribes did a better job of effecting or preserving the archaisms than others. In either case, it is obvious that we must allow for a fairly stable range of scribal preference and variation. While I find Schattner-Rieser’s view on archaisms and scribal change more convincing on grounds apart from language, the fact that scribes might consciously choose to write in an Aramaic literary idiom either more conservative or more progressive than their contemporary scribes or standard local dialects is nicely illustrated by two examples, one from the fifth century BCE texts discovered at Elephantine, and the other from the Aramaic inscriptions at Elymais dating to the Arsacid (or Parthian) period of rule, in the first to third centuries CE.

Margaretha Folmer has drawn attention to the fascinating case of two scribes at Elephantine who were father and son, Natan bar Ananiah and Mauwziah bar Natan bar Ananiah.43 In contrast to the situation of the Qumran scrolls, the names and dates included in the documents from Elephantine allow us to situate their Aramaic in time and place with confidence, and to compare scribes. Natan and Mauwziah, who belonged to a larger family of scribes and wrote in the middle to late fifth century BCE, are notable because they differ in some surprising ways regarding linguistic usage. The father generally wrote with a more progressive (“later”) orthography and morphology than did his son, who tended to be more conservative (“earlier”) in his style.44 This included Natan more often beginning words with the dental ד where Mauwziah had ז,‪45‬ occasionally using the root תק״ל “weigh” instead of Mauwziah’s שק״ל,‪46‬ employing the prefix conjugation causative verb without ה while Mauwziah included it (e.g., אנצל [TAD B1.3:13] vs. אהנצל [TAD B3.8:42]), and using the apocopated pronominal suffixes ך– (2fs) and וה– (3ms) instead of Mauwziah’s כי– and והי–. Moreover, Natan once wrote the noun for “land” as ארע, rather than the more expected ארק of his direct contemporary, Attarshuri bar Nabuzeribni. Conversely, Mauwziah adopted a more progressive spelling than his father for the suffix-conjugation 2mpl verb ending תן–, whereas Natan used תם–. Mauwziah also used the verb stem כה״ל “be able” as opposed to Natan’s use of the (in Folmer’s opinion earlier) synonym יכ״ל in similar contexts. Mauwziah preferred the oft-repeated temporal phrase מן יומא זנה עד עלם to his father’s מן יומא זנה ועד עלם (with conjunctive ו), and had the uncharacteristic practice of beginning the dependant clause of a conditional sentence with the prefix conjugation instead of the suffix conjugation of his father and other Elephantine scribes.47 It should be added that another Elephantine scribe, who was a close contemporary of Mauwziah, Haggai bar Shemaiah, used some of the same progressive practices as Natan. Haggai used ד instead of ז to an even greater extent than Natan,48 and like Natan he wrote ארע rather than ארק in TAD B3.4:5. Finally, Haggai preferred the apocopated spelling of the first person common plural pronoun אנחן, not אנחנא used by Mauwziah and other Elephantine scribes.

The features that fluctuate between these Elephantine scribes are similar to some of the features used to date the Qumran texts typologically and place them in a relative order. Even though we certainly find more variations present among the Qumran texts (which considerably outnumber those at Elephantine), the texts from Elephantine show us that scribes working at the same time, in the same town, presumably trained in the same scribal system, and even living under the same roof had their own idiosyncratic preferences. These preferences could give the false appearance of a text being earlier or later, but could also be combined in a way that would otherwise confuse attempts at a neat relative dating. Because we are fortunate enough to possess the names and dates on the Elephantine texts, we can see that attempts at relative dating would be wrong-headed from the start. Were the names and dates not preserved and relative dating was attempted, we would almost surely have gotten the order wrong.

Another cautionary tale may be drawn from the eleven brief, rock-cut inscriptions at Tang-e Sarvak and Tang-e Butān, Elymais, in south-western Persia.49 These inscriptions were apparently commissioned by the ruling aristocracy, and date to between the first and third centuries CE. Though they were written at a relatively late date, contemporary with other, more updated eastern Aramaic dialects such as those at preserved at Assur, Hatra, and Palmyra, Gzella observed that the inscriptions of Elymais use a noticeably more pure, classicizing form of Official Aramaic. It should be stressed that the inscriptions are very short, formulaic, and in a few cases fragmentary; there is really too little text preserved to make far-reaching claims. Nevertheless, it is true that enough is extant to recognize a dialect that, were we not able to situate it historically due to the place of its discovery and script, would surely be dated earlier than the second century CE.

To be sure, the Qumran, Elephantine, and Elymais texts are not all cut from the same cloth, and, in keeping with Koller’s recommendations, we ought to bear this in mind when doing any comparison.50 Some factors that must be taken into account are the media, genre, ideology, social setting, and geographic location of each group of texts. The Elephantine documents are mostly legal or administrative in nature and written on papyrus, in Upper Egypt, by scribes trained specifically for that task. At Elymais we have brief monumental, official inscriptions in stone, again by trained scribes and masons. In both cases we might expect a conservative scribal approach based on genre, though we have still seen modest variation, either within a corpus or against contemporary dialects from adjacent regions. In both cases this variation cut against straightforward typological dating models. The Qumran scrolls, in contrast to the Elephantine and Elymais texts, are largely religious and literary in nature, written on leather or occasionally on papyrus, and were copied repeatedly over an extended period of time. In most cases, the scribes seem to have been well-trained for their task, though this varies across the corpus.

Having addressed the scribal variation present at Qumran, the possibilities of historical archaisms versus archaizing, and some examples of contemporary texts using varied linguistic features, we are now ready to move on to a descriptive overview of the Aramaic of the scrolls from the Qumran caves.

5 A Descriptive Overview of Qumran Aramaic with Reference to Bordering Aramaic Dialects

David Crystal’s Dictionary of Linguistics and Phonetics defines a dialect as “[a] regionally or socially distinctive variety of language, identified by a particular set of words and grammatical structures.”51 Definitions like these by necessity leave open the fuzzy boundaries of where one dialect begins and another ends. Is Qumran Aramaic “regionally or socially distinctive”? Should we classify it as a language variety “identified by a particular set of words and grammatical structures”? Our answers will depend on the standards of variety allowed or expected, whether implicitly or explicitly. In the end, the identification of a dialect should be a heuristic tool for those working with, or interested in, these texts and how they fit into their broader historical, geographic, and social surroundings. Below, I seek to provide an accessible overview of Qumran Aramaic with the following guiding principles: 1.) To provide an easy entry into many of the salient features of the language for students of the Qumran Aramaic texts, especially relative to widely-used categories of Official Aramaic, Biblical Aramaic, and the Aramaic texts most closely following after the Qumran scrolls chronologically (e.g., the documents from elsewhere in the Judean Desert, or the Onkelos and Jonathan targums); 2.) To judge by these comparative observations whether the Qumran Aramaic corpus obtains levels of internal homogeneity and the required distance from other Aramaic text collections to justify calling it a distinctive dialect; and 3.) To establish the range of time during which the Qumran Aramaic texts were most plausibly written, based on linguistic factors alone. I must hasten to add that I neither attempt here to provide an exhaustive reference grammar of Qumran Aramaic, for which we are now fortunate to have Takamitsu Muraoka’s Grammar of Qumran Aramaic, nor to cover all aspects of the language. Rather, my goal is to touch on what are, in my opinion, some of the most salient features with respect to the guiding principles just laid out. Where a feature is not distinctive in relation to other surrounding Aramaic dialects, I typically will not discuss it. Much of what follows synthesizes previous scholarship, relying heavily on the pioneering work of earlier and contemporary Aramaicists such as Kutscher, Muraoka, Fitzmyer, Beyer, Greenfield, Puech, Sokoloff, Fassberg, Cook, Folmer, Schattner-Rieser, Gzella, Koller, and others. I occasionally add my own suggestions, drawn in large part from the profiles in this volume, to which readers may refer for further details on any given text.

5.1 Spelling (Orthography and Phonology)

The difference most readers first notice when comparing Qumran Aramaic texts to Aramaic works from outside the Qumran scrolls corpus is the spelling. As with any language, the pronunciation and spelling of Aramaic shifted with the passing of time and geographic distance, something that took place at varying places and in different ways, depending on the particular locale.52 Achaemenid Official Aramaic – the international language of the Persian Empire beginning with Darius I around 550 BCE – provides an important waypoint in the development of the language. At this time, the language was standardized within scribal circles throughout the empire, and remained remarkably stable in the available literary record over an impressive geographical expanse during the Persian and early Hellenistic periods.53 Because of this stability, and the fact that many Official Aramaic texts are signed and dated, the dialect provides an excellent anchor for the comparison of various other forms of Aramaic, both earlier and later. In Biblical Aramaic, the Aramaic letters of Ezra are typically characterized as being written in good Official Aramaic, while the Aramaic portions of Daniel exhibit some small indicators of subsequent development from the imperial standard.54 Turning back to the Qumran scrolls, we can see that, like Daniel, some things differ from Official Aramaic, though the two groups of texts still share many genetic similarities. The differences speak to incremental changes in pronunciation and writing practice that were taking place as highly-trained Jewish scribes plied their craft in Official Aramaic to the composition and transmission of new national literature. In what follows, I give an overview of various areas of spelling traditionally falling under the grammatical categories of orthography and phonology, focusing on where changes in pronunciation and writing practice are manifest in relation to Official Aramaic on the one hand, and Aramaic dialects contemporary with or subsequent to Qumran Aramaic on the other.

Generally speaking, the Qumran Aramaic texts tend to be full (or plene) with regard to spelling, providing matres lectionis of various sorts not found in most Official Aramaic texts, and distinct in some ways from the full spelling systems in later dialects, such as those of Targum Onkelos and Targum Jonathan, Christian Palestinian Aramaic, or the Jerusalem Talmud. In many cases, the full spellings in Qumran Aramaic are associated with the graphic representation of vowels, and my survey will begin with this phenomenon.55 These spellings were presumably connected with the reading tradition, and are thus often taken to be an indicator of how Aramaic was pronounced at the time and place in which the texts were composed or copied.56 However, it is very important to recognize at the outset that not all the scribes of the Qumran Aramaic texts employed the full spelling system to the same extent, and that, moreover, we find differences in this regard between copies of the same work, not just two different compositions. Good examples of different spelling preferences in two copies of the same work are 4Q530 (EnGiantsb) versus 4Q531 (EnGiantsc), 4Q113 (Danb) versus 4Q112 (Dana) and MT Daniel, and 4Q554a (NJb) versus the other New Jerusalem copies. These texts prove beyond doubt that full or defective spelling is not an accurate tool for dating when a text was composed. Still, the vast majority of Qumran Aramaic texts do use full spellings in a fairly consistent way, so that it is justified to call it a general trait of the corpus, with occasional exceptions. Even these exceptions could be considered to fall within the realm of scribal preference described above. The following list surveys the full spellings most commonly used to represent vowels in Qumran Aramaic, though any given copy will have different constellations of agreement or disagreement with the items on this list. As with Qumran Hebrew, or indeed most other ancient languages, we should not expect consistency in spelling.57

  1. The letter א is used to represent vowels in a variety of situations:58

    1. We find א internal to words of different sorts, often combined with other letters to signify the vowel (e.g., ראיש “head” [1Q20 (apGen) 14.9, 4Q566 (Prophecyc?) 1.4], באיש “bad” [4Q203 (EnGiantsa) 8.14], אסיאנהון “their physician” [4Q548 (VisAmramf) 1–2ii.3], ראסין “stadia” [4Q554 (NJa) 1i.15], קאם “standing” [4Q542 (TQahat) 1ii.4]). For some of these words the א is an original, etymological feature, but for others it is not. Cook noted that א is used after the letters ו and י (e.g., עואן “iniquity” [4Q560 (Magical Booklet) 1.4]) to distinguish a word from others with the same written form (i.e., homographs), something rare outside of the Qumran texts.59 It is also found separating vowels at the boundary between morphemes, as in גנואין “shame” (4Q541 [apocrLevia?] 9i.6). Some spellings of words with etymological א are found regularly in earlier Official Aramaic texts (e.g., באיש), but such spellings without the etymological connection are uncommon (e.g., קים versus קאם in Qumran Aramaic). Both kinds of spelling are found in the third- to second-century BCE scribal exercises on clay bowls from Maresha (באיש, ‪דארה‬), and to a more limited extent in the later Judean Desert documents.60 Most later Aramaic texts, such as Targum Onkelos and Targum Jonathan, lose the internal א, etymological or otherwise (e.g. ריש, ‪ביש‬, בתר). Generally speaking, this feature is more pronounced at Qumran than in earlier groups of texts, and wanes noticeably in later dialects. Though the sample size is currently very small, the Maresha texts are important, since they show that these forms were present in Palestine as early as the third century BCE.61 It is also noteworthy that similar full spellings with א are found in Qumran Hebrew.62

    2. א is also used at the end of words finishing in a vowel, such as הווא “they were” (e.g., 4Q530 [EnGiantsb] 2ii+6–12(?).15), שגיא “much many” (e.g., 1Q20 [apGen] 20.7, 4Q196 [papToba] 2.12), and היא “she” (e.g., 4Q197 [Tobb] 4ii.17, 4Q552 [Four Kingdomsa] 2iii.1). Cook suggests that this serves a graphic purpose (some call it a digraph), and as with the examples above it seems tied to marking vowels.63 While it is also common in Qumran Hebrew, it is rare elsewhere.64 A noteworthy full spelling of this sort, uniquely shared by Biblical Aramaic and Qumran Aramaic against other dialects, is גוא “interior, within.”65 We would rather have expected the גו of surrounding dialects, both earlier and later.

  2. The letter ו is frequently employed to represent /o/ and /u/ vowels:66

    1. Other Aramaic corpora, both earlier and later, also use ו to signify the final and medial long vowels /o/ and /u/. However, the high frequency of this practice stands out in a large majority of Qumran Aramaic manuscripts, and separates it in particular from the more defectively-spelled, earlier Official Aramaic texts. Generally speaking, in the Masoretic tradition Ezra and Daniel are also written more defectively than most Qumran Aramaic scrolls. Part of the increased usage in Qumran Aramaic is due to the fact that many of the scribes marked etymologically short vowels with ו.‪67‬ This includes, most notably, the word כול “all,” but also many other words (e.g., אורחא “road” [4Q196 (papToba) 13.1], אומ[ה “people” [1Q72 (Danb) 2.4], עולימא “youth” [1Q20 (apGen) 2.2], and the second masculine singular suffix תון–). Again, this also happens in Qumran Hebrew, and becomes more frequent in later Jewish and Christian Aramaic dialects.68

  3. The letter י marks many /e/ and /i/ vowels:69

    1. As with ו, full spellings with י are not restricted to Qumran. Yet, the extent to which that letter is used for /e/ and /i/ vowels at Qumran is more extensive than in most earlier Aramaic corpora. At the same time, it is not as widespread as in some later text groups. For example, the etymological short /i/ in the first syllable of a word (e.g., at the beginning of the passive-reflexive stem איתפעל) is not represented with י in the Qumran scrolls, but is present in the later Aramaic texts from the Judean Desert (e.g., איתפרע “I will take vengeance” [Nahal Hever 50.9]), and the Cairo Geniza copy of the Aramaic Levi Document (e.g., ‪מינהון‬ [7:5]). We do find י marking short /e/ or /i/ vowels in later syllables of the pael and aphel conjugations in some texts at Qumran.70 As in Qumran Aramaic, י marking short /e/ or /i/ vowels in Qumran Hebrew is quite restricted.71

A different kind of full spelling takes place with the appearance of the letter נ in a limited number of grammatical situations, a phenomenon that may again be tied to pronunciation.72 This נ can, in fact, appear to be preserved in the case of a word containing an original, etymological נ in the root that in other dialects would be assimilated into the word (as with the prefix conjugation of נת״ן and נפ״ק), or added in a word with gemination, the doubling of a consonant (a process sometimes referred to as nasalization, nunation, or dissimilation by means of נ), though it may be that the underlying phonological process is the same in both of these cases.73 Examples are ינפק “he will go out” in 4Q201 (Ena) 1i.5 (compare with יפוק in 1Q20 [apGen] 22.34; this process is particularly seen in the prefix conjugation of פ״ן verbs) and מנדע “knowledge” in 4Q213a (Levib) 1.14 (compare with מדע in 4Q212 [Eng] 1iv.13). This sort of dissimilation by way of נ varies in Qumran Aramaic from copy to copy, and even within individual texts, two good examples being 1Q20 (apGen) and 11Q10 (Job). However, the נ is present on a fairly regular basis, and in most cases is comparable to the patterns of assimilation and dissimilation in Official Aramaic and Biblical Aramaic.74 The second century CE documents from the Judean Desert still show mixed use of dissimilation by way of נ, though in later Jewish and Christian dialects there is a noticeable decline in such usage, albeit with differing rates and patterns of change depending on the text. In a few other cases, certain words may assimilate letters, but these tend to be less noticeable and frequent than assimilation or dissimilation of נ. One of the more prominent cases is the assimilation of ל in certain conjugations of סל״ק “go up,” or the assimilation of the second ל in על״ל, both of which are also found in Official Aramaic and Biblical Aramaic.

Another type of augmentation by נ is at the end of a small group of words that, without the נ, end in an open vowel (א or ה), as discussed already by Kutscher.75 The main examples of this are תמה “there” (always in Ezra [4x] and Official Aramaic) and כמה “how?” (not present in Official Aramaic, and two times in Daniel), which in Qumran Aramaic also appear as תמן and כמן. In Qumran Aramaic, only 4Q529 (Words of Michael) and 11Q10 (Job) use תמה, and all other texts תמן (20x). The evidence for כמן is less striking, with nearly twenty occurrences of כמא/כמה in eight manuscripts, including a number in 1Q20 (apGen), and only three instances of כמן, all of them also in 1Q20 (apGen).

In contrast to the characteristically full spellings of Qumran Aramaic, there are occasionally defective spellings that suggest some distance from Official Aramaic. An example of this is the word כען “now,” which is consistently spelled this way in Qumran Aramaic’s forty-two occurrences, but fluctuates between the fuller form כענת and defective כען and כעת in Official Aramaic. In Biblical Aramaic, Daniel has only כען (7x), while Ezra is mixed between כענת (3x), כען (6x), and once כעת. There has also been discussion of what linguists call monophthongization, the collapsing of a dipthong into a single vowel, for the Official Aramaic phonemes /aw/ and /ay/, which some suggest contracted to /o/ and /e/ in Biblical Aramaic and Qumran Aramaic. If this were the case, it would show phonological development from Official Aramaic to Biblical Aramaic and Qumran Aramaic, but the issue is not as clear cut as some suggest. The idea began with Beyer, based on guesswork about spoken vowels drawn from the consonantal texts. It moved from there to Cook, Muraoka, Gzella, and others.76 The problem is that the evidence is, in fact, very thin or non-existent for Qumran Aramaic, with the possible exception of the hollow-root participle using א (e.g., קאם), to be discussed further below.77 This contrasts with the later Judean Desert documents, which do contain multiple consonantal indicators that seem to suggest the monopthongization of some diphthongs. Indeed, this comprises a significant orthographic or phonetic departure from Biblical Aramaic and Qumran Aramaic. For the latter two text groups, clear evidence for the phenomenon is negligible.78

Some spelling changes do, however, demonstrate differences in how the language was read or spoken phonetically from the Official Aramaic to the Biblical Aramaic and Qumran Aramaic corpora. These changes have to do with a more broadly attested set of phonological shifts taking place over the course of the Persian, Hellenistic, and Roman periods. The main examples are listed below:79

  1. ז‬ changed to ד80

    During the floruit of Official Aramaic, there was an interdental consonant /ḏ/, for which the tongue was apparently set somewhat behind the teeth, producing a sound slightly harder than /z/, but softer than /d/. In Official Aramaic, the consonant is spelled predominantly with the “softer,” more spirantized character ז, but already we can see it hardening – or dentalizing – in the occasional spelling with the full dental ד.‪81‬ Good examples of this are זי “that, which,” זנה “this,” אחז “seize, hold onto,” זהב “gold,” and זחל “to fear,” each of which is sometimes spelled with ד instead of ז in Official Aramaic, in certain cases quite often. In Biblical Aramaic and Qumran Aramaic the situation is essentially reversed. In these texts ד predominates and ז is an occasional or absent spelling (e.g., regularly די, ‪דנה‬/דן, ‪אחד‬, דהב, and דחל). The published third-century BCE Maresha bowls still have ז,‪82‬ but there are hundreds of unpublished bowls and ostraca from that site dating to the same period, and Esther Eshel has confirmed that these show mixed spellings, which include the prefixed –ד.‪83‬ This is very important, since it shows that this “late” form was already being used in Judea in the third century. Somewhat surprisingly, the later Judean Desert documents contain many more ז spellings than Biblical Aramaic or Qumran Aramaic, even though we know they come from a later period.84 In later Jewish and Christina dialects, however, the transition to ד reaches its completion.

