Hebrew Adverbialization, Aramaic Language Contact, and mpny ʾšr in Exodus 19:18

For use by the Author only | © 2015 Koninklijke Brill NV

Semitic Languages in Contact

Edited by

Aaron Michael Butts

LEIDEN | BOSTON


Contents

Preface vIIList of Figures xiAbbreviations xiiContributors xx

A Thamudic B Abecedary in the South Semitic Letter Order 1Ahmad Al-Jallad and Ali Al-Manaser

Ethiopian Semitic and Cushitic. Ancient Contact Features in Ge‘ez and Amharic 16

David Appleyard

Hebrew Adverbialization, Aramaic Language Contact, and mpny ʾšr in Exodus 19:18 33

Samuel Boyd and Humphrey Hardy

The Distribution of Declined Participles in Aramaic-Hebrew and Hebrew-Aramaic Translations 52

Yochanan Breuer

The Proto-Semitic “Asseverative *la-” and the Innovative 1sg Prefĳixes in South Ethio-Semitic Languages 68

Maria Bulakh

Egyptianizing Features in Phoenician and Punic Inscriptions from Egypt 97

David Calabro

Head-Marking in Neo-Aramaic Genitive Constructions and the ezafe Construction in Kurdish 114

Eran Cohen

Notes on Foreign Words in Hatran Aramaic 126Riccardo Contini and Paola Pagano

Language, Writing, and Ideologies in Contact: Sumerian and Akkadian in the Early Second Millennium bce 158

C. Jay Crisostomo

vi contents


Inner-Semitic Loans and Lexical Doublets vs. Genetically Related Cognates 181

Lutz Edzard

Structural Change in Urban Palestinian Arabic Induced by Contact with Modern Hebrew 198

Uri Horesh

Language Contact as Reflected in the Consonant System of Ṭuroyo 234Otto Jastrow

Lexical Borrowings in the Eastern European Hasidic Hebrew Tale 251Lily Kahn

Possible Ugaritic Influences on the Hurrian of Ras Shamra-Ugarit in Alphabetic Script 267

Joseph Lam

The Lexical Component in the Aramaic Substrate of Palestinian Arabic 280

Mila Neishtadt

The Classifĳication of Hobyot 311Aaron D. Rubin

Expression of Attributive Possession in Tunisian Arabic: The Role of Language Contact 333

Lotfĳi Sayahi

Aramaic Loanwords in Gǝʿǝz 348Jürgen Tubach

Language Contact between Akkadian and Northwest Semitic Languages in Syria-Palestine in the Late Bronze Age 375

Juan-Pablo Vita

Semitic Languages in Contact—Syntactic Changes in the Verbal System and in Verbal Complementation 405

Tamar Zewi and Mikhal Oren

Index 423

© koninklijke brill nv, leiden, 2015 | doi 10.1163/9789004300156_004For use by the Author only | © 2015 Koninklijke Brill NV

Hebrew Adverbialization, Aramaic Language Contact, and mpny ʾšr in Exodus 19:18

Samuel Boyd and Humphrey Hardy

Adverbializers are a subclass of subordinators, or subordinating conjunc-tions, that mark an intra-clausal, adverbial relation.1 Hebrew expresses adverbialization in a variety of ways. In this article, two syntactic patterns of adverbialization with prepositions are surveyed from the perspective of his-torical linguistics. Both of these arrangements are found in the earliest strata of Hebrew. Later evidence demonstrates a shift in the usage of one pattern in favor of another. We propose a historic process through which these patterns emerged in Biblical Hebrew (BH), track the diachronic distribution in Hebrew, demonstrate the role of language contact on the changes, and apply this framework to a particularly difffĳicult interpretive problem in the Hebrew Bible.

1 The Development of Adverbializers

The primary means of adverbialization in BH is a function of prepositions. Two productive structures consist of prepositions governing a following clause or an embedded relative clause: (1) PREP+S and (2) PREP+REL+S.

אחרי הסבו אתו (1) ʾaḥăre hesabbu ʾoṯo AFTER-PREP lead.around-SC.3.M.PL DOM+it ‘After they led it around’ (1Sam 5:9)

* Abbreviations: 1 = fĳirst person; 3 = third person; advz = adverbializer; BH = Biblical Hebrew; c = common gender; dom = direct object marker; LBH = Late Biblical Hebrew; m = masculine gender; N = noun; np = noun phrase; pc = prefĳix conjugation; pl = plural; pp = preposition phrase; prep = preposition; rel = relative; s = sentence; SBH = Standard Biblical Hebrew; sc = sufffĳix conjugation; sg = singular.