  2. ק changed to ע, and ע changed to א85

    There is a long-term shift in pronunciation of a proto-semitic velar or pharyngeal consonant formed with the back of the tongue (more forward) or pharynx (slightly farther back) that is represented by the “harder” letter ק in the earliest Aramaic texts and continues on into Official Aramaic. This is seen in the Official Aramaic words עק “wood” and לעבק “with haste.” Already in Official Aramaic the early velar ק started to shift to the more pharyngeal ע, as seen in the mixed usage of ארק and ארע for “land, earth,” a process that is more advanced in Biblical Aramaic and Qumran Aramaic (where only ארע is present). We see this phonetic progress, for example, in Official Aramaic לעבק now being written לעבע (11Q10 [Job] only) or לעובע (3x) in Qumran Aramaic. In like manner, Official Aramaic עק is now sometimes written as עע (4x).86 This word allows us to see a second shift, from ע to the weaker laryngeal א, since in Biblical Aramaic (5x) and in Qumran Aramaic (twice) we find instead the spelling אע.‪87‬ The development of this word thus shows nicely the process of change: ‪עק‬ > ‪עע‬ > אע.‪88‬ In terms of external evidence, the fourth-century BCE ostraca from Idumaea published by Ephʿal and Naveh still have עק, while אע is found in the third- to second-century BCE Maresha bowls, showing that the shift to that form had already begun by that time.89 The latter part of this process is also caught in motion among the Qumran Aramaic texts with the verb חע״ך “to laugh,” which is found in 1Q67 (Unclassified frags.), 4Q543 (VisAmrama), and 4Q544 (VisAmramb), but six times as חא״ך in 11Q10 (Job). A further extension of these shifts can be seen in later Jewish and Christian dialects, where Biblical Aramaic עלע “rib” changes to אלע, Qumran Aramaic עב״ע changes to אב״ע, and Qumran Aramaic ער״ע “to meet” shifts to אר״ע.‪90

  3. ש and ס

    A number of words that originally contained an/ś/ sound rendered in Official Aramaic by the letter ש show the slight beginnings of a shift to ס in Qumran Aramaic that would eventually predominate in later Jewish and Christian dialects.91 Presumably this was because the phonetic value of ש and ס were equalizing during the Second Temple period.92 While it is true that we find some spellings with ס at Qumran (e.g., עסר in 4Q201 [Ena] 1ii.5 vs. עשר in 4Q209 [Enastrb] 2ii.6), and that nearly twenty lexemes show this interchange in Qumran Aramaic, the texts overwhelmingly have the older spellings with ש, as also in Biblical Aramaic. It should also be noted that certain Qumran manuscripts, such as 4Q201 (Ena) and 11Q10 (Job), show a stronger preference for ס, sometimes disagreeing with conventions in other copies of the same text. The case is quite different for the second-century CE Judean Desert documents, which exhibit a noticeably increased use of ס in the relevant situations.93 Later Jewish and Christian Aramaic dialects display a strong move to ס.‪94

  4. ם changed to ן

    Another noticeable change from Official Aramaic is the phonetic and graphic movement from מ to נ in some situations.95 These are mostly found in independent pronouns and pronominal suffixes, such as Official Aramaic ‪אנתם‬, כם–, and הם–, versus Qumran Aramaic with mostly ‪אנתון‬, כון–, and הון (note again the full spellings with ו).96 There is also Official Aramaic הם/המו, for which Qumran Aramaic has mostly אנון.‪97‬ Biblical Aramaic is caught between the shift, with both Ezra and Daniel having some of each type (see the section on pronouns, below), though Ezra leans toward Official Aramaic and Daniel more toward Qumran Aramaic for this characteristic. The second-century documents from the Judean Desert still use some of the forms with מ, but later Jewish and Christian dialects switch over completely to נ.‪98

Along with these phonetic shifts, many Qumran Aramaic manuscripts show a widespread preference for א over ה, which seems to imply a weakening, or quiescence, of ה.‪99‬ In other cases, the letters are simply interchanged freely, with no apparent pattern for which one is chosen, and this suggests phonetic equalizing. Preference for א is seen most starkly in the prefix for the causative conjugation, which is mainly אפעל in Qumran Aramaic, rather than the הפעל of Official Aramaic and Biblical Aramaic.100 There are, however, plenty of exceptions to the rule at Qumran, and Cook opines that the choice of א or ה “seems to have depended on the whim of the scribe.”101 11Q10 (Job) and 4Q529 (Words of Michael) typically or always use the הפעל form, and usage is mixed in manuscripts such as 4Q196 (papToba) and 4Q542 (TQahat). We may pair this with Muraoka’s further observation that the representation of א or ה in the causative stem at Qumran is largely restricted to the imperative, infinitive, and suffix conjugation forms, but is rare as an infix in the prefix-conjugation and participles. By contrast, Biblical Aramaic often (though not always) has the ה infix in these latter two verb forms (e.g., יהקם in MT Dan 5:21; מהקים in MT Dan 2:21).102 For Muraoka, we thus witness “a diachronic shift in progress” at Qumran. Looking outside of Qumran, we see that the fourth century BCE ostraca from Idumaea still use הפעל.‪103‬ Later dialects commonly have the אפעל spelling. A similar situation obtains for spellings of the passive/reflexive conjugation, אתפעל and התפעל, though here the surrounding chronological evidence is different: As opposed to the use of הפעל in Official Aramaic, that dialect always uses אתפעל for the passive/reflexive conjugation, as do the later Jewish and Christian dialects.104 In other words, Biblical Aramaic and some Qumran Aramaic texts (e.g., 11Q10 [Job]) stand out for their idiosynchratic use of the התפעל, while most Qumran Aramaic texts fit better with Official Aramaic and later forms of Aramaic.105 The other major places where we see the exchange between א and ה are: 1.) Representing the final vowel of ל״א verbs or other words ending with open vowels, such as the independent pronoun אנה/אנא or demonstrative pronoun דנה/דנא;‪106‬ 2.) For the definite article at the end of nouns to indicate the determined state (overwhelmingly א [e.g., מערבא “west” in 1Q20 (apGen) 17.10], but occasionally ה [e.g., מערבה in 2Q24 (NJ) 4.10]); and 3.) For the sufformative /ā/ of the feminine noun in the absolute state (typically ה [e.g., אנתה “woman, wife” in 4Q544 (VisAmramb) 1.8], but sometimes א [e.g., אנתא in 4Q197 (Tobb) 4i.13]). The firm rule in Official Aramaic is א for the definite article and ה for the feminine noun sufformative. While the Qumran Aramaic texts still largely adhere to this rule, there are certainly more exceptions than in Official Aramaic (and Biblical Aramaic). In some cases, however, it is clear that the divergences are based on the preferences of an individual scribe, as we also find occasionally in Official Aramaic.107

Noteworthy differences between full and defective spellings are also found among the endings of pronouns, pronominal suffixes, and demonstrative pronouns.108 These forms are often referenced by those situating Qumran Aramaic among other dialects, and so they merit brief treatment here. For ease of discussion, the most common Qumran Aramaic forms are provided, with alternative Qumran spellings separated by a slash, and relatively infrequent or singular Qumran spellings in parentheses.

Some of these forms, used consistently across Qumran Aramaic and Biblical Aramaic, distance the Qumran texts from other corpora. Both earlier and later dialects also use אנה and אנחנה (or אנחן), but later Jewish and Christian dialects witness the growth of new apocopated forms for אנחנה, such as אנן and חנן. The main thing distancing Qumran Aramaic from surrounding dialects is the use of a final א in אנחנא, which is purely orthographic and varies in Qumran Aramaic and Biblical Aramaic (both Daniel and Ezra). More significant is the second person masculine אנתה, also the usual ketiv spelling in Daniel (though not Ezra), which is a full form missing from other, surrounding Aramaic text groups. All other known Aramaic texts, including the later Judean Desert documents, use shorter forms such as אנת (the Official Aramaic form) and את. Some debate has taken place over whether the Qumran Aramaic/Daniel form represents a long or stressed /a/ vowel already present in Official Aramaic, but not written there, or signifies a novel development in the pronunciation behind the Qumran Aramaic/Daniel form, perhaps influenced by Hebrew.109 אנתון is a full spelling in Qumran Aramaic and Biblical Aramaic distinguished from the Official Aramaic אנתם on one hand, and אתון of later Aramaic corpora on the other. The Official Aramaic spelling of the third-person masculine singular pronoun is הו, and the feminine הי, though some Early Aramaic texts and the fifth-century BCE Sheikh Fadl inscription from Egypt have הא or הוא/‪היא‬. The latter forms are the ones found almost exclusively in Qumran Aramaic, Biblical Aramaic, and subsequent Jewish dialects, with the exception of 4Q550 (Jews at the Persian Court) 7+7a.1 (הו) and 4Q204 (Enc) 5ii.30 (הואה) among the Qumran scrolls, and the noteworthy use of הו and הי in the later Judean Desert documents. The spelling הואה in 4Q204 (Enc) is taken by Fassberg, Schattner-Rieser and others to be a Hebraism,110 and some have suggested that the switch from earlier הו/הי to הוא/היא is due, at least in part, to the influence of Hebrew orthography.111

As with אנתון, Qumran Aramaic and, to a slightly lesser extent, Biblical Aramaic witness a shift with the third person plural אנון and אנין, the latter being attested only once (4Q201 [Ena] 1iii.15). The older Official Aramaic forms were המו (or הם) and הני (one occurrence), of which only the masculine form has come through to Qumran Aramaic and Daniel as המון. This earlier form is used exclusively in 11Q10(Job) and what little is preserved of 4Q242 (PrNab). The distribution between המון and אנון is split in Biblical Aramaic (three times each), but the later אנון predominates in the remaining Qumran Aramaic texts, with over forty occurrences. Some of the other, post-Official Aramaic dialects use forms beginning with ה,‪112‬ and it of interest that the second century CE Nahal Hever documents still use המון frequently alongside אנון. אנון or related forms (e.g., אינון, ‪הנון‬) are found in later Jewish and Christian Aramaic. Cook suggested that המון was used only for the direct object in Biblical and Qumran Aramaic, though in fact the closely related form המו is used as a plural subject in Ezra 5:11, and the same is true of המון at 11Q10 (Job) XXVIII.2.113 Furthermore, אנון is clearly used as a direct object in Ezra 5:4 and Dan 6:25, as it is in Qumran Aramaic.114 Taking all of this evidence into account, we should be careful of assigning specific syntactic roles to אנון and המון – they seem to be used arbitrarily based on the preferences of a given scribe. In light of the relatively late use of המון in 11Q10 (Job) and the Nahal Hever documents, we should also avoid dating texts with great specificity based on this criterion.

The first-person suffixes are unremarkable in relation to surrounding dialects, except that the full נא– ending of the plural in Qumran Aramaic and Biblical Aramaic (fluctuating between א and ה in both) is instead mixed in Official Aramaic between נא– and the defective form ן–.‪115‬ The spelling of the second-person suffix across the range of dialects is typically the short ך– of Official Aramaic, but the full spelling כה– (כא– in 1Q20 [apGen] 2.17, 5.9) occurs fairly regularly in a cross-section of Qumran Aramaic texts. This full ending is a distinctive trait of the Qumran Aramaic manuscripts, and as with אנתה there has been debate over whether this is an idiosyncratic graphic representation of an original Official Aramaic final vowel /a/, or a Hebraism.116 Although the short form ך– is always used in Biblical Aramaic of the Masoretic tradition, we saw above that this is not true of 4Q113 (Danb). The long feminine singular suffix כי– is present in Official Aramaic along with the short form ך–, though in most later Jewish Aramaic dialects (with the exception of Christian Palestinian Aramaic) a short feminine form ך– is the norm.117 The non-Qumran Judean Desert documents are mixed; Qumran Aramaic stands out among the other dialects for its exclusive use of the long form.118 As noted by Kutscher, the long Aramaic form also influences Qumran Hebrew, where it sometimes appears as an Aramaism.119 Masculine plural כן–/כון– was usually כם– in Official Aramaic, with an occasional כן–/כון– showing that the transition from mem to nun was already underway in fifth-century BCE Egypt. For the most part, we find the later endings with nun at Qumran, though 1Q20 (apGen) has the ending כם– at least once (8.16), perhaps twice (10.7?). It seems to me best to interpret this as an archaism in Aramaic rather than a Hebraism, though either interpretation is possible.

The third person singular masculine ה– (after a consonant) or הי– (after a vowel, והי– on masculine plural nouns, as well as prepositions such as על and קודם) of Qumran Aramaic and Biblical Aramaic is the same as in Official Aramaic and the later Targum Onkelos/Targum Jonathan dialects.120 Some other later dialects have shortened forms of והי–, such as Jewish Palestinian Aramaic וי– and Samaritan Aramaic יו–. One of these is found at Qumran in 11Q18 (NJ) (עלוי; 8.3, 9.4), a copy dating to the Herodian period, and so relatively late in view of the wider corpus. Forms from dialects further east are different. The shorter feminine singular suffix ה– is the standard in Official Aramaic and Biblical Aramaic.121 Although הא–/הה– does appear a few times in Official Aramaic texts, and regularly in Targum Onkelos/Targum Jonathan after a vowel, this longer ending is much more widespread in Qumran Aramaic than in Official Aramaic, and also appears in more grammatical situations than in Targum Onkelos/Targum Jonathan.122 Other later dialects have only ה–. Scholars have compared the long feminine singular ending to the second person כה–, and an analogous debate has taken place over its origins, whether it represents an original Aramaic vowel (Kutscher, Cook, Fassberg) or is a Hebraism (Fitzmyer). As for the third person plural forms, we find a roughly analogous situation to that of the independent המון and אנון. The typical Official Aramaic forms are masculine ‫–הם/–הום‬ and perhaps feminine הן–, though the masculine ending does occasionally have nun rather than mem.123 As with כם– and ‫–כן/כון‬, the evidence of ‫–הם/–הום‬ and ‫–הן/–הון‬ illustrates the halting transition from mem to nun beginning to take place already in Official Aramaic. In Biblical Aramaic, Ezra has the Official Aramaic הם– and Daniel הון–, followed by most Qumran Aramaic texts. Some mem forms persist into later periods, as with מנהם in 4Q112 (Dana) (MT Daniel מנהון), אלהיהם in 4Q570 (Unid. Text D) 6.6, and more conspicuously the later Nahal Hever documents.124 The full spelling of the feminine suffix (הין–; e.g., 4Q531 [EnGiantsc] 1.4) is found in Qumran Aramaic in addition to the defective form.

The singular masculine short demonstrative דן makes its first appearance in Qumran Aramaic, signalling a clear development from Official Aramaic זנה/זנא (rarely דנה/דנא) and Biblical Aramaic דנה, although the short form זן does occur in a variety of inscriptions predating or contemporary with Official Aramaic.125 The Qumran Aramaic דן is a forerunner of the later Jewish and Christian Aramaic forms דין (when a sentence subject), הדין (when an adjective), and דנן.‪126‬ It may be that a single occurrence of הדן is found at 4Q544 (VisAmramb) 2.2.127 Although far surpassed in number by דן, the longer form דנה does occur in thirteen manuscripts at Qumran, in some cases (e.g., 1Q20 [apGen] and 4Q537 [TJacob?]) in mixed usage with the short form. The later Judean Desert documents have a surprising mixture of זנה, ‪דן‬, דנן, and (mostly) דנה/דנא.‪128‬ The Qumran Aramaic and Biblical Aramaic far demonstrative דך (Ezra; דכן in Daniel)129 is mixed between זך, ‪דך‬, and זנך in Official Aramaic. We find דך and זך alongside the independent forms הוא and היא serving as demonstratives in the later Judean Desert documents, while subsequent Jewish and Christian dialects shift further to using הוא and היא.‪130‬ Aside from the interdental to dental shift from ז to ד, the feminine דא matches the Official Aramaic זא, which also has once דה.‪131‬ Like the masculine form, it later shifts to דא (subject) and הדא (adjective) in Jewish and Christian Aramaic. The plural near demonstrative אלה, standard in Official Aramaic, may occur once at Qumran in 4Q536 (Birth of Noahc) 2ii.12, and twice in Biblical Aramaic (Jer 10:11, Ezra 5:15).132 The Qumran Aramaic and Biblical Aramaic form is otherwise always אלן/אלין. The other Judean Desert texts witness to both אלה and אלין, though it should be remembered that Nabatean preserves אלה.‪133‬ Later Jewish and Christian Aramaic dialects used forms similar to, or developed from, Qumran Aramaic and Biblical Aramaic, such as אלן, ‪אלין/אליין‬, and הלין/הילין. The far demonstrative plural אלך is stable across a wide spectrum of dialects encompassing Official Aramaic (along with the alternate אלכי), Qumran Aramaic, Biblical Aramaic, and the later Judean Desert documents. Subsequent Jewish and Christian forms are אלוך and הליך, or simply the independent pronoun (אינון, ‪הנון‬ or the like). At Nahal Hever, and perhaps once at Qumran, we also find the independent אנון used in a far demonstrative sense.134

5.2 Formation of Verbs and Nouns (Morphology)

The formation of verbs and nouns is another area where elements distinguishing Qumran Aramaic from surrounding dialects may be recognized. A few of these features appear to be part of the same phonetic/orthographic phenomena discussed above for the pronouns, and so will be dealt with first.

In the suffix conjugation there are four sufformatives that warrant comment:135

  1. The second-person masculine singular ending is often the short form ‫–ת‬ found in Official Aramaic (e.g., חזית “you saw” 1Q20 [apGen] 14.17), though a longer ‫–תה/–תא‬ suffix is found a number of times across a range of copies (e.g., ‪חזיתא‬ 1Q20 [apGen] 14.14). Biblical Aramaic shows this same mix, with long endings at Dan 2:41, 4:14, and 5:27. While later Jewish and Christian dialects tend to use the short form, the long form is also attested in some places.136 As with the pronoun אנתה and pronominal suffixes ‫–כה‬ and ‫–הא‬, scholars have debated the possibility of an original Aramaic vowel versus Hebrew influence.137

  2. The second-person masculine plural verb suffix in Official Aramaic is mixed between ‫–תם‬ and ‫–תן‬/‫–ת ון‬.‪138‬ In Biblical Aramaic and Qumran Aramaic, however, it is always ‫–תון‬ (rarely written defectively as ‫–תן‬), as is the standard in later dialects. Here we see the same basic shift from ם to ן endings, as in several of the pronouns.

  3. In Official Aramaic and the consonantal text (ketiv) of Biblical Aramaic the third-person singular feminine verb suffix is the shared masculine ‫–ו‬.‪139‬ Several Qumran Aramaic manuscripts, however, attest to a distinctive feminine ending ‫–א‬ (e.g., שלמא [1Q20 20.6], הויא [4Q202 1ii.2]).140 This same ending is found in later Jewish dialects, such as that of the Onkelos and Jonathan targums.141

  4. The first person common ending is ‫–ן‬ in Official Aramaic, but in both Biblical Aramaic and Qumran Aramaic we find uniformly the longer ending ‫–נא‬, and we may note that the same is found in the third to second century BCE Maresha bowl inscriptions.142 ‫–נא‬ is also used in the subsequent Jewish and Christian dialects.

One of the most striking traits of the verbal system used only in Biblical Aramaic, Qumran Aramaic, and the other Judean Desert documents, is an idiosyncratic practice of using a –ל prefix for the third person singular and plural forms of the verb הו״א “to be” in the prefix conjugation (e.g., להוה, ‪להוון‬, להוין).‪143‬ This practice is unknown in Official Aramaic, and some have posited that Biblical Aramaic and Qumran Aramaic inherited an old –ל jussive-precative prefix known from a group of Early Aramaic inscriptions influenced by Akkadian, which they put to their own use as a stand-in prefix to avoid verb forms that may be confused with the Tetragrammaton.144 As Schattner-Rieser observed, the fact that the Tetragrammaton is never otherwise used to refer to God in Qumran Aramaic corroborates this probable avoidance.145 Recently, Fassberg has proposed a completely different line of development, in which avoidance of the Tetragrammaton does not play a part.146 Whatever the case, the old prefix, were it indeed known, was available for all verb stems. Yet in Biblical Aramaic and Qumran Aramaic it was restricted to the root הו״א.‪147‬ Moreover, Biblical Aramaic and Qumran Aramaic may use the –ל prefix in an indicative sense, unlike the old Early Aramaic and Akkadian jussive-precative prefix. Although a –ל prefix was used in Late Eastern Aramaic dialects, it was not restricted there to הו״א, and Fassberg is justified in distancing this from the Biblical Aramaic and Qumran Aramaic usage.148 The practice is not found in later Jewish Aramaic dialects, and thus serves as a truly distinctive feature of Biblical Aramaic, Qumran Aramaic, and the other Judean Desert Aramaic texts.

Some differences between Qumran Aramaic and the later Judean Desert documents have been indicated above, but two further items should be added to these. First, Aaron Koller noted an interesting discrepancy in the morphology of the אתפעל/התפעל conjugation between Official Aramaic, Biblical Aramaic, and Qumran Aramaic on the one hand, and the second century CE Judean Desert documents and Nabatean on the other.149 In the former group of texts we find the first letter of verbal roots beginning with ש, ‪ס‬, and ז (i.e., a sibilant) undergoing metathesis with the infixed –ת– of the conjugation, as in ‪אזדמנו (זמ״ן‬, 1Q20 [apGen] 21.25), ‪הסתכל (סכ״ל‬, 11Q10 [Job] 24.5), and ‪משתפך (שפ״ך‬, 4Q533 [EnGiantse] 4.2). This is the expected norm across almost all Aramaic dialects, with the exception of Nabatean and the Judean Desert documents, which show “a consistent lack of the expected sibilant metathesis for [H]ithpe‘el verbs.”150 Examples from the Judean Desert texts are אתזבן (זב״ן, P. Yadin 7:16), התשדר (שד״ר, P. Yadin 53:3), and יתשכח (שכ״ח, P. Yadin 54:10).151 Muraoka has also noted the unusual full spelling איתפרע “exact punishment” in Nahal Hever 50.9, though not much can be made of one irregularity.152

The second difference is seen in the morphology of the infinitive. In the dialects up to and including Qumran Aramaic, the standard way of forming the infinitive of the base peal (G) stem was with a prefixed ל and מ, examples being למדחל “to fear” (4Q198 [Tobc] 1.1) and למעבד “to do” (4Q212 [Eng] 1iv.16).153 For the derived stems there is typically a ל prefix and a א/ה suffix, as in לאלפה “to teach” (4Q201 [Ena] 1iii.15) and להלבשה “to clothe” (11Q10 [Job] 29.7). However, already in Official Aramaic, and in Qumran Aramaic as well, we occasionally find derived-stem forms that include a prefixed מ, for example with למתיה “to bring” in Official Aramaic (את״ה, TAD A2.4:11) and at Qumran למחזיא (4Q542 [TQahat] 1ii.6) and למעמרא (4Q544 [VisAmramb] 1.1; though, importantly, לעמרה in the parallel at 4Q545 [VisAmramc] 1a–bii.13).154 This latter combination becomes the norm in most later Jewish and Christian dialects, and in the Judean Desert documents we see an undeniable increase in use of this form when compared to Biblical Aramaic and Qumran Aramaic (e.g., למשפיה, ‪למעמקה‬, and למעבדא).‪155‬ Moreover, Muraoka and Schattner-Rieser note some other innovative infinitive forms in these texts, like למפרוע (Nahal Hever 7.17, 57), a construction typical of later Jewish dialects, and למנעלו (Nahal Hever 7.26, 57), identical to the Syriac pattern.156 It seems, then, that we see a clear development in the morphology of the derived stem infinitives in some of the Judean Desert documents, similar to forms that will become well-established in later dialects.

Steven Fassberg has drawn attention to another full spelling that appears to reflect a morphological or phonological shift taking place in some Qumran Aramaic manuscripts.157 He cited five instances where a peal imperfect form is spelled with a long vowel, graphically representing /o/ or /u/ in the final or penultimate consonant (e.g., ינטור in 4Q534 [Birth of Noaha] 7.4) instead of the expected /a/ vowel inferred for Official Aramaic (i.e., ינטר).158 Muraoka has added a number of other occurrences, though some of these are from the later Judean Desert documents and Cairo Geniza copy of the Aramaic Levi Document, and so should not be considered as Qumran Aramaic in a straightforward way.159 The longer form is that known from later Jewish and Christian dialects, such as Christian Palestinian Aramaic.160

The use of א in some situations to indicate the internal vowel of hollow-root verbs was already noted in the section on orthography and phonology above, and scholars have debated the extent to which this is a morphological change. These forms, all found in the participle construction, are present in Biblical Aramaic and Qumran Aramaic (e.g., ‪קאם‬, Dan 2:31; קאמיא, Dan 7:16; דארין, Dan 3:31; ‪קאם‬, 2Q24 [NJ] 4.16; ‪דאלין וזאעין‬, 1Q20 [apGen] 0.7; ‪דאנין‬, 4Q544 [VisAmramb] 1.10; ‪ד̊אר̊‬, 4Q209 [Enastrb] 23.3).161 Little evidence for the relevant forms is preserved in the later Judean Desert texts, but the א participle is well known from later Jewish and Christian dialects. When we look at Official Aramaic, however, we find no occurrences of the participle with א representing the internal vowel, but rather י only.162 Beyer and others after him assumed that this implies a process of monopthongization, combining the double vowel of a diphthong /ay/ into one sound /e/.163 In fact, there is no way of being sure that this is the case, and Schattner-Rieser adopts the alternative explanation that the א simply serves the purpose of distinguishing the active and passive participle forms.164 Whatever the phonetic reality was behind the graphic change, this feature constitutes a clear difference between Official Aramaic and Biblical Aramaic/Qumran Aramaic.

Finally, there are two verbal lexical items that deserve mention. The first pertains to the words signifying one’s ability to do something, “to be able,” כה״ל and יכ״ל. In Official Aramaic, both of these words are used to describe the state of one’s ability, almost exclusively in negative statements: לא נכל נגרך דין ודבב “We shall not be able to institute against you a suit or process …” (TAD B3.4:12–13); לא נכהל נרשנך דין ודבב “We shall not be able to bring against you suit or process …” (TAD B3.12:25). The lexeme כה״ל is used in the Sefire inscriptions and Saqqara papyrus, while יכ״ל occurs already in the Deir Alla inscription, so that both words are present in Early Aramaic. In Official Aramaic, כה״ל is more frequent, with around forty preserved instances, compared to around twenty-five for יכ״ל.‪165‬ In Biblical Aramaic, usage is tilted in the other direction, with all occurrences being found in Daniel: twelve יכ״ל and four כה״ל. One may rightly ask whether there is enough evidence to extrapolate a shift in use from these texts, but it is interesting that in Qumran Aramaic we find only יכ״ל (at least seventeen times in nine texts).166 Neither כה״ל nor יכ״ל appears in the later Judean Desert documents, but כה״ל falls out of use in favour of יכ״ל in all later Jewish and Christian dialects. Biblical Aramaic and, to an even greater extent, Qumran Aramaic appear to signal the earliest stages of this shift.

Fassberg has noted a comparable situation with two Aramaic roots meaning “to go” or “to walk,” הו״ך and הל״ך.‪167‬ Like Early Aramaic and Official Aramaic, Biblical Aramaic and Qumran Aramaic make a basic distinction between the two forms, using הו״ך for the prefix conjugation (imperfect) and infinitive forms, and הל״ך as a suppletive paradigm for the suffix conjugation (perfect) and participle.168 In a response to Fassberg, Jan Joosten averred that it is, in fact, אז״ל “to go” that serves as the suppletive paradigm for הו״ך, not הל״ך, which has a different semantic range.169 While this distinction is generally upheld in the Aramaic of Targum Onkelos and Targum Jonathan, in other Jewish and Christian dialects הל״ך and אז״ל take over the old uses of הו״ך.

Noun morphology in Qumran Aramaic has focused largely on an apparent shift in the vocalization of a tri-consonant noun class realized in Official Aramaic and Biblical Aramaic as qeshut or qeshot (קְשֹט “truth, righteousness,” often referred to as the qtul pattern), with the full vowel placed between the second and third consonants.170 This pattern fits into a more general tendency in these dialects to have the weight of the full vowel placed in the last syllable for the absolute-state noun (e.g., גְבַר, ‪כְסַף‬, לְחֶם, ‪תְקָף‬). While we cannot know with certainty how these words were pronounced when the texts were written – the vocalization of the Masoretes for Biblical Aramaic deriving from a much later period – some full spellings in Qumran Aramaic suggest that the Official Aramaic and Biblical Aramaic pronunciation had shifted to qosht/qusht or, more likely, to a vocalization with two full vowels: qushut, qushot, qoshot, or qoshut.171 Support for qosht/qusht (= qutl) vocalization is seen in preserved spellings such as קושט (1Q20 [apGen] 2.7, 4Q542 [TQahat] 1ii.8), תוקף (1Q20 [apGen] 20.14, 4Q203 [EnGiantsa] 7a.3), אורך (1Q20 [apGen] 11.11, 4Q554 [NJa] 2ii.12), and רושם (4Q530 [EnGiantsb] 2ii+6–12[?].19). Such spellings have at times been ascribed to Hebrew influence, since this is a well-known noun pattern from that language.172 However, Kutscher ultimately rejected a possible Hebrew provenance in favor of an internal Aramaic development. Working from the fact that the more expected pattern from Official Aramaic and Biblical Aramaic also appears in Qumran Aramaic (e.g., תקוף in 1Q20 [apGen] 13.16, 4Q531 [EnGiantsc] 22.3; אנוס in 4Q550 [Jews at the Persian Court] 1.7; and ארוך in 5Q15 [NJ] 1ii.7), sometimes in close proximity to the alternate forms just listed, Kutscher was the first to hypothesize that the implied vocalization must, in fact, be qushut or qoshot, noting similar forms in Qumran Hebrew.173 This full vocalization was confirmed by the publication of 4Q542 (TQahat), which contains the spelling קושוטא at 1ii.1.174 Muraoka rightly judged that there is a low possibility of this spelling being a scribal mistake, reinforcing the impression that a number of the Qumran Aramaic texts reflect a more fully-vocalized noun pattern than in Official Aramaic and Biblical Aramaic, even if the origins of this development remain obscure.175

5.3 Ways of Communicating: Some Larger Syntactic and Idiomatic Traits of Qumran Aramaic

Thus far my overview has focused on the writing system, phonology, and word formation of Qumran Aramaic at the level of individual words or morphemes. However, some of the most interesting features of any language are how it arranges words and clauses in order to form a larger system of written or spoken communication. Attaining a thorough grasp of this topic is significantly more difficult than for words and morphemes alone, since a higher-order study of this sort includes many moving parts, which can be measured in any number of ways. In the case of Qumran Aramaic, it is also hampered by the fragmentary nature of the scrolls. I do not claim here to give anything approaching a comprehensive overview of how Qumran Aramaic communicates in relation to surrounding Aramaic text groups, but simply to offer a sounding of some salient facets of Qumran Aramaic’s “way of speaking” that overlap significantly with the linguistic domains of syntax and semantics. Most of these facets have been discussed already by others. I will begin by treating use of the verb in sentence structure, and then move to other, non-verbal constructions.

A pervasive verbal construction shared by Biblical Aramaic and Qumran Aramaic (along with many other earlier and later dialects, including Official Aramaic) is the so-called periphrastic phrase, in which the verb הו״א “to be” is combined with another verb, usually to convey ongoing action.176 הו״א is always in a finite form, thereby governing the tense of the phrase. It is typically followed by the participle conjugation of the second verb, which defines the specific content of the action (e.g., ואנה הוית יתב “And I was dwelling” 1Q20 [apGen] 21.7; וא]נ̊תון̇ ת̊הוון ממרין ומקשין “And y]ou will be agitating and making trouble” 4Q537 [TJacob?] 9.2). Occasionally, this order is reversed (e.g., חזה הוית “I was looking” 1Q20 [apGen] 13.11; ידיע להוא לכון “Let it be known to you” 4Q203 [EnGiantsa] 8.6).177 הו״א is found rarely in the imperative, in most cases conveying a durative command to “be doing” something habitually (e.g., והוא קד[י]ש̇ין ודכין מן כ̇ל[ ער]ברוב “Be set a[p]art and clean of all[mi]xture” 4Q542 [TQahat] 1i.8–9).178 While Fassberg was right to point out that חזה הוית and ידיע להוא are frozen expressions also found in Biblical Aramaic, his and Muraoka’s assumption that Qumran Aramaic usage must be influenced by Biblical Aramaic (Daniel and Ezra respectively) is not convincing.179 Is it not possible that חזה הוית was a visionary expression used more widely in the Jewish sphere, drawn upon by Daniel and other Qumran texts as a common idiom? The issue should not be judged in advance based on an assumption that Daniel had an authoritative status not enjoyed by some other Jewish Aramaic texts at this time, even if we may, in the end, deem this to be the case. In addition, the question of dating for texts other than Daniel is not yet a settled issue. As for ידיע להוא, that expression is well attested in Official Aramaic (ידיע יהוה, without the Jewish Aramaic –ל prefix on הו״א), especially if we now consider the Bactrian texts from the early Hellenistic period in the Khalili collection.180 Against this wider Official Aramaic background, any argument for the exclusive influence of Ezra loses its force. If there is anything to be gleaned from Qumran Aramaic using the periphrastic construction, it is that it conforms to the situation also found in Biblical Aramaic and Official Aramaic.

A couple of special verbal features are shared by Daniel and 1Q20 (apGen), setting these two works off from the surrounding Aramaic dialects. The first is use of the prefix conjugation as a preterit, carrying the sense of something that is recalled as having taken place in the past.181 This occurs six times in Daniel, always governed by a preceding suffix-conjugation verb or temporal clause indicating action in the past. One example is Dan 4:2: חלם חזית וידחלנניוחזוי ראשי יבהלנני “I saw a dream, and it frightened me … and the visions of my head agitated me.”182 The same thing happens three times in 1Q20 (apGen), the third instance being part of Noah’s dream in column 13: והוית תמה̇ ע̇ל זיתא ד̇ן̇ … [ארבע] רוחי שמיא נשבן בתקוףויתברן לה “I was amazed at this olive tree … [the four]winds of heaven blowing powerfully … and breaking it to pieces” (1Q20 [apGen] 13.15–16).183 There is some question of whether the agreement between Daniel and 1Q20 [apGen] on this grammatical feature may be ascribed to genre, since these are two of the best-preserved Aramaic story collections from the Second Temple period, both of which often cast narrative in the past. Whatever the case, this shared feature is striking in view of our available evidence.

A second mutual feature is the impersonal plural participle אמרין used to indicate something said by a vaguely specified subject “they.” This is always in reference to a pronouncement being made by an authority, either earthly or heavenly. In Daniel we find it three times, one being the command in Dan 3:4 to fall down and worship the golden image: וכרוזא קרא בחיל לכון אמרין עממיא אמיא ולשניא “And the herald called out loudly, ‘To you they are speaking, O peoples, and nations, and language groups …’”184 Here the identity of “they” is not specified, though “they” clearly have the stamp of authority. The same thing is found in 1Q20 (apGen) 6.15, as part of one of Noah’s visions: קל אשמע לך אמרין יא נוח “I heard a voice, ‘To you they are speaking, O Noah!’”185 Note the similar constructions of <ל prefix + pronominal suffix + אמרין + addressee(s)> in both examples, an arrangement also used Dan 4:28. In all cases but Dan 3:4 the construction is used as part of a dream-vision, and to my knowledge is not found in earlier Aramaic, unless we consider the phrase כה אמרן [כ]תבן of Sefire 1C:1 to be a pair of participles (“Thus they are saying [and thus they are wr]iting”) rather than the suffix conjugation, as assumed by Fitzmyer.186

A wider agreement between Daniel and a number of Qumran Aramaic texts is found in their common adoption of the historic participle in narrative situations, something that again might have to do with genre. Gzella discussed the phenomenon of the Aramaic participle extending its verbal functions following the period of Official Aramaic, including its use as a substitution for the suffix conjugation (“Perfekt”) indicating past action.187 The first place this happens with any certainty and regularity is the book of Daniel, as in the sentence באדין נפקין שדרך מישך ועבד נגו מן גוא נורא “Then Shadrach, Meshach, and Abednego departed from the midst of the fire” (Dan 3:26).188 The same function is found regularly in Qumran Aramaic, a good example being the beginning of the angelic dispute over Amram in 4Q544 (VisAmramb) 1.10: והא תרין דאנין עלי וא̇מ̇רין[ “And behold, two (beings) were having a dispute over me and saying[…”189 It is significant that this novelty in narrative use of the participle is employed so similarly in Biblical Aramaic and Qumran Aramaic.190

Another use of the participle that distinguishes Biblical Aramaic and Qumran Aramaic from Official Aramaic is as the verb in a “performative utterance,” the function of which “is to perform an act, rather than describe one.”191 In both Classical Biblical Hebrew and Official Aramaic, the customary way to make such an utterance was to use the suffix conjugation: ברכתכי לפתח “I bless you by Ptah” (e.g., TAD A2.1:2); הגדתי היום ליהוה אלהיך כי… “I proclaim today by the Lord your God that …” (Deut 26:3). The situation changes in Biblical Aramaic and Qumran Aramaic, where we now find a mixture of the suffix conjugation and, more often, the participle: מהודעין אנחנה למלכא די … “We (hereby) inform the king that …” (Ezra 4:16); ו̊כ̊ען לכה עמרם ברי אנא מפק̇[ד… “And now you, Amram my son, I comma[nd …” (4Q542 [TQahat] 1ii.9–10).192 This shows that in Biblical Aramaic and Qumran Aramaic the shift from the suffix conjugation to participle for performative utterances was partially complete, on the way to later dialects, where the participle is used with yet greater diversity.193

One area where some differences may be perceived between Biblical Aramaic (particularly Daniel) on the one hand, and Qumran Aramaic on the other, is word order, or preferred sentence structure. This is a topic deserving of extended treatment, though only a few of the most basic observations will be made here. Despite the many interesting facets of syntax, I will focus primarily on verb placement relative to the main subject and object(s) of a clause or sentence. As a preliminary caveat, it is important to emphasize again the need to recognize generic differences among our texts and the role genre may play in word order. What is more, the very fragmentary nature of most Qumran texts regularly interferes with a clear understanding of their syntax. Folmer has observed that “[f]unctional grammar, as any other syntactic treatment, is most useful when texts to be compared, belong to the same genre, are in good shape, and are clear and unambiguous in meaning.”194 Though our knowledge of Second Temple period Aramaic has grown considerably over recent decades, not least with the publication of the Aramaic scrolls from the vicinity of Qumran, these criteria are in many respects still not met.195 Despite this handicap, there is enough extant literature for some provisional observations to be made on word order.

Studying word order can help us to understand how languages change over time and geographic space, thus providing a useful comparative tool that is less susceptible to scribal interference than orthography, and perhaps also morphology.196 In addition, it can show how a single dialect may use word order to present ideas with different nuances, such as placing a thought in continuity or discontinuity with the main topic of the text.197 For native speakers of a language, deviation from standard word order can be felt immediately, and such deviation may even encode a message of its own. An amusing place to see this at work in the English-speaking realm is the language of the Jedi master, Yoda, in the Star Wars films. Standard English word order is subject-verb-object (SVO), as in the sentence “You will see things through the Force.” Yoda, however, often places verbs at the end of his sentences, and the direct object early (i.e., object-subject-verb [OSV]). Thus, he says instead, “Through the Force, things you will see.” While an English speaker will surely be able to make sense out of this sentence, in the context of the film the abnormal syntax lends Yoda’s speech an aura of foreignness, timelessness, and wisdom. In the context of our Aramaic texts, it is important to note as background information that the eastern languages of Akkadian and Old Persian are widely acknowledged to prefer a subject-object-verb (SOV) sequence, tending to place the verb at or towards the end of clauses or sentences.198 Classical Biblical Hebrew, like most other Semitic languages, generally prefers instead a verb-subject-object (VSO) order, with the verb placed at or near the beginning of a clause or sentence.199 Of course, as in any language, we find plenty of exceptions to these general rules.