1 The use of the term ‘adverbializer’ to indicate an element that converts a constituent into an adverb (Gensler 2000) is not in view here. Rather, the standard description of this term from language typology is employed designating a type of subordinating conjunction which forms an adverbial clause (Thompson and Longacre 1985).

34 Boyd and Hardy


עד שקמתי (2) ʿaḏ šaq-qamti UNTIL-PREP REL+arise-SC.1.C.SG ‘Until I arose’ (Judg 5:7)

The syntax of these patterns fĳinds its origin in the Semitic noun phrase (NP), particularly, the genitive construction. Through examining these modifying structures, the development of adverbializers in Hebrew is detectable.

In cross-linguistic comparison, the modifĳication of nouns may be classifĳied into three constructions: genitives, adjectives, and relatives.2 These may be exemplifĳied in BH with the following: (3) a genitive (i.e. the Semitic construct phrase), (4) an adjective phrase, and (5) a relative construction. The fĳirst two expressions are types of NPs, and the last consists of a relative clause.

בני האלהים (3) bne hɔ-ʔɛlohim sons.of the-gods ‘the gods’ sons’

ההרים הגבהים (4) hɛ-hɔrim hag-gḇohim the-mountains-M the-high-M.PL ‘the high mountains’

הטובה אשר עשה לדוד (5) haṭ-ṭoḇɔ ʔăšɛr ʕɔśɔ lḏɔwiḏ the-good REL do-SC.3.M.SG FOR+David ‘the good that he did for David’

A fourth construction, though less frequent, is known in BH. The modifĳication strategy in Example (6) evidences a NP (kɔl-yme ‘all of the days of ’) in con-struct with a clause (hiṯhallaḵnu ʔittɔm ‘we walked with them’).3 It is weakly diffferentiated from the genitive construction. As such, the construction may demonstrate a mediating position between the syntax of genitive and relative constructions.

2 Gil 2013.3 This expression has been understood alternatively as having a covert relativizer, that is, as a

null-headed relative (Holmstedt 2001; 2006).

35Hebrew Adverbialization, Aramaic Language Contact


כל־ימי התהלכנו אתם (6) kɔl-yme hiṯhallaḵnu ʔittɔm all.of+days.of walk.about-1.C.PL WITH+them ‘all of the days we walked with them’

Goldenberg classes these kinds of constructions as examples of “endocentric modifĳication” or simply “attributive.”4 He demonstrates the correspondence between the genitive and relative structures throughout the Semitic languages. Examples from Syriac and Akkadian are provided in Chart 1 with (a) a nominal head, (b) a pronominal head, and (c) a nominal head and an appositional co-referential genitive construction. In this paradigm, the fĳirst two constructions are parallel to Example (3) above and the third is parallel to Example (5).

CHART 1 Attributive Noun Phrases

Apposition Head Attributive Gloss

Syriac (a) deḥlaṯ alɔhɔ ‘fear of god’(b) d- alɔhɔ ‘N of god’(c) deḥlṯ ɔ d- alɔhɔ ‘fear of god’

Akkadian (a) bīt ilim ‘house of god’(b) ša ilim ‘N of god’(c) bītum ša ilim ‘house of god’

Regarding phrases of the relative type as in Example (5), Goldenberg claims that these patterns are “shown to be syntactically equivalent to genitive nomi-nal forms.”5 The constructions with a verbal attributive (Chart 2), then, express the same basic relation as those with a nominal attributive. It is also impor-tant to note that relatives are derivative in each of the Semitic languages.6 That is to say, the constructions with relatives are secondary and developed inde-pendently in the individual daughter languages. As a result, these common NP types need to account for the construction similarities.

4 Goldenberg 1998.5 Goldenberg 1998: 56–57.6 Huehnergard 2006.

36 Boyd and Hardy


CHART 2 Attributive Relative Clauses


Akkadian (a) kasap išqulu ‘the money (that) he paid’(b) ša izzazzu ‘N that stands’(c) elippum ša- uṭebbû ‘the boat that he sank’

Following Goldenberg, the forms of NP modifĳication in BH may be collapsed into two construction types.

Chart 3 presents these two BH varieties using the previous examples. Type 1 is the genitive construction with either a nominal or verbal attributive that may be headed by either a preposition or a relative. Type 2 is the attributive adjective construction—hɛ-hɔrim hag-gḇohim ‘the high mountains’.