Muraoka and Porten observed that Official Aramaic often follows the classical Semitic practice of placing the verb first in a clause or sentence, although, as Folmer has shown in her exhaustive survey, there are a great many exceptions and alternatives.200 In fact, there are so many exceptions that Muraoka and Porten felt it necessary to qualify the situation as follows:201

In our corpus we find a considerable number of cases where the verb, either pc. [prefix conjugation] or sc. [suffix conjugation] form, follows an explicit subject or object or both. This verb-final position, which contradicts the classical Semitic word order VSO, has generally been attributed to a foreign influence, viz. Akkadian on the one hand, which in turn is said to be ultimately influenced by Sumerian, a non-Semitic language, and Persian on the other, also non-Semitic. However, in view of a substantial number of cases in which the verb occupies a non-initial position, it is more accurate to speak of free word-order in our idiom.

The main point for non-specialists to take away from this is that Official Aramaic presents a complex, mixed picture when it comes to clause and sentence structure, even if some general trends may be observed.202 An important ancillary point is that, in clauses and sentences where the verb is placed first and there is an indirect pronominal object mediated by a preposition (something that occurs frequently, especially with the preposition ל), this indirect pronominal object is typically placed in the second position, directly following the verb.203 The examples below illustrate this phenomenon, which is sometimes referred to by linguists as pronoun enclisis, or by the German term Pronominalregel.204

The word order of Biblical Aramaic, and particularly Daniel, has been characterized by a number of scholars as “free,” with a penchant not to place the verb first.205 Coxon, moreover, observed that there is “a marked preference in the biblical documents for the sequence object-verb-subject, and in sentences possessing no direct object the sequence subject-verb is preferred.”206 In fact, the general consensus is that the Aramaic of Daniel is even more “free” than is the norm in Official Aramaic, with the verb coming at or near the end of a clause or sentence more often in the former than in the latter. This variety in word order is nicely illustrated in the group of examples discussed by Yakubovich.207 In the opinions of Kutscher and Coxon, this is because Biblical Aramaic is of a more “eastern” type (i.e., more influenced by Akkadian word order) than is generally true in Official Aramaic, which has more “western” traits.208 Cook, Buth, and Hayes built on these observations by suggesting that some of the word order variation in the Aramaic of Daniel (and in Official Aramaic for Buth) could be ascribed to an increasing complexity of pragmatic functions in the language – i.e., the variation in word order expanded what the language could do.209 For example, in sentences with only a verb and object (the subject being implied by the verb), Cook suggested that those with verb-object (VO) order in the suffix conjugation marked temporal sequence or consecution, thereby indicating one event occurring sequentially after another.210 Buth argued that Official Aramaic essentially remained a verb-subject-object (VSO) dialect, but that this is hidden by the fact that the verb or other elements could be moved in order to mark background-discontinuity (verb placed at the end), or foreground-continuity (often subject-verb order in Daniel).211 This “functional” or “pragmatic” sort of approach has now been carried a step further by Yakubovich in a very interesting study.212 Shepherd proposed a “distributional approach” in which various verb forms and word orders are related to “a remarkable diversity of communication levels.”213 These newer studies suggest that word order in Daniel may not be as “free” (i.e., ungoverned) as it first appears, but they do affirm earlier assessments stressing the great variety of word order in that book.

Study of the word order of Qumran Aramaic to date has centered on two texts, 1Q20 (apGen) and 11Q10 (Job). There is good reason for this, since they present far more running text than in other Qumran Aramaic manuscripts. Outside of 1Q20 (apGen) and 11Q10 (Job), it is often difficult to say anything conclusive about word order, since the context necessary to judge the full meanings of clauses and sentences is broken or missing altogether. Moreover, our representative sample of a given text is very small in most cases. While these factors must be borne in mind, a few trends will be cautiously tendered below.

Muraoka provided a detailed analysis of the word order in 1Q20 (apGen), and his basic results are corroborated by the profile of that text earlier in this volume.214 These results were adumbrated already by Kutscher in his early study of the scroll’s language.215 The most important point to take away from the various sets of data is that 1Q20 (apGen) heavily favors placement of the verb at or toward the beginning of a clause or sentence, which stands in contrast to the much more mixed (or complex) orderings of Daniel. Over two-thirds of the time 1Q20(apGen) has the verb in the first or second position, before the direct object (verb-subject, subject-verb, or verb early with its subject implied from a preceding clause). Only around ten percent of the time does the object precede the verb. In other words, while the Aramaic of 1Q20 (apGen) is not exactly like the verb-early Semitic word order typically found in Biblical Hebrew prose, it is much closer to that syntactic system than is the Aramaic of Daniel. Another syntactic feature setting 1Q20 (apGen) somewhat apart from Official Aramaic and Biblical Aramaic is the strong preference of 1Q20 (apGen) to place the complement (typically the object) after an infinitive, as in למשבק תמרתא “to leave the date-palm” (1Q20 [apGen] 19.15).216 Some have observed that in Official Aramaic and Biblical Aramaic, however, there is more flexibility in order, with the object occasionally preceding the infinitive, e.g., פשרא להודעתני “the interpretation to make known to me” (Dan 4:15), דינן למרשה עליך “suits to bring against you” (TAD B2.11:8).217 While this may be true for a broad comparison, it is important to recognize that some sub-groups or individual scribes in the Official Aramaic corpus clearly preferred placement of the complement after the infinitive, as in 1Q20 (apGen).218 Kutscher and some after him have seen the object + infinitive order as an “eastern” trait, while infinitive + object is more “western.”219 Whatever the case, the variation noted above suggests that this was a grammatical area open to stylistic choice by an author or scribe in Official Aramaic.

What is said above for 1Q20 (apGen) seems to hold true, generally speaking, for a number of other Aramaic works at Qumran. These include, but are not limited to, 4Q537 (TJacob?), 4Q538 (TJudah/Words of Benjamin), 4Q542 (TQahat), 4Q550 (Jews at the Persian Court), the Visions of Amram, Tobit, the Book of Giants, and the other narrative Enochic materials (not including the highly formulaic portions of the Astronomical Book). Though the evidence is by no means complete or unequivocal, all of these texts appear to employ a grammatical approach similar to that of 1Q20 (apGen), with a majority of preserved clauses or sentences having verb-first or verb-early ordering, and objects or other complements following the infinitive. We may call this either more Semitic or more “western,” but however we label it, the differences in word order are notable.220

One area where we see Qumran Aramaic aligning closely with Official Aramaic is that of the aforementioned pronoun enclisis, or Pronominalregel.221 A combined preposition and pronominal suffix (grammatically, an indirect object) is regularly placed directly after the verb in Qumran Aramaic, as seen in the following examples:222

This construction is found over and over again in Qumran Aramaic, most notably in 1Q20 (apGen) because of its extensive stretches of preserved text. Since pronoun enclisis is also a rule followed often in Biblical Hebrew,223 at times we see that the underlying Hebrew could be used to explain the phenomenon in 1Q20 (apGen) (e.g., ויתן לו מעשר מכל (Gen 14:20)//ויהב לה מעשר מן כול (1Q20 [apGen] 22.17). However, the same word order is found in many passages of 1Q20 that do not correspond to portions of Hebrew Genesis, as well as in other Qumran Aramaic works. Combined with the fact that pronoun enclisis is native to Official Aramaic, there seems no need to attribute its use in Qumran Aramaic to Hebrew influence, although the correspondence did allow for a happy mirroring of syntax for those portions of 1Q20 (apGen) that follow closely Hebrew Genesis.

Pronoun enclisis is not as clearly seen in Daniel, where the preposition + pronoun is often (though not always) placed before the verb or active participle, and is sometimes separated from it by several intervening words. Note, for example, ולי הדברי ורברבני יבעון “And my advisors and nobles sought me” (Dan 4:33), which would be an unusual construction for Official Aramaic and Qumran Aramaic (and where we might expect instead ובעו לי הדברי ורברבני). An exception to this seems to be Dan 7, in which pronoun enclisis is more frequent.

11Q10 (Job) presents a special case for several reasons. First of all, it translates a notoriously difficult Hebrew source, which is itself known for some syntactic peculiarities. This includes what Muraoka called “the Sumero-Akkadian type of word order” of the Hebrew (MT) Job, with verbs often placed at or toward the end of clauses.224 11Q10 (Job) generally keeps this word order and, surprisingly, when it does not agree with Hebrew Job the syntax is skewed even further toward the “Sumero-Akkadian” style.225 This clearly contrasts with the great majority of other Qumran Aramaic texts, and was one of the factors leading Muraoka to posit that 11Q10 (Job) was most likely produced in “the East.”226 This idea has not gained wide acceptance, but no matter where the scroll was produced, its word order clearly diverges from the large majority of the Aramaic Qumran scrolls.

Another area of correspondence between Official Aramaic, Biblical Aramaic, and Qumran Aramaic is found in the syntactic and lexical rules governing the linking of verbs of movement to their directional objects. An insightful study of this phenomenon has already been published by Folmer, and much of what I present here simply summarizes her discussion.227 However, some further clarity may be gained on the Qumran Aramaic side of things from the profiles included earlier in this volume, along with the examples gathered in Muraoka’s grammar.228 In Official Aramaic, there is a general rule that most verbs of movement, such as אזל “go” and עלל “enter,” tend to use either no preposition (i.e., direct linking) or the preposition ל for an inanimate directional object, but the preposition על for an animate object, particularly a person. Examples of this are נחת אנת למנפי “you come down to Memphis” (TAD A3.8:11) for the first scenario, and אנה אתית עליך בביתך בסון “I came to you at your house in Syene” (TAD B3.13:2) for the second. It is important to add that, in Official Aramaic, this tendency is not always followed, with ל occasionally being used with an animate directional object.229 It has also been suggested that certain verbs of movement may act differently than others in this regard, although this impression depends partly on the varied practices of some textual sub-groupings.230 Turning to Biblical Aramaic and Qumran Aramaic, we find that the same general rules are observed there, though they are applied more methodically than in Official Aramaic, and with fewer cases of direct linking.231 A few further examples will suffice to illustrate the standard practice:232

Directional verb with an inanimate object, typically a place:

Directional verb with an animate object, typically a person:

Preliminary investigation into other contemporaneous and later dialects, including the second century CE Judean Desert documents and letters, suggests a difference from the fairly uniform approach of Official Aramaic, Biblical Aramaic, and Qumran Aramaic with respect to this practice.233 For example, Folmer observed that the later Judean Desert texts tend to use ל– in all scenarios, while Targums Onkelos and Jonathan use לות for animate directional elements (e.g., ואזל עשו לות ישמעאל “And Esau came to Ishmael” Gen 28:9).234 An exception may be Syriac, which Muraoka proposed complements verbs of movement in a way similar to Qumran Aramaic.235

A number of non-verbal phrases and words also warrant mention, based on either previous discussions or their saliency in the wider scope of surrounding dialects. One such word is the preposition קדם/קודם “before” used in its spatial (rather than temporal) sense, which has been discussed most fully by Jan Joosten.236 Joosten pointed out that, in Official Aramaic and Biblical Aramaic, there is a noticeable tendency to use קדם in a special, deferential sense as a term of respect reserved mostly for those of high rank, such as authority figures and deities. When used in this way, קדם indicates a respectful conceptual distance from the object of the preposition, and replaces the more common spatial prepositions –ל and על. However, in Official Aramaic and Biblical Aramaic the coordination of קדם with kings and deities is not complete; on the one hand, we occasionally find it used with someone who is not a king or deity, and on the other, kings or deities may also be coordinated with prepositions other than קדם. Consider, for example, the following two passages:

By the time of Targums Onkelos and Jonathan, the language of which has often been placed in close proximity to that of 1Q20 (apGen; e.g., by Kutscher), we find that קדם has become much more closely tied to the name of God than in Biblical Aramaic. In Joosten’s words, “dans le Targum, l’emploi de la preposition est devenue systématique dans le language religieux.”237 To illustrate this tighter affiliation, he cited a similar usage of the haphel conjugation of רג״ז “to be afraid” with קדם in Ezra and Targum Jonathan to 1 Kings:

According to Joosten, we see here a development from a preposition that overlapped to a notable (but not complete) extent with the domains of royal and divine address in Official Aramaic and Biblical Aramaic, to a systematic and nearly exclusive application in the religious domain with the targums and other, related corpora. But what of Qumran Aramaic? As in Biblical Aramaic, we often find קדם used in connection with God, but in some Qumran texts from the Enochic group we find the preposition also extended to other divine or quasi-divine beings, such as the angels or giants.238 Though Joosten categorized this differently than occurrences referencing a deity, he is correct in saying that it is “une extension de l’emploi théologique” of Biblical Aramaic and Official Aramaic. Also like Official Aramaic and Biblical Aramaic, the correspondence with God is not complete in Qumran Aramaic. So, for example, we still find in 1Q20 (apGen) 22.21 the following construction:

Alongside these similarities, Joosten suggests that קדם is not used in Qumran Aramaic for human kings, as in Official Aramaic and Biblical Aramaic. This is, in fact, not correct, since we find the locution several times, in 4Q550 (Jews at the Persian Court) 1.4, 4.2, 8.2; 4Q244 (psDanb) 1–3.1 (קודם רברבני מלכא), and 1Q20 (apGen) 19.25 (וקרית קודמיהון referring to the officials of the Pharaoh).239 We might also consider the use of קדם with Joseph under this category, since he too was part of the royal court (4Q538 [TJud] 2.3). In addition, there is the problematic fact that Joosten does not allow Daniel to be considered among the Qumran texts, when it is undeniably part of that corpus. Nevertheless, it is true that we might have expected the idiom more often with the many other instances of the noun מלך in Qumran Aramaic, especially in 1Q20 (apGen). Joosten holds that the author of 1Q20 (apGen) does not know of the distancing formula for language associated with the court, though I suspect it has less to do with knowledge of Official Aramaic than with the intentions and preferences of the author of the Genesis Apocryphon.

When we turn to the second-century CE Judean Desert texts, we find a different picture, with that corpus using קדם a number of times to refer to ordinary individuals and their families in legal contexts.240 This difference could, perhaps, be ascribed to the genre of these texts, with their official, legally-binding status. Whatever the correct explanation may be, the Qumran texts line up quite well with the Official Aramaic and Biblical Aramaic profile when compared to the more systematic application of קדם in Targums Onkelos and Jonathan on the one hand, and the legal use of the later Judean Desert texts on the other. A possible exception to this is 1Q20 (apGen), which does not use קדם when referring to kings.

The specialized particle of negation אל is well-known from Official Aramaic and earlier Aramaic texts, where it is used with a jussive form of the prefix-conjugation (imperfect) verb expressing a command or exhortation. The form is visibly shortened in the plural, by dropping the otherwise expected final nun of that conjugation (e.g., אל תדחלו vs. לא תדחלון).241 This practice continues into Biblical Aramaic, with six preserved instances, at Ezra 9:12 (twice; אל תתנואל תשאו), Dan 2:24 (אל תהובד), 4:16 (אל יבהלך), and 5:10 (twice; אל יבהלוךאל ישתנו). In Qumran Aramaic we find twenty-nine occurrences of אל with the jussive verb in fourteen different manuscripts, as at 4Q213 (Levia) 1i.13 (א̊ל תמחלו) and 4Q542 (TQahat) 1i.5 (אל תתנו). This is thrown into relief when we look to later dialects of Aramaic, such as Targums Onkelos and Jonathan, which no longer use אל, instead regularly replacing the אל of their Hebrew source text with לא. In this respect, Qumran Aramaic is closely aligned with Official Aramaic and Biblical Aramaic against the targums and other Christian and Jewish dialects of later periods, in which אל has disappeared.

The particle ית, used in some Aramaic dialects to mark the direct object in a manner roughly analogous to Hebrew את, has been discussed often in overviews of Biblical Aramaic and Qumran Aramaic. In Early Aramaic and Official Aramaic ית is not present, though by the time of Targums Onkelos and Jonathan, Nabatean, and other Roman-period Jewish and Christian dialects, it is widely-attested.242 Biblical Aramaic and Qumran Aramaic lie somewhere between these two extremes, with both heavily preferring the prefix ל– to mark the direct object when desired. In Biblical Aramaic ית is used once, with a pronominal suffix, at Dan 3:12 (יתהון “them”). Nine sure occurrences are found in Qumran Aramaic, concentrated especially in the New Jerusalem, the Visions of Amram, and the Job translation.243 Most of these are cases of the particle without a suffix, as at 4Q554 (NJa) 2iii.14 (ית טלולה “the roof”) and 4Q559 (papBibChronology) 3.9 (ית אהרון “Aaron”). In several of the Qumran Aramaic examples, it is clear that ית acts as a conscious substitution for Hebrew את, due to either an underlying Hebrew source text (11Q10 [Job] XXXVIII.9) or another form of influence from Biblical Hebrew, as when the author of 4Q559 (papBibChronology) borrows and adapts a standard Biblical Hebrew genealogical formula from Genesis.244

When we observe the translation technique of the later Targums Onkelos and Jonathan, we find that the correspondence between ית and את is much more complete. While the fact that some of the Qumran Aramaic instances of ית can be categorized as Hebraisms does set these few texts apart from the single occurrence of that particle in Biblical Aramaic, Qumran Aramaic generally resembles Biblical Aramaic in its extremely limited use of ית.

The conjunction להן “but rather, except for” is attested over a dozen times in Qumran Aramaic, in at least nine different compositions, and ten times in Biblical Aramaic (in both Ezra and Daniel). It is well known from Official Aramaic, but disappears completely in later Aramaic, including in the targums and other contemporaneous or later Christian and Jewish dialects. It is yet another lexical, or morpho-syntactic, clue aligning Qumran Aramaic more closely with Official Aramaic and Biblical Aramaic than with other, later dialects.

A similar situation obtains for the prepositional phrase ברא מן “except for, outside of,” found in four Qumran texts.245 The phrase first appears in fifth- to fourth-century Elephantine and Northern Saqqara texts,246 and again in one of the Nabatean papyri from the Bar-Kokhba period.247 Apart from these occurrences, the phrase is unknown, the most closely related constructions being לבר מן and מברא in later dialects (translating Hebrew מחוץ in the targums).

Another prepositional phrase with a comparable distribution is עלא מן “above,” found in Biblical Aramaic and Qumran Aramaic.248 It also occurs once in Official Aramaic, where the more common construction is מן עלא.‪249‬ The phrase is not found, however, in later Jewish and Christian dialects, where the more expected forms are עלא, ‪מן עלא‬, ‪–לעיל ל‬, ‪לעיל מן‬, מלעיל, or something similar.250 Like the items discussed above, עלא מן in Daniel and the other Aramaic Qumran scrolls is a feature distinguishing their Aramaic idiom from later Jewish dialects, and placing them closer to Persian and Hellenistic period usage.

As a final example of this sort, we may note the two occurrences of the preposition בטלל “on account of, because of, by the protection of” in 1Q20 (apGen) 19.16, 20.251 Derived from the noun טלל “shade, cover, protection,” בטלל does not appear in Biblical Aramaic, though this is likely due to the fact that there was no occasion for its use. In Official Aramaic, however, it is used often, in precisely the same way as in Qumran Aramaic. In both Official Aramaic and Qumran Aramaic we find the independent form without assimilation of the second lamed, and the suffixed form with such assimilation:

This word is not found in Jewish and Christian Aramaic dialects diachronically later than Qumran Aramaic and, together with the words and phrases collected above, contributes to an important cache of agreements between Official Aramaic, Biblical Aramaic, and Qumran Aramaic against other dialects.253

In concluding this overview of Qumran Aramaic, it is important to acknowledge the significant extent to which Hebraisms and, to a lesser degree, other loanwords characterize the corpus. While Hebraisms were noted already by Avigad and Yadin in their early edition of parts of 1Q20 (apGen), and were expanded further by Beyer, Fitzmyer, Fassberg, and Puech, the most comprehensive, important work has been that of Christian Stadel.254 In addition to cataloguing his own and previous suggestions in Hebraismen in den aramäischen Texten vom Toten Meer, Stadel went on to argue that the Hebrew influence in our Aramaic texts may be assigned primarily to a scribal familiarity with the Hebrew scriptures, and not acquaintance with a contemporary Hebrew vernacular.255 This observation is based on the use of Hebrew words such as רוזן “prince” (4Q530 [EnGiantsb] and 4Q542 [TQahat]) or עליון “Most High” (1Q20 [apGen], 4Q246 [apocrDan], 4Q536 [Birth of Noahc], 4Q541 [apocrLevib?], 4Q543 [VisAmrama], and 4Q552 [Four Kingdomsa]), morphological features like the Hebrew plural endings of עלמים “ages” (1Q20 [apGen]) and שמים (4Q531 [EnGiantsc]), and short expressions, including בני אדם (4Q544 [VisAmramb]). Although Stadel’s theory neither adequately explains every Hebraism, nor provides the only plausible explanation in many cases, it does suffice for a sizable portion of them. Consequently, his work tells us something vital about the scribes responsible for producing much of the Qumran Aramaic literature: they were intimately involved with their Hebrew ancestral texts, to the point of adopting specific words and phrases from them. Yet, they chose to compose their own works in an Aramaic literary idiom clearly extending from, and closely related to, Official Aramaic.

The international scope within which these same scribes worked is reflected linguistically in other, non-Hebrew loanwords. In Daniel these words are primarily Persian and Greek, as discussed at an early stage by S.R. Driver, and frequently thereafter.256 Greek loanwords are largely absent from the other Qumran texts, but Akkadian and Persian words are well-attested and appear with relative frequency across a number of works.257 It is worth noting, however, that a significant number of these occur in 11Q10 (Job), which is of a somewhat different character than most other Aramaic works at Qumran, for reasons already discussed above.

The foregoing overview is intended as a way to gain a basic familiarity with the dialect of Qumran Aramaic, focusing especially on how the Qumran texts comport with Official Aramaic, Biblical Aramaic, and the later Jewish and Christian dialects. As a final point, it is worth addressing briefly the topic of homogeneity and heterogeneity in Qumran Aramaic. Many appraisals of the dialect have treated it as a relatively coherent collection of texts, linguistically speaking. Cook represents well this view.258

Despite claims to the contrary, I feel that the grammatical and lexical variation in the QA [Qumran Aramaic] corpus, while undoubtedly present, is not suggestive of a wide chronological range, and most of the documents themselves were adaptively copied and used during the first century BCE and perhaps part of the first century CE. QA is a generally uniform synchronic corpus, consisting of a small selection of partially preserved didactic-religious texts written in a formal or literary register.

Others, such as Schattner-Rieser and Koller, have stressed instead the linguistic heterogeneity of the corpus, with various texts being allocated to different time periods and social or geographic locations based on isolated linguistic traits.259 The extent to which either approach is correct depends upon the scope of one’s view. When set against the broad sweep of Aramaic dialects, from Early Aramaic to Late Aramaic, Cook is certainly justified in his claim that “QA is a generally uniform synchronic corpus.” However, if we zoom in on Qumran Aramaic exclusively, as Schattner-Rieser does, the finer discrepancies among texts exhibiting some “older” traits (e.g., 4Q242 [PrNab]), those with some “younger” ones (e.g., 1Q20 [apGen]), and linguistic outliers like 11Q10 (Job), stand out more prominently.

Here, however, is where the range of scribal variation attested between manuscripts becomes of paramount importance for any attempt to date the Aramaic Qumran scrolls based on language. Given the range laid out above, is it conceivable that an earlier copy of the Genesis Apocryphon had some “older” linguistic traits that were updated in the process of copying, or that a later, now lost copy of the Prayer of Nabonidus had some perceived archaisms updated by the copying scribe? Since we witness these sorts of changes taking place very clearly among the extant scrolls, the theoretical answer is surely “Yes,” irrespective of what actually happened with either of these texts. Of course, it is impossible to gain clarity on the scribal history of any of our texts, but we can say without doubt that a variety of linguistic changes were made by scribes, who obviously felt comfortable exercising their discretion in this domain. Consequently, it is crucially important to recognize the contingent nature of a position like Schattner-Rieser’s. While we might speak cautiously of dating language based on earlier and later traits in individual copies – bearing in mind the cautionary examples from Elephantine and Elymais discussed above – it is considerably more hazardous to do so for specific compositions.

At the same time, the outer boundaries of the range of scribal variation exhibited in Qumran Aramaic do set some limits on where the dialect may be placed in time and space. Scribes tended to update their new copies in ways that were constrained to certain features within a more-or-less fixed chronological range. For example, some of our scribes evidently would have had a choice between אנת and אנתה “you,” but not את; between דן and דנה “this,” but not דין or דנן. Even texts with the “earliest” traits, such as the Aramaic portions of Ezra, the Prayer of Nabonidus, and to a lesser degree Daniel and Jews at the Persian Court, apparently postdate the cache of Official Aramaic texts found at Elephantine and sundry other locations, even if by only a relatively short period of time. “Later” texts like the Genesis Apocryphon, Tobit, the Visions of Amram, and the Book of Giants exhibit an appreciably greater distance from Targums Onkelos and Jonathan – and even the second century CE Judean Desert texts – than from the texts just listed above. It is difficult to gauge, however, the extent to which differences between Qumran Aramaic and the later Judean Desert texts is due to their obvious disparities in genre and social location, over and above any purely diachronic factors.

In relation to surrounding dialects, then, the Qumran Aramaic scrolls – with a few possible exceptions, and not including the later Judean Desert texts – exhibit a set of linguistic characteristics tight enough to mark them off as a distinctive group, falling chronologically somewhere between Official Aramaic on the one hand, and the second century CE Judean Desert texts and early targums on the other. We have seen that, apart from some superficial orthographic features that demonstrably changed over time with copying, the morphology, semantics, and general linguistic function of Qumran Aramaic regularly falls closer to Official Aramaic and Biblical Aramaic than the later Aramaic literature. The combined facts that a few of the Qumran Aramaic copies are dated by palaeography or carbon dating to the second century BCE, and that these same copies can be placed, linguistically speaking, with the relatively “later” Qumran Aramaic texts, proves that literary works containing such “later” features – even the orthographic ones – need not be dated to later than the second century BCE. A few supporting examples for this claim are the second century BCE copies 4Q201 (Ena), 4Q208 (Enastra), and 4Q542 (TQahat), which show little or no linguistic discrepancy from later copies of works like Tobit, the Book of Giants, and the Aramaic Levi Document. This important observation stands in direct contrast to the conclusion of Kutscher for 1Q20 (apGen), who did not have all of the evidence available at the time of his judgment, and consequently dated the language of that text to the late first century BCE or early first century CE. Only now, with more texts published, can we see that he need not have set the date later than the second century BCE, based on the second-century manuscripts just listed. A related issue, which cannot be addressed here, is whether the last stages of literary productivity in this Aramaic dialect coincided with an apparent resurgence of Hebrew literary activity in the Hasmonean period, evidenced in texts like Jubilees and the sectarian literature from Qumran, the addition of the Hebrew (Hasmonean-period) chapters to Daniel, and the probable translation of Tobit into Hebrew (4Q200 [Tobe]).260

It is more difficult to say where we should fix the earlier end of the diachronic spectrum for Qumran Aramaic. The language of texts with a greater accumulation of “earlier” features still feels somewhat later than the bulk of our fifth- to fourth-century Official Aramaic texts from Elephantine, Bactria, and elsewhere. But what are we to make of this? It is possible that some of the differences are due to factors other than diachronic distance, such as literary genre or a difference in location, whether social or geographic, as Koller and Gzella each suggested in his own way.261 What is more, since our Qumran scrolls tend to date to the second and first centuries BCE (mostly the latter), how much of that difference might be attributed to the process of transmission and copying, in keeping with the scribal changes documented above? An important piece of limiting evidence may be of some help here: Even though the bulk of the Aramaic Qumran scrolls date to roughly a one-century period – from the late second to late first centuries BCE – a range of dialect-internal linguistic difference is still present. That is to say, not all linguistic variation has been flattened or homogenized in our manuscripts, even though they were copied within a relatively narrow window of time. So, most of the “earlier” features of Ezra, Daniel, and the Prayer of Nabonidus persist in the Qumran manuscripts, even if we do witness small adjustments here and there where copies overlap. This suggests that, although scribes obviously felt comfortable making an array of minor changes to the orthography, morphology, and even syntax of the texts with which they worked, there was apparently no full-scale, deliberate program of updating their language per se. Much of the basic linguistic character of a composition remained in tact despite these minor modifications, with the most intensive loci of change being orthographic, such as matres lectionis and the scribal implementation of phonetic shifts (e.g., interdental ז changed to dental ד, or the interchange of ה and א due to aspirantization). This being the case, our best course is to reckon by the scrolls currently available, assuming that they represent the basic style of language in which they were composed, but at the same time recognizing that they may have been updated to some limited extent, and in restricted ways. Being unable to gauge at present the impact of factors like literary genre and social or geographic location, it seems best to allow a generous margin of time for the developments between Official Aramaic and Biblical Aramaic/Qumran Aramaic to have taken place. (The letters of Ezra may be viewed as a special case, and are dealt with further below.) Given the many uncertainties involved, a century offers a reasonable amount of time for most of the changes found in the available scrolls, leaving us with a soft terminus post quem for Biblical Aramaic/Qumran Aramaic around the transition from the fourth to third centuries BCE. In summary, working from the evidence now available, we may posit a working chronological range of roughly two centuries for Biblical Aramaic/Qumran Aramaic, from the late fourth to late second centuries BCE.

6 Biblical Aramaic, Qumran Aramaic, and the Problem of Terminology

In this overview, I have hitherto retained the common distinction between Biblical Aramaic and Qumran Aramaic, though in view of the combined Qumran evidence now available, the validity of that division ought to be seriously questioned. When the distinction was first introduced by Kutscher, only 1Q20 (apGen) was available for study. One can understand, then, why he compared Qumran Aramaic (i.e., the Aramaic of 1Q20 [apGen]) to the much more well-established and well-studied domain of Biblical Aramaic, along with other dialects such as that of Targum Onkelos. As new texts were discovered and published, the separation between Biblical Aramaic and Qumran Aramaic was maintained, even when the Aramaic portions of Ezra and Daniel (as well as Jer 10:11) were added to the “Qumran Aramaic” corpus. Yet with these additional texts, particularly Ezra and Daniel, the separation of Biblical Aramaic and Qumran Aramaic has become confused and untenable, not least because there is a direct terminological overlap between the two: Ezra and Daniel are, strictly speaking, simultaneously Biblical Aramaic and part of “Qumran Aramaic.” How, then, can we speak of comparing Biblical Aramaic and Qumran Aramaic, unless we wish to cordon off the Qumran copies of Ezra and Daniel from the rest of the Qumran corpus based simply on their later theological (not linguistic) claims to canonical status? While the separation of texts along canonical boundaries has been the approach of the editors of the Discoveries in the Judean Desert series and a number of other publications in Dead Sea Scrolls research, in this case it draws an artificial boundary between texts otherwise displaying a natural affinity in both language and content.262

Along with this terminological problem, it has become much more difficult to uphold the sorts of linguistic differences first delineated by Kutscher between Biblical Aramaic (primarily Daniel) and Qumran Aramaic (= 1Q20 [apGen]). We now possess over one hundred additional Qumran Aramaic manuscripts, many of which show micro-variation internal to the literary dialect encompassing Biblical Aramaic and Qumran Aramaic. For example, there are “non-biblical” Qumran Aramaic scrolls that contain early traits once said to set Biblical Aramaic apart from Qumran Aramaic. 4Q570 (Unid. Text D) uses pronominal endings with ם, like Ezra. As in Daniel, 11Q10 (Job) and 4Q529 (Words of Michael) prefer spelling the causative verb with the ה rather than א prefix. The list of such features could be expanded, blurring the border between Biblical Aramaic and Qumran Aramaic once drawn by Kutscher – a border disputed already in 1963 by Rowley for Daniel and 1Q20 (apGen) alone.263 All of this makes it very difficult to maintain the distinction any longer. Instead, we should recognize Biblical Aramaic (the consonantal text only) and Qumran Aramaic as parts of a contiguous, vibrant Jewish literary dialect active during the Persian and Hellenistic periods. The linguistic variation present in our Biblical Aramaic and Qumran Aramaic texts is the combined result of scribal stylistic preference, scribal updating, and limited diachronic development over the period of one or two centuries. As we might expect, some scribes (or copies) exhibit idiosyncrasies, even while clearly participating in the larger dialect. Folmer has helpfully shown that something similar is true of Official Aramaic, so often referred to as a remarkably standardized literary idiom.264

Based on the points above, I would argue that an emic, historically-oriented perspective discourages the separation of Biblical Aramaic and Qumran Aramaic along later canonical lines, as valuable as that distinction may remain for some non-linguistic discussions. This perspective, in turn, encourages a reappraisal of the Biblical Aramaic texts relative to Qumran Aramaic. As noted frequently by those commenting on Biblical Aramaic, Ezra and Daniel should be treated separately from one another, even if in practice this advice often is not followed. One reason for such a distinction is the different compositional histories of the two books. Ezra is an essentially Hebrew book integrating Aramaic documents – the Artaxerxes correspondence – though the precise contours of how this was done and the historicity of the Aramaic documents are still debated. For this reason, Ezra as a composition is really quite different in kind than our other Qumran Aramaic texts. It would be more accurate to place it with the post-exilic Hebrew literature, recognizing that it incorporates some Aramaic material pertinent to our discussion. Daniel’s situation is very different. The most compelling theories of that book’s composition have been put forward by German scholars – crystallized most convincingly by Reinhard Kratz and Rainer Albertz – and have been subsequently adopted by others like John Collins and Carol Newsom.265 These scholars continue to disagree over details of Daniel’s composition history, and the finer points of their descriptions are not always persuasive. However, the essential theory uniting their approaches is that a collection of originally independent (but closely related) Persian- or Hellenistic-period Aramaic tales about Daniel and his companions was brought together at some point before the Hasmonean era, forming an Aramaic compilation comprising all or much of Daniel 1–7.266 Some suggest that Daniel 1 and 7 (and various bits of other chapters, especially Daniel 2) were added at later stages, but more certain is the notion that the Hebrew chapters 8–12 were successively added during the Hasmonean period, as a further elaboration of chapter 7. If this is correct, then the core (or all) of the Aramaic portion of Daniel – and perhaps Daniel 1, if at a late stage it was translated from Aramaic into Hebrew – was originally a collection of Persian- or Hellenistic-period Aramaic compositions, like the bulk of the Qumran Aramaic texts, and unlike the book of Ezra. Thus, in kind this section of Daniel is akin to many other Aramaic works kept at Qumran, though the attention that it – and especially chapter 7 – received resulted in the eventual addition of the Hasmonean-period, Hebrew chapters 8–12, and either the addition of an originally Hebrew Daniel 1 or its translation from Aramaic into Hebrew.267 These fundamental compositional differences between Ezra and Daniel, and the resulting differences in their relationships to the Qumran Aramaic corpus, must be borne in mind while discussing their language: As a literary work, Daniel is much more closely aligned with the rest of the Qumran Aramaic corpus than Ezra, even if the Hebrew-Aramaic edition of Daniel, as a final product, distances it from other Qumran Aramaic texts in important ways.268

Aside from some probable updates to spelling and one or two possible idiomatic adjustments, it is widely accepted that the Aramaic of Ezra’s official letters is essentially Official Aramaic, irrespective of whether or not they preserve actual historical correspondences.269 The basic, if somewhat updated, Official Aramaic character of Ezra’s Aramaic has been most recently reiterated by Folmer and Williamson.270 Some potentially late elements in the greeting formulas of Ezra’s letters have been noticed by Schwiderski, and used to argue that the letters are best seen as products of the Hellenistic period: Use of the preposition –ל (rather than על or אל) in Ezra 5:7 and 7:12 to introduce the recipient, and the salutation שלם “peace, greetings” placed at the end of the greeting phrases in Ezra 4:17 and 5:7.271 The persuasiveness of the first feature has been challenged, and does not inspire great confidence as a tool for dating. The second has stood up quite well to scrutiny, but we may question how confidently a Hellenistic-period date may be advanced based on this single, minor feature alone. In the end, it is best to acknowledge the essentially Official Aramaic character of Ezra’s Aramaic, leaving the requisite room for some scribal change over time, on analogy with what we see in the Qumran Aramaic corpus.

I will not delve into the features used to distance Daniel from Qumran Aramaic, since they have been listed exhaustively by others such as Kutscher and Rowley. Many such features were reviewed near the beginning of this chapter. As noted above, these traits do not hold up very well when Daniel is compared to the entire Qumran Aramaic corpus, and in my opinion the only argument for retaining a distinction between Daniel and Qumran Aramaic is canonical – not linguistic – in nature. The Aramaic of Daniel fits comfortably within the full scope of Qumran Aramaic, a kinship reflected not only in small linguistic details, but also in many shared idiomatic expressions. A number of these were first pointed out by Rowley for Daniel and the Genesis Apocryphon, a list to which Fitzmyer later added.272 These shared expressions could now be expanded yet further, not just with reference to the Genesis Apocryphon, but other texts as well. A few of Rowley’s and Fitzmyer’s examples will suffice to demonstrate the point, further grounding the observation that the authors of Daniel and other Qumran Aramaic texts quite literally “speak the same language”:

Based on these and other idiomatic parallels, Fitzmyer was surely justified in his judgement that, “[f]rom such a list it can be seen that the language of the Genesis Apocryphon is not far removed from that of Daniel.”273 All of this is not even to mention the strong affinities between Daniel and other Qumran Aramaic texts in terms of genre and content, something emphasized by Dimant and others.274

Based on these similarities, it has sometimes been suggested that Daniel and Ezra influenced other Qumran Aramaic texts in some of their phrasing, reifying the idea that the Aramaic of biblical books has a claim to greater antiquity, higher religious authority, or both. For example, addressing a cluster of Aramaic works gathered under the heading “Legendary Narratives and Court-Tales,” Dimant stated that, “[t]he place of the book of Daniel is particularly intriguing, for a number of texts from this group build upon or are influenced by it.”275 Speaking of the Qumran Aramaic corpus, Fassberg remarked that “[t]he inverted word order in the compound tenses hwy + participle (חזה הוית, ‪ידיע להוא‬) also reveals links to and dependence on literary Aramaic, in this case, the biblical books of Daniel and Ezra.”276 Muraoka concurred, judging that Qumran Aramaic occurrences of the inverted periphrastic construction חזה הוית “are best interpreted as modelled on the same combination in Daniel 2.34, 7.4, 9, 11.”277 These scholars may be correct in asserting the influential role of the Biblical Aramaic books, but it is at least worth examining the grounds for their conclusions. Each of the scholars cited above seems to take it as self-evident that non-biblical Qumran Aramaic works would more likely draw on Daniel or Ezra than other possible relationships. In commenting on the spelling of the causative verb with a prefixed ה or א, Muraoka submitted that, already during the period when the Qumran texts were written, Daniel and Ezra had achieved a more authoritative status than other Qumran Aramaic texts.278

That the /h-/ is quite common in BA in the PC [prefix conjugation] and participle as well shows that BA represents a stage earlier than most of the QA documents. The fact that Qumran Daniel (and Ezra) fragments show complete agreement with the MT, not altering H to A, does not have to signify that the scribe’s or scribes’ Aramaic was considerably more archaic than that of many of his or their colleagues, but he or they hesitated to tamper with what was, in his or their eyes, a sacred document.

This statement seems to imply that the other, non-biblical Aramaic works kept at Qumran either were not considered sacred, or that their sacrality was of a somewhat lesser distinction than that of Daniel and Ezra, with the possible result that scribes felt comfortable tampering with the “non-biblical” texts (specifically the causative spelling), but not the “biblical” ones. In reaching this conclusion, Muraoka drew on the earlier study of Pfann, who noted the tendency of the Qumran copies of Daniel and Ezra to follow closely the MT in a handful of (mostly) orthographic features, notably the ה prefix for the causative verb.279 Pfann stopped short of making the claim that this and a few other features were preserved by the copying scribes because these “biblical” texts held a different authoritative status than other Aramaic works at Qumran, but he did single out Daniel and Ezra as unique among the then-published Aramaic scrolls regarding the stability of their textual transmission. As noted above, however, plenty of other Aramaic Qumran scrolls also act in unique ways with regard to language (one thinks most immediately of 4Q212 [Eng] and 11Q10 [Job]), and every one of the traits singled out by Pfann can now be recognized in other, non-biblical texts. In reality, the Qumran Daniel and Ezra copies do not differ appreciably in relation to the MT consonantal text than do other Qumran Aramaic works with overlapping text in two or more copies, such as Tobit and the Visions of Amram. While there may be non-linguistic grounds for accepting the more authoritative status of Daniel, Ezra, or both during the Second Temple period, from a strictly linguistic viewpoint we can say only that there are a few features suggesting that these texts be placed tentatively earlier in the period of Jewish literary activity in Aramaic, with Ezra somewhat earlier than Daniel. Nevertheless, there is clear evidence that the spelling of each book was updated slightly over the course of the Hellenistic and later periods.280 This updating is seen, for example, in the probable change of Daniel’s original passive-reflexive verbs with an א prefix in Qumran copies (אתפעל) to ה prefixes in the MT version.281 This fits well with the broader trends in the Qumran Aramaic corpus, and texts in antiquity more generally.

To sum up, it is evident that Ezra varies as much from Daniel as Daniel does from “later” Qumran Aramaic works like Tobit or the Book of Giants. Consequently, there is good reason to view Biblical Aramaic and Qumran Aramaic as part of the same literary dialect, inherited directly from Official Aramaic and adapted to use as a Jewish language of literary composition over the course of approximately two centuries.282 Texts like Ezra’s letters and the Prayer of Nabonidus contain some linguistic characteristics that place them closer to Official Aramaic, and presumably toward the earlier end of this period. Others, such as Tobit, the Genesis Apocryphon, and the Visions of Amram betray a somewhat later phase, at least in the copies available to us. Daniel and the early Enochic texts fall somewhere between these two poles. However, along with any attempt to sketch dialectical contours in greater detail we must account for the reality that the scribe of any given copy could exercise his own style and preferences during the course of transmission, including archaizing or updating. The dialect thus represents a living scribal tradition – an inconsistently moving target – and as a result we must leave sufficient room for variation from work to work, copy to copy. The fact that Ezra and Daniel were later canonized in Jewish and Christian circles should not result in their being cut off from the literary tradition of which they (especially Daniel) were clearly part.

7 Summary and Conclusions: Early Jewish Literary Aramaic (EJLA)

In this chapter, I offered a descriptive account of linguistic features characterizing Qumran Aramaic, arguing that it should be treated as a Jewish literary dialect thoroughly enmeshed with Biblical Aramaic – especially the Aramaic of Daniel – so as to obfuscate any sharp division between the two. The reasons for treating these text groups as a single unit, rather than two distinct dialects as is so often done, are both linguistic and generic, based on the many clear bonds in both language and content between the Biblical Aramaic and Qumran Aramaic works. This vibrant literary idiom was a direct descendant of Official Aramaic, though adapted for use in a Jewish context.283 The specifically Jewish character of the language is subtle, but may be seen, for example, in the many Hebraisms scattered throughout the corpus (often drawn from the ancestral Hebrew literature), and in the distinctive ל– prefix for the prefix conjugation of the verb הו״א. Based on the linguistic features surveyed in this chapter, it makes good sense to place the beginnings of this Jewish literary use in the late Persian period, with Ezra preserving what appears to be the earliest Aramaic (notwithstanding the arguments of Schwiderski), followed by that of Daniel, the Prayer of Nabonidus (4Q242), and perhaps several other works such as Jews at the Persian Court (4Q550) and some of the Enochic texts. As already mentioned, it can be shown convincingly that we need not date the slightly later-looking Aramaic of texts like the Genesis Apocryphon, Tobit, the Visions of Amram, the Book of Giants, and the Testament of Qahat – all of which are similar enough to discourage attempts at relative dating – any later than the late second century BCE based on the palaeographic and carbon dates of some early manuscripts (e.g., 4Q201 [Ena], 4Q542 [TQahat]). Since it is widely acknowledged that all or most of the Qumran Aramaic scrolls are copies, not autographs of the original composition, from a linguistic viewpoint alone we may plausibly place the composition of Aramaic texts like these at or before the first half of the second century BCE.284

If Biblical Aramaic and Qumran Aramaic merit treatment as an acceptably cohesive dialect, albeit evolving over approximately two centuries, it remains only to discuss the vexed topics of taxonomy and nomenclature. The perennial problem of classification and choosing where dialects begin and end – and consequently what to call the resultant classes and dialects – is witnessed on both large and small scales in Aramaic Studies.285 On the large scale, one needs look no further than the most recent American and German systems, put forward by Joseph Fitzmyer and Klaus Beyer respectively.286 Fitzmyer’s goal was to unify and simplify the somewhat confused organizational landscape existing up until his proposal, in 1966, of a five-phase chronological model. Based, no doubt, on the relative simplicity and cogency of his system, Fitzmyer’s approach has been widely accepted and used, especially in English-language scholarship. However, some have justifiably complained that this simplicity gives the false impression of unity across what are, in fact, extremely diverse types of Aramaic.287 Beyer’s taxonomy, constructed on three major phases (Old Aramaic, Middle Aramaic, and Modern Aramaic), is considerably more sophisticated than Fitzmyer’s, which is at the same time a strength and a weakness; a strength in that its greater level of refinement allows for talking about the language with more precision through time, place, and social location, but a weakness in that many of his finer shades of distinction – proposed with astonishing confidence – are open to serious questioning.288 These two systems also have some unfortunate terminological discrepancies, with the identical terms “Old Aramaic” (Das alte Aramäisch) and “Middle Aramaic” (Das Mittelaramäische) covering very different periods for each scholar. Importantly for our purposes, both systems separate Biblical Aramaic and Qumran Aramaic within their larger taxonomies: For Fitzmyer, Biblical Aramaic is included in his second major diachronic phase, Official Aramaic (roughly 700–200 BCE), while Qumran Aramaic is placed with the following Middle Aramaic period (roughly 200 BCE–200 CE). It should be obvious by now that I would contest 200 BCE as a natural point of chronological division, since it likely bifurcates the active period of the combined Biblical Aramaic-Qumran Aramaic Jewish literary dialect described in this chapter. If Fitzmyer’s system were kept, however, I would argue that Qumran Aramaic is better placed in the earlier, Official Aramaic period. Beyer’s separation is less dramatic, including Biblical Aramaic (Das Biblisch-Aramäische) and “Hasmonaean” (Das Hasmonäische, under which the Qumran literature is placed) as distinct dialects of Post-Achaemenid Imperial Aramaic (Das nachachämenidische Reichsaramäisch).

On a smaller scale, more concerned with Qumran Aramaic, we have just seen that Fitzmyer situated the Qumran texts as a part of his considerably broader category Middle Aramaic, alongside other local text groupings such as the later second century CE Judean Desert documents and letters, Nabatean, Hatran, and Palmyrene. For Beyer, the Qumran literature was part of Hasmonaean Aramaic, also including the later Judean Desert texts, a few Judean inscriptions (which are very short and fragmentary), some bits of the Mishna and Tosefta, “the older layer” (die ältere Schicht) of the Galilean and Babylonian targums, and a few other documents.289 Beyer’s notion that Qumran Aramaic is chronologically Hasmonaean has been adopted recently by Gzella, in an important monograph sketching the historical development of the Aramaic language in antiquity. Echoing the sentiments of Beyer, Gzella writes that,

As elsewhere in the Hellenistic Near East, the change in leadership and the resulting independence of Judaea with the rise of the Hasmonaean dynasty (142 until 37 BCE) had an immediate impact on the linguistic situation. It coincides with the appearance of a local literary variety of Aramaic in Jerusalem and Judaea that is best attested in some hundred-twenty to hundred-thirty religious compositions discovered in the Qumran scrolls.290

Much is assumed in this statement that is either unproven or incorrect, depending on one’s opinion. It is not entirely clear whether Gzella is referencing the Aramaic of the original compositions, the mostly first-century BCE copies, or both.291 If it is the original compositions, there is still considerable debate over where these texts were written. I tend to agree with Gzella’s view that these are, for the most part, texts written in Hellenistic Palestine. Yet others, such as Henryk Drawnel and Jonathan Ben-Dov, have argued for the Mesopotamian extraction of at least some of these texts. Even less agreeable is the placement of the Qumran texts (en masse) in the Hasmonean period. They were copied in that period, to be sure, but the character of the Aramaic by no means requires them to be dated this late, and, as argued above, we would do better to place them instead in the Hellenistic period (or perhaps even the late Persian period, depending upon the text). On non-linguistic grounds, too, many scholars have argued for the Hellenistic dating of texts like Tobit, the Visions of Amram, the Book of Giants, and the Genesis Apocryphon.292 Consequently, it is best not to prejudge the matter by calling Qumran Aramaic “Hasmonaean,” when in all likelihood “Hellenistic” would be the more accurate chronological adjective. In fact, in contrast to Beyer’s and Gzella’s view, it seems more likely that the early Hasmonean period marked the last, waning years of literary activity in Aramaic (though copies continued to be made) and the revival of Hebrew as the preferred literary language of Jews in the re-established land of Israel. For these reasons the identification of the Qumran texts as belonging to Hasmonaean Aramaic ought to be rejected.

A classificatory and terminological alternative with direct relevance to Qumran Aramaic was first proposed by Jonas Greenfield in 1969: Standard Literary Aramaic.293 Greenfield described Standard Literary Aramaic as a supra-local literary language developed from, and heavily dependent upon, Official Aramaic; what Wise has aptly described as “a new Semitic koine that replaced earlier Official Aramaic.”294 An important point distinguishing Greenfield’s conception from those of Fitzmyer and Beyer was the fact that he understood both Biblical Aramaic and Qumran Aramaic to belong to Standard Literary Aramaic, though with Biblical Aramaic being more “eastern” and somewhat earlier, while the Qumran Aramaic texts were western, “written on Palestinian soil.”295 Greenfield drew a sharp line between this fairly standardized literary idiom and whatever types of spoken Aramaic may have existed in Palestine during the same period (hence, the adjective “literary”), and stressed the need to pay attention to differences between the various Qumran Aramaic compositions. Unfortunately, he gave no clear indication of the traits distinguishing Standard Literary Aramaic from other dialects, most pertinently Official Aramaic. This has led some to question the category, as in Gzella’s judgment that it had not “been defined precisely … so the existence of such a standard idiom besides the Achaemenid chancellery language as a medium of administration remains doubtful.”296

Despite the nebulous character of Greenfield’s formulation, Standard Literary Aramaic has been widely referenced by those studying the Aramaic language, and was developed further by Michael Sokoloff and Steven Fassberg.297 Sokoloff placed Qumran Aramaic under what was essentially Greenfield’s Standard Literary Aramaic, but chose to call it instead Jewish Literary Aramaic because of some specifically Jewish features of the language already discussed above.298 In several places Fassberg has now combined these insights, citing Qumran Aramaic as the primary example of what he called either Standard Jewish Literary Aramaic, or Standard Literary Jewish Aramaic.299

While, in my opinion, Fassberg’s Standard Jewish Literary Aramaic comes closest to capturing the essence of the combined Biblical Aramaic (especially Daniel) and Qumran Aramaic dialect that I have endeavored to describe in this chapter, it has the single drawback of not indicating, in general terms, where the dialect fits diachronically into broader developmental models of the Aramaic language. For this reason, I suggest that the language of Biblical Aramaic and Qumran Aramaic be combined under the rubric Early Jewish Literary Aramaic. This label has the advantage of signalling the literary character of the language (following Greenfield and many others), its several Jewish characteristics (with Sokoloff and Fassberg), and its diachronic placement at the beginning of specifically Jewish Aramaic dialects, such as Jewish Palestinian Aramaic and Jewish Babylonian Aramaic. The word “Early” also avoids fixing the dialect to a specific politico-historical period, such as “Hasmonaean” or “Hellenistic” (e.g., Beyer and Gzella) with which it is unlikely to coincide neatly. Of course, under this general heading we remain free to explore the finer contours and development of the dialect attested between the various compositions and copies, including between the now-biblical works (traditionally, Biblical Aramaic) and those that did not achieve canonical status (traditionally, Qumran Aramaic). We may compare texts with more “eastern” and more “western” features (e.g., the syntactic differences between Daniel and the Genesis Apocryphon), or with “earlier” and “later” features (e.g., the Prayer of Nabonidus and Tobit), all the while recognizing that these differences occur among what are much more substantial and numerous linguistic similarities, warranting their treatment together. Early Jewish Literary Aramaic is closely related to Official Aramaic on the one hand, and to the Onkelos and Jonathan targums on the other. It is also relatively similar to some of the other writings placed under Standard Literary Aramaic by Greenfield, or under Hasmonaean by Beyer and Gzella. In my opinion, the designation Early Jewish Literary Aramaic allows us to picture more clearly, from a linguistic viewpoint, a tremendously creative, fertile branch of Jewish literary activity in Aramaic during the Second Temple period. Written by highly-trained scribes working in the broader tradition of Official Aramaic, this language was the vehicle of an impressive international, ethno-religious literature that drew heavily upon earlier Hebrew traditions, and most likely spanned from the late Persian period to the waning years of Hellenistic rule in Palestine and the ascension of the Hasmoneans. From the Hasmonean period onward, this Jewish Aramaic literary tradition appears slowly to have given way to a new wave of literary productivity in Hebrew.

1

I have attempted to summarize the issues at play here in an earlier article, Machiela, “Translation.”

2

This question leads to the vexed topic of terminology associated with the Aramaic of the Qumran texts, which will be discussed below.

3

Kutscher, “Genesis Apocryphon.”

4

Kutscher’s Middle Aramaic is to be distinguished from Fitzmyer’s later definition of the same label, which encompasses an earlier period than that intended by Kutscher. Kutscher never clearly defined Middle Aramaic, but one may gather from his related comments that it includes the various Jewish targums (especially Onkelos), Galilean Aramaic, Christian Palestinian Aramaic, and Nabatean. For a slightly fuller explanation, see Kutscher, “Dating,” 288–89. Middle Aramaic has been used in this same sense by Kutscher’s student, Michael Sokoloff. The definition of Fitzmyer has been followed by many scholars, such as Schattner-Rieser, Grammaire, 20. A completely different dating taxonomy is proposed by Klaus Beyer, ATTM1, 23–76 (with an updated, English translation published as Language). Beyer use the term Mittelaramäisch in a way distinct from both Kutscher and Fitzmyer, adding to the confusion around this term. Beyer placed most of the Qumran Aramaic texts under his category Hasmonäisch (i.e., Hasmonean Aramaic; ATTM1, 34–35), with the notable exception of 4Q201 (Ena), which is the only text he lists as Jüdisch-Altostjordanische (227).

5

Fitzmyer, Commentary, 28.

6

Rowley, “Notes.”

7

Rowley, “Notes,” 129.

8

Rowley, “Notes,” 129.

9

van der Ploeg and van der Woude, Job, 3–4. Sokoloff, Job, 9–16.

10

Sokoloff, Job, 25.

11

Sokoloff, “Notes.”

12

Sokoloff, “Notes,” 203. On the potentially complicating factor of this scroll being written in a different geographic location than other Qumran scrolls, see section 4.2 in the following chapter on scribal practices.

13

García Martínez, “New Jerusalem,” 456. See also the comments of Puech in DJD 37:98, and the earlier assessment in DJD 3:184.

14

Wise, Thunder, 103–151 (published earlier in Muraoka, Studies, 124–67); Cook, “Kohath” (quotation at 218). “Typological series” is a phrase first used by Cook (“Kohath,” 216) to describe the approach developed by Kutscher and Sokoloff.

15

Though not dealt with directly in what follows, see also the corroborating comments of Wacholder, “Judaeo-Aramaic,” 259.

16

Gzella (see below) is correct to point out that the rather acerbic, haughty tone of Wise’s essay may contribute to its being widely ignored in scholarly treatments. Nevertheless, some of his points deserve serious consideration.

17

Cook, “Kohath,” 216.

18

Sokoloff, Job, 25.

19

Cook, “Kohath,” 217–18. At the end of his article, Cook explicitly rejects Wise’s diglossia solution, preferring instead the more geographically-oriented “local orthography” explanation. On this point, see also Cook, “Dialectology.”

20

Cook, “Kohath,” 218. Cook’s two main examples are the Qumran Enoch manuscripts and 4Q542 (TQahat).

21

Cook, “Kohath,” 219.

22

Gzella, “Dating”; idem, Cultural History, 231.

23

Gzella, “Dating,” 72.

24

Gzella, “Dating,” 75–6.

25

Gzella, “Dating,” 78.

26

Koller, “Dialects.”

27

Koller, “Dialects,” 212–13. See also Schattner-Rieser, Grammaire, 25; Cook, “Dialectology,” 7–8.

28

In concluding this section, attention should also be drawn to the comments of Christian Stadel, who issues a similar warning for typological dating based on how various text groups employ the word כל. See Stadel, “Syntagm,” 44.

29

As noted by Cook, “Kohath,” 217. See Tisdall, “Daniel,” 237–45; Schäder, Beiträge, 242, 245–46.

30

Kutscher wrote of the “language” and “spelling” of the “scroll” (“Genesis Apocryphon,” 15, 27–28), which could be taken to mean than he was referring to this copy only in his analysis. See also Avigad and Yadin, Genesis Apocryphon, 38–9.

31

Avigad and Yadin, Genesis Apocryphon, 38–9; Fitzmyer, Commentary, 26–28.

32

The third example was noted already by Cook in “Kohath,” 217, n. 43.

33

For background on the differences between the ketiv and qere in Daniel, see Morrow and Clarke, “Ketiv/Qere.” While Cook (“Kohath,” 217, n. 43) understands this variant in 4Q112 (Dana) to be an archaism, it is also technically possible that it is a Hebraism. This possibility gains some weight from the presence of a stark Hebraism in the plural ending of חרטמים at 4Q112 (Dana) 3i, 17.11, versus חרטמין of MT Dan 2:27. In either case, the general point of scribal intervention is illustrated.

34

Kutscher and Sokoloff imply that the ithpeel, like the aphel, is a later spelling. However, Cook (“Dialectology,” 14–16) has pointed out correctly that the situation is, in fact, the opposite for the passive-reflexive stem.

35

The topic is discussed briefly by Gzella (Cultural History, 46, n. 111), though even there the term is left quite open.

36

Another potentially fruitful area of study for our purposes is the type of scribal corrections made in the available manuscripts. This will be discussed in the next chapter, on scribal practices.

37

Parallels are only provided where there is clearly overlapping text. For example, where two copies preserve part of the same verse, but no certain letters or words are shared in common, that correspondence is not included.

38

One might legitimately question the inclusion of MT Daniel and Ezra among the Qumran manuscripts, and ask whether this may skew the analysis, since Leningradensis was not found in the caves, and the origin of its text cannot be placed in time and space with certainty. However, I include it here on the grounds that its text is widely recognized to be very ancient, indeed to antedate in some cases the Qumran copies. Moreover, the types of changes exhibited between the Qumran Daniel-Ezra copies and the MT consonantal text are very similar to those exhibited between parallel Qumran copies for other works. Finally, the MT consonantal text generally corresponds closely to the Qumran copies, although there are many small differences among them. In fact, similar changes are regularly seen taking place in both directions, with (for example) the MT including a mater lectionis against a Qumran manuscript in one place, and vice versa in another. The case is quite different for the Cairo Geniza copy of the Aramaic Levi Document, and for that reason I did not include it in the list of parallels. For such a list, see the editions of Greenfield, Stone, and Eshel, Aramaic Levi, and Drawnel, Aramaic Wisdom.

39

Scrolls are listed in order of manuscript number and according to literary composition. In general, parallels are listed only once, at the occurrence of the first scroll in the list. Occasionally, when there is an overlap between three witnesses, a parallel is listed twice because distinct overlaps of a passage may be preserved between two of the three witnesses.

40

The relevant texts may be found in Drawnel, Aramaic Wisdom. See now also an additional fragment of the Genizah text published by Bohak, “Aramaic Levi,” and discussed further by Drawnel, “Fragment.”

41

Schattner-Rieser, “archaïsmes,” 102–6, 111.

42

Schattner-Rieser, “archaïsmes.” See also the comments of Gzella, Cultural History, 231.

43

Folmer, Aramaic Language, 715–17.

44

For a listing of the documents written by Natan and Mauwziah, along with their dates and other details, see Porten and Yardeni, TAD 2, 188.

45

See especially the pronominal prefix דיל– “belonging to –” and noun דכר “male.” This phenomenon occurs six times in Natan’s writing, and once in Mauwziah’s. For Natan, see TAD B2.7:7, 11, 16; B2.6:17, 20, and B3.1:23. For Mauwziah, see TAD B8.8:2. It should be noted, however, that the use of ז still far outweighs that of ד in Natan’s writing.

46

For Natan see TAD B2.6:4, B3.1:5 and B3.2:8, and for Mauwziah a number of times in TAD B3.8. Natan also uses שק״ל frequently.

47

For כה״ל and יכ״ל, compare TAD B2.6:31, 35 (Natan) to TAD B2.10:9, 10 (Mauwziah). An example of the conditional sentence construction is Mauwziah’s הן יאמר כות לא ישתמע לה (TAD B3.8:42) versus Natan’s הן מטת מריבתא לרשא ירבה מריבתא כרשא חד כחד (TAD B3.1:6). For even more features, consult Folmer, Aramaic Language, 715–17.

48

TAD B3.4:12; B3.6:3; B3.11:3; B3.10:10, 14; B3.12:30, 31.

49

See Gzella, “Arsacid,” 112–22. The Tang-e Sarvak inscriptions were originally published by Henning, “Tang-i Sarvak.” For the editio princeps of the Tang-e Butān inscriptions, see Bivar and Shaked, “Inscriptions.”

50

Koller, “Dialects.”

51

Crystal, Dictionary, 142.

52

On the topic of linguistic change in Aramaic, see the recent history of the language by Gzella, Cultural History, especially on pages 45–52.

53

Gzella, Cultural History, 157–201. Here Gzella gives a helpful descriptive overview of Official Aramaic and the texts included under that category.

54

See Gzella, Cultural History, 205–8.

55

The same basic trend is seen in Qumran Hebrew, as overviewed by Elisha Qimron (Grammar, 55–98) and Eric Reymond (Qumran Hebrew, 13–63).

56

This is not, however, to imply that the vowels represented by these letters were not spoken at an earlier time. As will be seen below, this remains a matter of scholarly debate.

57

On this point see Reymond, Qumran Hebrew, 35–37.

58

For further details and examples see Muraoka, Grammar, 21, 24–26, 28–29.

59

Cook, “Aramaic,” 362.

60

See DJD 37 and Yadin, Bar-Kokhba.

61

Hundreds of additional Aramaic ostraca from Maresha have been discovered in recent years. These are planned for publication in the coming decade as part of a project led by Prof. Esther Eshel at Bar-Ilan University.

62

Reymond, Qumran Hebrew, 43–47.

63

Cook, “Aramaic,” 362–63.

64

Qimron, Grammar, 82–86; Reymond, Qumran Hebrew, 56–57. Note the frequent use of שגיא in Official Aramaic. In later Aramaic the typical spelling is סגי. היא and the like is used in later Jewish Aramaic (probably influenced by Biblical Aramaic and/or Hebrew), while in Official Aramaic we find הי.

65

E.g., 5Q15 (NJ) 1i.13, 4Q206 (Ene) 4i.17, 11Q18 (NJ) 13.5, MT Dan 3:25, 4:7, and 7:15.

66

See Muraoka, Grammar, 23, 27–28.

67

Cook, “Aramaic,” 362.

68

Qimron, Grammar, 58–61, Reymond, Qumran Hebrew, 47–51.

69

See Muraoka, Grammar, 22–23, 26–27.

70

For examples see Muraoka, Grammar, 26.

71

Reymond, Qumran Hebrew, 39–43.

72

So Muraoka, Grammar, 11, though see the comments of Cook, “Aramaic,” 363. On the phenomenon in its grammatical context see Muraoka, Grammar, 6–11. Another nice orientation to the issue, though done before most of the Aramaic Qumran scrolls were published, is that of Coxon, “nasalization.”

73

As noted by Schattner-Rieser (Grammaire, 44–45) and Muraoka (Grammar, 10). For a word with an etymological נ, such as the third masculine singular prefix conjugation of נת״ן, the process would be: Lengthening of the word through addition of the prefix ינתן > assimilation of the נ and compensatory gemination (doubling or lengthening) of the second radical יתתן > dissimilation of gemination by adding נ, as is usual for geminated consonants, resulting in ינתן. In this scenario, the process for words with and without etymological נ is the same, once the etymological נ has been assimilated and the following consonant doubled.

74

Certain words reveal some differences, typically with Qumran Aramaic not having the נ and thereby suggesting possible chronological development. One such case is מדינתא “the province, land,” which always keeps the נ in Official Aramaic and Biblical Aramaic, but in two out of seventeen occurrences in Qumran Aramaic exhibits assimilation. Also see the verbal roots על״ל and סל״ק. However, the situation changes with, e.g., the word א(נ)ף “face, nose,” where Qumran Aramaic and Biblical Aramaic always have dissimilation, but Official Aramaic sometimes does not. Since we are dealing with small sample sizes, it is difficult to know how much weight to place on differences like these, and in the end it seems accurate to say that Official Aramaic, Biblical Aramaic, and Qumran Aramaic prefer the preservation or addition of נ in many of the same situations, with Qumran Aramaic exhibiting a slight decline from Official Aramaic and Biblical Aramaic.

75

Kutscher, “Genesis Apocryphon,” 4.

76

Beyer, ATTM1, 53, 116–20; Cook, “Aramaic,” 364; Muraoka, Grammar, 30–31; Gzella, Cultural History, 207.

77

In the very few other cases where Qumran Aramaic may show this process, it could also be attributed to scribal error, as noted by Muraoka.

78

For Biblical Aramaic I refer, of course, only to the consonantal text, and not the later Masoretic vocalization.

79

For a fuller account see Muraoka, Grammar, 3–6. One shift not dealt with below is that from ש /ṯ/ to ת /t/, discussed by Schattner-Rieser (Grammaire, 36) and Muraoka (Grammar, 6). Both refer to אשור (typically אתור in Aramaic at this time) in 1Q20 (apGen) 17.8 (it also occurs in 12.18), but both times in 1Q20 (apGen) אשור refers to the personal name of Shem’s son (cf. Gen 10:22), all of which are spelled according to the Hebrew conventions of Genesis. As such, this name should be dismissed as evidence for any such shift. More useful is the clear shift from the verbal root שו״ב “to return” in Official Aramaic to תו״ב in Biblical Aramaic and Qumran Aramaic.

80

See Schattner-Rieser, “archaïsmes,” 104–5; Muraoka, Grammar, 4.

81

There can be no doubt that ד was already use occasionally in the fifth century BCE, as seen clearly, for example, in text 44.3 of the Persepolis ritual objects published by Bowmen, Persepolis, 115. This text, written on a green chert ritual plate, uses the demonstrative pronoun דנה, while the many other similar texts have זנה. The plate with דנה is securely dated to 452/51 BCE.

82

Eshel, Puech, and Kloner, “Maresha.” See also the evidence adduced by Schattner-Rieser, “archaïsmes,” 105, and the discussion of Koller, “Dialects,” 206.

83

Personal communication on May 13, 2015.

84

This is yet another reminder to avoid simple linear models that posit, e.g., texts using ז as being chronologically earlier, and those using ד as being later.

85

See especially Sokoloff, “Dialects,” 748–50; Muraoka, Grammar, 6.

86

This includes the word עעיתה “wooden frame/lattice?” (2x), though there has been some discussion as to whether it is synonymous with עע.

87

This is another process of dissimilation, on which see Muraoka, Grammar, 12. In 4Q211 (Enastrd) 1i.3, the weakening may have been represented in the other character, based on the spelling עא.

88

As noted by Sokoloff, the same development is seen in Official Aramaic ער״ק > Qumran Aramaic ער״ע > and אר״ע in later dialects.

89

See the discussion in Koller, “Dialects,” 204–207. Ephʿal and Naveh, Ostraca, texts 25:2 and 167:2. Eshel, Puech, and Kloner, “Maresha.”

90

Sokoloff, “Dialects,” 748–49.

91

The change is also seen nicely in a comparison of the Qumran copies of the Aramaic Levi Document, which typically have ש, and the Cairo Geniza copy, which for the same words has ס.

92

So Muraoka, Grammar, 4.

93

See some of the references in Schattner-Rieser, Grammaire, 36.

94

For example, see the comments of Levias, Grammar, 14. In some cases (e.g., Syriac) the sin (ש) disappears altogether.

95

Schattner-Rieser (Grammaire, 43) sees this as part of a larger shift toward nasalization or nunation. See also Schattner-Rieser, “archaïsmes,” 107–9.

96

Notable exceptions at Qumran are 4Q112 (Dana) and 4Q570 (Unid. Text D).

97

4Q242 (PrNab) and 11Q10 (Job) are the exceptions; see further below.

98

For the Judean Desert documents see the examples given by Muraoka, Grammar, 41.

99

See Schattner-Rieser, “archaïsmes,” 106.

100

With the exception of the Hermopolis Papyri, which also have אפעל.

101

Cook, “Aramaic,” 373. An excellent overview of the topic is available in Muraoka, Grammar, 109–111.

102

See Muraoka, Grammar, 110, and the other literature cited there.

103

As noted by Koller, “Dialects,” 207, n. 25.

104

As pointed out by Cook, “Dialectology,” 14–16. For many examples see Muraoka, Grammar, 113–14.

105

Beyer (ATTM1, 463, 466) and Cook (“Aramaic,” 374) attribute the forms in Biblical Aramaic and 11Q10 (Job) to Hebrew influence from the התפעל stem, though Muraoka (Grammar, 112) disputes this claim. Bauer-Leander (Grammatik, 108) thought, rather, that it was formed on analogy with the הפעל.

106

See Muraoka, Grammar, 23–24. A nice example of the verb with weak final radical is תתקרה (4Q543 [VisAmrama] 2a–b.4) versus תתקרא (4Q563 [Wisdom Composition] 1.2). The letter א is more typically found in the final position, though it can alternate with ה even within the same work (e.g., מתחזא [4Q204 (Enc) 1i.21] and מתחזה [4Q571 (Words of Michaela) 1.13]).

107

As in a number of the Hermopolis Papyri (TAD A2.1, A2.2, A2.3, A2.4 and A2.5), which use the negative particle לה, rather than the standard לא.

108

Muraoka, Grammar, 37–50. Beyer, ATTM1, 423–25, 449–52. On many of these forms see also Cook, “Vowels.”

109

See Cook, “Vowels,” 63–64; Muraoka and Porten, Grammar, 43–44; Muraoka, Grammar, 41–43.

110

This spelling also occurs in Qumran Hebrew. See Fassberg, “Hebraisms,” 51; Schattner-Rieser, Grammaire, 53; Cook, “Aramaic,” 365. In my opinion, the spelling suggests strongly that this copy was made at Qumran, exhibiting some features of what Tov has called the Qumran Scribal Practice.

111

So Muraoka, Grammar, 37.

112

See Cook, “Dialectology,” 11.

113

Cook, “Aramaic,” 365, cites Rosenthal (Grammar, 23 [§29]) in support of this claim, though it should be noted that Rosenthal wrote that “With one exception … הִמּוֹ and הִמּוֹן happen to be used as direct objects immediately following a verb in the perfect tense” (italics added).

114

Hence, in Ezra 5 we find precisely the opposite linguistic situation as that suggested by Cook.

115

The defective form appears once in Qumran Aramaic at 4Q213a (Levib) 3–4.2.

116

For the former opinion see Beyer, ATTM1, 424, 449–50; Cook, “Vowels,” 60–61; Qimron, “Suffix”; and with some further nuance, Sokoloff, “Dialects,” 751. On the Hebraism theory see Fassberg, “Hebraisms,” 51–53.

117

The form is not present in Biblical Aramaic. The fact that both the short and long forms occur in Official Aramaic may testify to the fact that from early on the suffix was pronounced /–ki/, though the phonetic consistency behind these ending is disputed. On this point see Fassberg, “Suffix”; Sokoloff, “Dialects,” 752.

118

One can see the stark difference between Qumran Aramaic and the Judean Desert documents in the examples provided by Muraoka, Grammar, 40.

119

Kutscher, Isaiah Scroll, 158–61. For an alternative explanation see Qimron, Grammar, 266–67.

120

See Cook, “Vowels,” 56–58.

121

For further discussion and reference to earlier opinions see Fassberg, “Hebraisms,” 53–54; Muraoka, Grammar, 40.

122

Fassberg (“Hebraisms,” 54) and Sokoloff (“Dialects,” 751–52) pointed out that the long ending was primarily a feature of 1Q20 (apGen), which was true at the time. However, the ending has now come to light in many other Aramaic Qumran scrolls (e.g., 4Q197 [Tobb] 4ii.2, 4Q213a [Levib] 3–4.6, 4Q537 [TJacob?] 12.3, 4Q541 [apocrLevib?] 9i.4, and 4Q549 [VisAmramg?] 2.2), so that it can said to have a much wider distribution.

123

The feminine is either poorly attested or, according to Muraoka and Porten (Grammar, 46, n. 215), not attested at all.

124

Muraoka, Grammar, 41. As with some of the other forms already discussed, it is at least conceivable that these are Hebraisms, though an archaism from earlier Aramaic seems more likely.

125

For references see Schwiderski, DARI1, 291. זנה is also found in a second century BCE inscription from Mt. Gerizim (Naveh, “Samaria”).

126

On some of these forms see the study of Folmer, “Pronouns.”

127

So Puech (DJD 31:327), Schattner-Rieser (Grammaire, 63), and Muraoka (Grammar, 47), though it may also be the more expected דן prefaced by an interrogative ה, as suggested by Kobelski (Melchizedek, 32). On other dialectical forms see Cook, “Dialectology,” 8, 10–11; “Aramaic,” 367. The later grammatical functions noted by Cook may support Kobelski’s view.

128

See Schattner-Rieser, Grammaire, 62–63.

129

Muraoka’s interpretation of בדכן in 4Q209 (Enastrb) 23.5 (Grammar, 49) is clearly mistaken based on the parallel in 4Q210 (Enastrc) 1iia+b+c.16.

130

Or very occasionally other forms, such as דיכי in Targum Onkelos/Targum Jonathan. Muraoka and Porten (Grammar, 58) note that הו is occasionally used in a demonstrative way in Official Aramaic.

131

Muraoka and Porten, Grammar, 56–57. The use of הוא and היא as far demonstrative pronouns is completely absent from Qumran Aramaic, where they are used primarily as the second component of a predicate clause, e.g., אבי הוא “He is my father” (4Q197 [Tobb] 4iii.8) or a subject. Dan 2:32 and 6:11 do have הוא as a far demonstrative, but Biblical Aramaic usage otherwise resembles Qumran Aramaic.

132

As Schattner-Rieser (Grammaire, 63) points out, the word אלה in 4Q536 (Birth of Noahc) may also be the proper noun “God,” with the context supporting either interpretation. The fact that this scroll otherwise falls into line with Qumran Aramaic may support Schattner-Rieser’s opinion, though see also Puech, DJD 31:169. אלה also appears in the single verse of Aramaic in Jeremiah, in 4Q70 (Jera) 5i.4. See also the profile for 4Q536 (Birth of Noahc) and the relevant discussion there.

133

See Cook, “Dialectology,” 8, 10.

134

Muraoka, Grammar, 49.

135

I will not deal here with the suffix-conjugation third-person masculine plural verb ending –ון (rather than Official Aramaic –ו), discussed by Fassberg, “Verbal System,” 75. What he considers his most secure example, in 1Q20 (apGen) 5.16, is actually in the prefix conjugation, severely weakening the presence of this feature in Qumran Aramaic. See also Muraoka, Grammar, 99.

136

See Fassberg, “Verbal System,” 70–71.

137

See Muraoka, Grammar, 98.

138

See, e.g., Muraoka and Porten, Grammar, 97, 100–101.

139

The qere tradition of Biblical Aramaic, however, vocalizes these verbs as if they had the /-ā/ ending seen graphically in the Qumran Aramaic form.

140

Opportunities for use of the form are rare in the preserved texts. It is found at 1Q20 (apGen) 5.12, 13.16, 20.6, 22.28; 4Q201 (Ena) 1iii.16; and 4Q202 (Enb) 1ii.2.

141

See Fassberg, “Verbal System,” 75–76; Schattner-Rieser, Grammaire, 70.

142

Eshel, Puech, and Kloner, “Maresha,” 41.

143

There are a few exceptions to this rule, collected by Schattner-Rieser, Grammaire, 71, though not all of the readings are certain. For discussion see Sokoloff, “Dialects,” 750–51; Fassberg, “Verbal System,” 68–69; and Fassberg, “להוא.” In this study Fassberg includes exceptions from the Cairo Geniza copy of the Aramaic Levi Document in his numbers, though placing this copy with the Aramaic Qumran scrolls for these purposes is highly questionable from a methodological viewpoint (the same thing is done in Muraoka’s Grammar).

144

For a review of the different suggestions, see Rubin, “Preformative.” Also, Kaufman, “Reflections,” 150.

145

See the discussion appended to Fassberg “Verbal System,” 81.

146

Fassberg, “להוא.”

147

In Fassberg’s opinion (“להוא,” 16), this may simply capture the earliest stages of ל as a prefix in Jewish Aramaic, with its attachment to הו״א being due to the fact that its “consonants are weak and thus prone to morphological innovation.”

148

Fassberg, “Verbal System,” 69. On these later dialects see especially Rubin, “Preformative.”

149

Koller, “Dialects,” 203–4. For the situation of אפעל versus הפעל forms, and אתפעל versus התפעל forms, see the discussion above on the orthographic, and likely phonetic, shift between ה and א in Qumran Aramaic.

150

The quotation, also cited by Koller (“Dialects,” 203–4), is found in Yadin, Bar Kokhba, 23. On the Nabatean texts, see especially Morgenstern, “Nabataean,” 139.

151

As noted by Koller and Morgenstern, the non-metathesized forms also occur occasionally in Qumran Hebrew. For examples see Qimron, Grammar, 239.

152

Muraoka, Grammar, 113.

153

See Muraoka, Grammar, 103–106. Rarely this stem will not have the מ; for the common exception of לאמר in Official Aramaic see Muraoka and Porten, Grammar, 108.

154

Muraoka and Porten, Grammar, 108–109; Muraoka, Grammar, 104–105.

155

Cf. Muraoka (Grammar, 105) for references. See also Ezra 5:9 למבניה, though the proper interpretation of the form there is disputed, as the ending may indicate the determined state or a pronominal suffix.

156

Muraoka, Grammar, 104–105; Schattner-Rieser, Grammaire, 71–72.

157

Fassberg, “Verbal System,” 73–74. Though see already Kutscher, “Genesis Apocryphon,” 13.

158

One of the examples cited by Fassberg is not correct: His במעול (1Q20 [apGen] 6.4) should instead be read במעיל. As a result, his “miqtol” forms are not represented. The other occurrences given by Fassberg are at 4Q530 (EnGiantsb) 2ii.15, 4Q534 (Birth of Noaha) 7.4, 4Q541 (apocrLevib?) 24ii.5, 4Q542 (TQahat) 1ii.3, and 11Q10 (Job) XXXIII.9. Most of these match the Masoretic vocalization of Biblical Aramaic.

159

Muraoka, Grammar, 107–108. Also Cook, “Aramaic,” 372.

160

As noted by Kutscher, “Genesis Apocryphon,” 13.

161

As pointed out by Schattner-Rieser (Grammaire, 77) and others, the Biblical Aramaic qere appears to follow the vocalization expected of a form with י representing the internal vowel, and therefore seems at odds with the ketiv.

162

Muraoka and Porten (Grammar, 131, n. 607) do, however, wonder if this is simply due to a lack of adequate attestation for the relevant forms.

163

Beyer, ATTM1, 53, 116–20; Cook, “Aramaic,” 364; Muraoka, Grammar, 30–31; Gzella, Cultural History, 207.

164

Schattner-Rieser, Grammaire, 77.

165

For the occurrences, see the relevant entries in Schwiderski, DARI1.

166

The one proposed occurrence of כה״ל at 4Q530 (EnGiantsb) 5.2 is very uncertain, and no argument can be built on it.

167

Fassberg, “Verbal System,” 68–69.

168

4Q542 (TQahat) 1ii.13 also uses הל״ך for the nominalized ithpaal form באתהילכותהון “by their conduct.”

169

See Joosten’s response at the end of Fassberg’s article “Verbal System,” 78–80. I am inclined to agree with Joosten on this point.

170

Typically spelled in both corpora defectively: e.g., קשט, as at Dan 2:47, 4:34. For the pattern in Official Aramaic see Muraoka and Porten, Grammar, 80–81.

171

For the most recent and extensive treatment of the topic see Muraoka, Grammar, 66–70.

172

With reference to 1Q20 (apGen), Kutscher (“Genesis Apocryphon,” 12–13) stated that, “[t]he qutl pattern of the noun appears in the indeterminate state as qotl (qutl)…, exactly as in Hebrew. In ‘correct’ Aramaic it should have been qtol, e.g. קשוט, ‪תקוף‬.” Of the word קושוטא in 4Q542 (TQahat) 1ii.1, Puech (“Qahat”) wrote, “La forme QWŠWT (avant correction) et ii 1 montre non l’hésitation entre la forme qutl (type hébreu) et qtol (plus araméenne) mais un compromise.”

173

Kutscher, “Cave I,” 181; Kutscher, Language, 396–98. This possibility was initially eschewed by Muraoka (“Segolate,” 231–32), but he later accepted it as likely (Grammar, 69).

174

As argued by Cook, “Kohath,” 207–9.

175

Muraoka, Grammar, 69, n. 290.

176

On this phenomenon see Fassberg, “Verbal System,” 71–72; Muraoka, Grammar, 175–79; Geiger, “Periphrastic,” 214–15.

177

For the few exceptions to the usual constructions see Muraoka, Grammar, 177. Muraoka argues (Grammar, 175, n. 71 and 72) that Fassberg is incorrect to include ידיע להוא and some similar forms in his list of this construction, though here Muraoka is surely making black and white an area that is rather grey. He asserts that this phrase “has nothing to do with the continuative, repetitive aspect.” It is true that it is not repetitive, but neither are some other periphrastic uses. That it is not “continuative” is not clear, since one might plausibly argue that the expression intends to convey that one ought to know and keep on knowing, i.e., to keep in mind.

178

It should be noted, however, that Gianto (“Aramaic,” 21) recently argued that it bears the sense rather of “trying” to do something in certain texts. See also Muraoka and Porten, Grammar, 206–207.

179

Fassberg, “Verbal System,” 72. A very similar statement is made by Muraoka, Grammar, 176.

180

Naveh and Shaked, Bactria.

181

Noted by Joosten in his response to Fassberg, “Verbal System,” 80. See also Muraoka, Grammar, 170.

182

Also Dan 4:2, 17, 33; 5:6; 6:20; 7:16.

183

See 1Q20 (apGen) 2.13 and 6.15. In the passage cited above we also see a nice example of the “historic present” participle (נשבן), on which, see Muraoka, Grammar, 174.

184

The other instances are at Dan 4:28 and 7:5.

185

This statement is especially close in wording to that in Dan 4:28.

186

See Fitzmyer, Sefîre, 19, 73. This would fit the profile of the expressions in Daniel and 1Q20 (apGen), especially that in Dan 7:5.

187

Gzella, “Erscheinungsformen.” Also see the examples cited by Li, “Participle.”

188

See Gzella, “Erscheinungsformen,” 401, who noted that employment of the participle in this way in Daniel is concentrated in Dan 3:3–7, 26–27; 4:3–5; and 5:5–10, and discussed other examples.

189

For a discussion of, and elaboration on, the Qumran evidence, see Schattner-Rieser, Grammaire, 120; Muraoka, Grammar, 174–75.

190

Li, “Participle,” wished to nuance the functions of the participle in the Aramaic of Daniel in comparison with previous approaches. While he disagrees with some ways in which the participle has been construed in earlier scholarship, my impression is that, even if adopting his conclusions, the functions of the participle in Qumran Aramaic would not differ greatly from that in Daniel.

191

Quoted from Rogland, “Performative,” 277.

192

See especially Loesov, “Present Time,” 119–23; and Rogland, “Performative.” The same shift takes place in Late Biblical Hebrew and Qumran Hebrew, on which see Rogland, Qatal.

193

For a discussion of the performative participle in Classical Syriac, see Rogland, “Syriac.” Note, too, the somewhat skeptical assessment of Li, “Participle,” 81–82.

194

Folmer, Aramaic Language, 522.

195

The same point, though with specific reference to lexical items, has recently been made by Cook, “Retroversion.”

196

Note, however, some of the changes to word order in the Qumran manuscripts of Daniel (vis-à-vis the MT) in the section on scribal variation above. It is clear, then, that even word order could on occasion be altered.

197

See especially Buth, “Functional Grammar.”

198

For Akkadian see, e.g., Huehnergard, Grammar, 19–20. For Old Persian, Hale, “Persian.”

199

See Joüon and Muraoka, Grammar, § 155c (p. 545); Buth, “Functional Grammar.” Buth spends some time discussing those who would instead describe Biblical Hebrew as a language preferring SOV word order.

200

Muraoka and Porten, Grammar, 296–313; Folmer, Aramaic Language, 521–87. Note, however, the earlier opinion of Kutscher (“Aramaic,” 363), who considered at least a good deal of Official Aramaic word order to be “eastern,” and more like Akkadian (with more free, non-Semitic word order).

201

Muraoka and Porten, Grammar, 299.

202

Buth (“Functional Grammar,” 93) writes that “Imperial Aramaic … is noted for very diverse word orders.”

203

This is also the case with prepositions other than ל.

204

The examples are from, respectively, TAD B3.3:4, TAD A2.2:16, and TAD B1.1:2. For many more examples, see Muraoka and Porten, Grammar, 296–97; Folmer, Aramaic Language, 585–87.

205

See, e.g., Rosenthal, Grammar, 56; Stefanovic, Daniel, 99–100. Cook (“Word Order”) has a somewhat qualified sense of “free,” meaning that no single order is dominant, but each is specialized to serve different purposes.

206

Coxon, “Syntax,” 119. He followed the classic treatment of Bauer-Leander, Grammatik, §§99–100. Cook observed instead that SVO is most common, and that verb-object (VO) slightly outnumbers object-verb (OV) in clauses or sentences without an expressed subject.

207

Yakubovich, “Structure.”

208

Kutscher, “Aramaic,” 400; Kutscher, “Genesis Apocryphon,” 2; Coxon, “Syntax,” 120. On the difficulty of fitting the early evidence into simplistic categories like “eastern” and “western,” see Gzella, Cultural History, 48–52. H.H. Rowley chose to characterize Daniel’s Aramaic as a Palestinian dialect of relatively late coinage (Aramaic, 106, 155–56), though this has been widely rejected in subsequent scholarship. It may be indicative of the ambiguity of the evidence that Rowley (153–54) considered Biblical Aramaic to be connected with “western” Aramaic.

209

Cook, “Word Order”; Buth, “Functional Grammar,” 94, 96–97; and Hayes, “Word Order.”

210

Cook, “Word Order,” 125.

211

A more extensive treatment is provided in Buth, Word Order. See, however, the critique of Buth’s view on Aramaic as a continuous VSO language by Kuty, Syntax; Muraoka, Grammar, 242.

212

Yakubovich, “Structure.”

213

Shepherd, Verbal System, 70. Shepherd suggests an opposition between the suffix conjugation, which is used primarily for “narration,” and the prefix conjugation, used primarily for “discourse” (see especially 73–74).

214

Muraoka, Grammar, 241–51.

215

Kutscher, “Genesis Apocryphon,” 33–34.

216

Kutscher, “Genesis Apocryphon,” 34; Muraoka, Grammar, 250.

217

Kutscher, “Genesis Apocryphon,” 33; Folmer, Aramaic Language, 536–42; Muraoka and Porten, Grammar, 308.

218

Folmer, Aramaic Language, 536–42.

219

Folmer, Aramaic Language, 754–55.

220

One attempt at this, which has remained unpublished, is the dissertation of E.H. Chandler, Word Order.

221

See Muraoka, Grammar, 249–50.

222

There are, of course, exceptions, as in 4Q541 (apocrLevib?) 9.6, which may be written in a heightened or more poetic prose that intentionally inverts the typical word order.

223

See Muraoka, Emphatic, 44.

224

Muraoka, “Aramaic,” 439.

225

Muraoka, “Aramaic,” 439–441.

226

Muraoka, “Aramaic,” 442.

227

Folmer, Aramaic Language, 589–621. See also Muraoka and Porten, Grammar, 268–71.

228

Muraoka, Grammar, 219–21; Muraoka, “Notes.”

229

On the possible reason for this, related to whether the subject moves or the object is transferred (with a more stationary subject), see Folmer, Aramaic Language, 607–9.

230

Folmer, Aramaic Language, 609–15; Muraoka and Porten, Grammar, 268–71.

231

Except in those places where קדם is used as a deferential preposition of respect for God or royalty, on which see below. “In BA the preposition ‘l always occurs when the directional element of the verb of movement is a PN [personal name], a noun or a pron. sf. denoting a living being … Both in Ezra and Daniel, the verb of movement is always linked to its directional element by ‘l when this element is a toponym or a noun denoting a location …” (Folmer, Aramaic Language, 617–18). Regarding Qumran Aramaic, Muraoka (Grammar, 219) writes that, “[i]n the overwhelming majority of cases, however, indirect, prepositional government is the rule: ל– marks a place as a destination and על a person.”

232

Muraoka (Grammar, 220–21) noted that the verb שלח acts somewhat differently than the other verbs of movements surveyed by him for Qumran Aramaic, tending to use instead –ל even for persons and other animate objects (e.g., ישתלח לכול בני [ע]מ̊ה in 4Q541 [apocrLevib] 9i.2). This observation is due in large part to 1Q20 (apGen) only (though note 4Q530 [EnGiantsb] 2ii+6–12(?).21). Folmer (Aramaic Language, 620) observed the same trend in Official Aramaic.

233

See Folmer, Aramaic Language, 619–21, though the evidence is not always compelling due to the small sample size available in some cases.

234

Folmer, Aramaic Language, 620. The standard preposition for the inanimate directional object in the targums is still –ל.

235

Muraoka, “Notes,” 39. See also Folmer, Aramaic Language, 620, n. 104, and the bibliography cited there.

236

Joosten, “devant.” For the Official Aramaic evidence, see also Folmer, Aramaic Language, 590. Joosten oberved that the same expression is used by some of the Septuagint translators, betraying the background influence of Aramaic on the Greek of the Septuagint. See Joosten, “Septante.”

237

Joosten, “devant,” 92.

238

Joosten, “devant,” 94–95.

239

As Joosten notes (“devant,” 95, n. 38), it probably also occurred at 4Q243 (psDana) 2.1.

240

See, e.g., Nahal Hever/Se’elim 7.5–6, 9.20, and 50.15 in DJD 27, or P. Yadin 7.22.

241

The same practice is found in Biblical Hebrew.

242

See Kutscher, “Genesis Apocryphon,” 20–21; Muraoka, “Aramaic,” 439.

243

4Q550 (Jews at the Persian Court) 5+5a.7; 4Q554 (NJa) 1iii.13; 4Q554a (NJb) 1.13; 4Q556a (Prophecyb) 5i–ii.7; 4Q559 (papBibChronology) 3.8, 3.9; 5Q15 (NJ) 1i.1; 11Q10 (Job) XXXV.9, XXXVIII.9. Other possible occurrences are at 4Q201 (Ena) 1v.5 and 4Q535 (Birth of Noahb) 3.4. It should also be noted that ית occurs in the phrase ולא ישבקון יתהון in the Manchester fragment of the Aramaic Levi Document from the Cairo Geniza. The Geniza copy typically corresponds to the earlier Qumran copies where comparisons are possible, and so it is probable that the phrase was once present in the Qumran copies of this work. For the text and accompanying discussion, see Drawnel, “Fragment.”

244

As noted already by Stadel, “Influences,” 395.

245

1Q20 (apGen) 22.23, 31; 1Q23 (EnGiantsa) 1+6+22.4; 4Q204 (Enc) 1i.24; and 11Q18 (NJ) 20.2.

246

For references see Schwiderski, DARI1, 179

247

In P. Yadin 1:36 (Yadin, Bar-Kokhba, 180).

248

Dan 6:3; 1Q20 (apGen) 20.7; 2Q26 (EnGiants) 1.2; and 4Q206 (Ene) 1xxvi.20, 4i.20.

249

Schwiderski, DARI1, 651. For the Egyptian Aramaic letter from Elephantine (ca. 475 BCE), written on an ostracon, see Schwiderski, DARI2, 151 (D:7.9(5).4–5), which reads, concerning a servant girl, יכתבוה על דרעה עלא מן כתבא זי על דרעה (“Tattoo her on her arm above the tattoo that is [already] on her arm”). The only other close formulation is in the Elephantine Ahiqar text (Schwiderski, DARI2, 87 [C:1.1 (Kol.11).162]), where we find the phrase לעלא מנה.

250

For later dialects, refer to Jastrow, DTTM, 1069; Sokoloff, DJA, 70; Sokoloff, DJPA, 315; Sokoloff, DJBA, 630.

251

Already noted by Kutscher, “Genesis Apocryphon,” 7. See also Cook, DQA, 95–96, who draws attention to the related construction מן טלל in 11Q10 (Job) XXVIII.7.

252

Compare the corresponding בגללך in Gen 12:13.

253

To this list we may now wish to add the observations of Stadel on the use of כל “all, every” followed by plural determined and indeterminate nouns in Official Aramaic, Biblical Aramaic, and Qumran Aramaic. The results of Stadel’s article are somewhat ambiguous, and Biblical Aramaic clearly has a different approach, but there does seem to be a general correspondence between how כל is employed in Official Aramaic, Qumran Aramaic, and one of the later Judean-Nabatean papyri. See Stadel, “Syntagm.”

254

Stadel, Hebraismen; Stadel, “Influences.”

255

Stadel, “Influences.”

256

Driver, Daniel, lvi–lxiii. See also the vibrant rejoinder by Tisdall, “Daniel,” who extensively interrogated the loanwords in Daniel.

257

See now the excellent lists in Muraoka, Grammar, 78–81, and the bibliographic items cited there. This is in keeping with the character of Official Aramaic, which also contains many Akkadian and Persian words (cf. Muraoka and Porten, Grammar, 342–56).

258

Cook, “Dialectology”; Cook, “Aramaic”; and especially Cook, “Retroversion,” 362–363.

259

Schattner-Rieser, Grammaire, 25; Koller, “Dialects,” 212–13. A well-known example of varied geographic distribution is Muraoka’s suggestion (“Aramaic”) that 11Q10 (Job) is of eastern origin, as opposed to the more western origin of 1Q20 (apGen).

260

On the evidence for a revival of Hebrew during the Hasmonean period, see now Machiela and Jones, “Revival.”

261

Koller, “Dialects”; Gzella, “Dating.”

262

Of course, in certain situations this distinction might remain useful, such as when writing a grammar focusing on Biblical Aramaic alone for students with a view to the Jewish or Christian canons. This should remain separate, however, from the broader dialectical question addressed here, and in this case Biblical Aramaic should be understood as a canonically-limited sub-corpus of the larger Biblical Aramaic/Qumran Aramaic dialect.

263

Rowley, “Notes.”

264

Folmer, Aramaic Language. The same point had been made earlier, though much more laconically, by Ginsburg, “Aramaic,” 232, and Greenfield, “Standard,” 116.

265

Albertz, Daniel 4–6; Albertz, “Setting”; Kratz, Translatio; Collins, Daniel, 29–38; Newsom, Daniel, 6–12.

266

On the story collection format of Daniel, see now Holm, Courtiers.

267

My own preference, for the reasons often provided since the initial argument of Charles (Daniel, xlvi–xlviii), is that of translation into Hebrew from Aramaic. This was most likely done because the Hebrew beginning and ending resulted in a “Hebrew book,” thereby facilitating Daniel’s Jewish use from the Hasmonean period onward.

268

It is worth noting that a somewhat different approach was taken with Tobit, which was apparently translated entirely into Hebrew, as seen in 4Q200 (Tobe). On this topic, see Machiela, “Hebrew of Tobit”; Dimant, “Hebrew Copy.”

269

The latter topic has been debated, and is by no means divorced from the question of language. For a recent overview of the issues and various opinions, see Doering, Letters, 122–25. The Aramaic sections of Ezra are 4:11–22, 5:7–17, 6:2–12, and 7:12–26. A useful account of older opinions, on which later scholarship tends to elaborate, is provided by Torrey, “Ezra,” 210–14.

270

Folmer, Aramaic Language, 41, 754; Williamson, “Documents,” 54–62. See also Gzella, Cultural History, 206–7.

271

Schwiderski, Briefformulars, 360, 362, 364–68.

272

Rowley, “Notes,” 128; Fitzmyer, Commentary, 35. See also Fassberg, “Verbal System,” 76–77.

273

Fitzmyer, Commentary, 35.

274

Dimant, “Qumran Aramaic,” 204. See also Gzella, Cultural History, 208.

275

Dimant, “Qumran Aramaic,” 204. Italics have been added.

276

Fassberg, “Verbal System,” 76. Italics have been added.

277

Muraoka, Grammar, 176. Italics have been added. On the same page he deems the phrase ו]א̊נה הוית חזה עד ארעא חפית מין in 4Q206 (Ene) 4i.18 to be “another epigone of Daniel.”

278

Muraoka, Grammar, 110. Italics have been added.

279

Pfann, “Daniel and Ezra.”

280

A recent example of the bifurcation between Biblical Aramaic (Daniel and Ezra, in particular) and the Aramaic Qumran literature is found in the overview of ancient Aramaic by Abraham Tal, “Aramaic.” Tal makes the typical distinction between Biblical Aramaic and Standard Literary Aramaic (ארמית ספרותית סטנדרדית, on which, see further below), the latter including the Qumran texts. Tal cites as evidence of this distinction, for example, the fact that Biblical Aramaic uses ה to signify a vowel at the end of words, while Standard Literary Aramaic uses א, as shown in the Genesis Apocryphon. However, this and the other examples cited by Tal are too simplistic, ignoring the full situation of the Qumran copies of Daniel and Ezra. Once that evidence is considered, Tal’s examples simply fail to convince, and his distinction loses its force. With regard to the case of final vowels just cited, see the “Interchange of ה, ‪א‬ (or י) for weak verb ending” section in the catalogue of scribal changes, above. See also Tal, “Raison d’Être,” 360.

281

This may have been done for reasons of consistency with the causative verb spellings, or even because it homogenized the forms with the Hebrew spellings throughout the rest of the book. As noted by Pfann and Cook, the forms with א align with the earlier Official Aramaic texts.

282

This accords with the summation of Gzella, Cultural History, 208. See also Flesher, “Aramaic,” at 86–89.

283

In fact, Beyer (ATTM1, 28–35; Language, 14–21) includes both Biblical Aramaic and Qumran Aramaic as later sub-types of Official Aramaic, which he calls Imperial Aramaic (Reichsaramäisch). The closest literary successor of our dialect is the Aramaic of the Onkelos and Jonathan targums.

284

Of course, this is not to discount the possibility that some later Fortschreibung took place after this original stage, as a number of scholars have suggested was the case for the “little horn” (most likely representing Antiochus IV Epiphanes) in Dan 7:8.

285

On this issue, see now the helpful overview of Gzella, Cultural History, 45–52.

286

Fitzmyer first published his classificatory system in a footnote of his original, 1966 edition of Commentary, 19, n. 60. A more extended, article-length discussion later appeared as “Phases.” Beyer’s system can be found in Beyer, ATTM1, 23–71; and Beyer, Language.

287

See recently Gzella, Cultural History, 47–48.

288

Gzella (Cultural History, 50–51) advocates for Beyer’s general system, while recognizing some of its shortcomings.

289

Beyer, ATTM1, 34; Beyer, Language, 20.

290

Gzella, Cultural History, 230. This follows very closely Beyer’s view in ATTM1, 34. It should also be noted that there are 120–130 Aramaic manuscripts from Qumran, not “compositions.” Many of these manuscripts are one of multiple copies of the same composition.

291

His statements about dating (Cultural History, 233) might lead one to think that he means only the preserved copies, but this is never made clear.

292

For an overview of this topic, see Machiela, “Compositional Setting.”

293

Greenfield, “Standard.”

294

Wise, Language, 282.

295

“Qumran Aramaic is also Standard Literary Aramaic but written on Palestinian soil. It is in this language that Tobit, Enoch, the Testament of Levi, and the ‘Daniel’ pseudographs were written.” Greenfield, “Standard,” 116–17.

296

Gzella, Cultural History, 49 (see also 165). So, too, Cook, “Dialectology,” 1; and Cook, “Perspective.”

297

See, in addition to those already named, Flesher, “Aramaic,” 87–89. This designation has been affirmed most recently (2018) by Tal, “Aramaic.” It seems to me, however, that Tal draws a sharper distinction between Biblical Aramaic and Standard Literary Aramaic than did Greenfield.

298

Sokoloff, “Dialects,” 746. He is followed by Flesher, “Aramaic.”

299

Fassberg, “Verbal System,” 67, 76–78; Fassberg “Language,” 136; Fassberg, “להוא.” This title is cited with approval by Joosten “Exorcise,” 348.

  • Collapse
  • Expand