CHART 3 BH Modifĳication of NPs


Type 1 (a) bne hɔ-ʔɛlohim ‘the gods’ sons’(b) kɔl-yme hiṯ hallaḵ nu

ʔittɔm

‘all of the days we walked with them’

(c) haṭ-ṭoḇ ɔ ʔă šɛr ʕɔśɔ lḏ ɔwiḏ ‘the good that he did for David’

Type 2 hɛ-hɔrim hag-gḇ ohim ‘the high mountains’

Classical grammatical and historical linguistic approaches have established that most BH prepositions originate from NPs.7 Aspects of the originating constructions are preserved even as these lexemes acquire new statuses and

7 Nearly every BH grammar presents the general word-class change of Noun to Preposition (see, e.g., Blau 2010: 283). Using grammaticalization theory, Hardy (2014) provides more details for the origin and development of BH prepositions that may be etymologically related to nominals.



are incorporated into the linguistic system as function words. In particular, prepositions—and by extension adverbializers—emerge from these noun- modifĳication constructions displaying the syntactic arrangement and mor-phology of their originating forms. The emergence of adverbialization in BH is situated at the nexus of genitive constructions and the development of prepositions. Combining these two components, a plausible development of adverbialization may be outlined. The syntax of BH prepositions patterns on the genitive constructions of the fĳirst type (Chart 4), corresponding to the previously discussed genitive of NPs. These preposition constructions consist of two varieties where the complement may be a NP (a) or a sentence either without a relative (b1) or with a relative (b2). In the case of the relative, the construction follows the genitive type of Chart 3 (b) kɔl-yme hiṯhallaḵnu ʔittɔm ‘all of the days we walked with them’, where the relative is the attributive in the larger construct phrase. The result is two productive BH patterns, PREP+S and PREP+REL+S, whereby adverbialization could be expressed.

CHART 4 BH Prepositions

PREP Complement Gloss

Type 1 (a) ʔaḥă re bne hɔ-ʔɛlohim ‘after the gods’ sons’(b1) ʔaḥă re hiṯ hallaḵ nu ʔittɔm ‘after we walked with them’(b2) ʔaḥă re ʔă šɛr ʕɔśɔ lḏ ɔwiḏ ‘after what he did for David’

2 Linguistic Variation in Hebrew Adverbializers

The structure suggested in the previous section is based on the reconstruction of the genitive construction and the emergence of prepositions functioning as adverbializers. Before turning our attention to outlining the proposed contact-induced change, we shall review more closely the linguistic variation in BH and later Hebrew. What will be demonstrated is that the adverbializer construc-tions followed a clear development trajectory of incremental change. At each step, a lone pattern emerged from polysemy, and in turn the grammatical func-tion was reshaped as a result of internal and external influences.

BH evidences both patterns of adverbialization discussed previously. However, the patterns are not evenly distributed. The PREP+S pattern is found

38 Boyd and Hardy


fĳifty-nine times.8 Of the nine diffferent prepositions (ʔɛl, ʔaḥar, ʔaḥăre, bilʕăle,

kmo, kʕal, min, ʕad, and ʕal), the most frequent semantic relation is the tem-poral function—either AFTER (ʔaḥar/ʔaḥăre) or UNTIL (ʕad). Three-fourths of the adverbializers consist of this latter temporal element. The only “insepa-rable” preposition is the expanded independent form kmo ‘like, as’. The dis-tribution of this PREP+S pattern within the BH corpus is diverse. There are nineteen examples in the Pentateuch and Deuteronomistic History, fĳifteen in the prophets, nineteen in the writings, and six examples in Chronicles and Ezra-Nehemiah. Isaiah (First and Second), Psalms, and Job witness the largest number of instances.

The PREP+REL+S pattern is attested more widely within BH, both pertain-ing to types and tokens. Twelve diffferent prepositions are evidenced amount-ing to more than seven hundred examples. The inseparable variety include: b- ‘in’, k- ‘like, as’, l- ‘to, for’, and me- ‘from’ (< min-). Eight separable prepositions are evidenced (ʔɛl, ʔɛt, ʔaḥar, ʔaḥăre, min, ʕad, ʕal, and taḥat). The overwhelm-ing majority of these examples (515 instances) consists of a lone type, kaʔašɛr ‘just as’. The distribution of the PREP+REL+S pattern throughout the BH cor-pus is not altogether diffferent from that of the previous construction.

Comparing the two patterns, one fĳinds considerable overlap in semantics, usage, and dispersion within the corpus. Seven of these constructions use the same preposition element (Chart 5). Six distinct prepositions are evidenced uniquely with PREP+S (bilʕăle, kʕal) or PREP+REL+S (ʔɛt, b-, l-, taḥat). Two lex-emes demonstrate formal variances relating to the contrast of the indepen-dent form as compared to the dependent: kmo and kaʔašɛr; min and meʔašɛr. Both forms are found in early texts, so-called Standard Biblical Hebrew (SBH), and the later corpus, Late Biblical Hebrew (LBH). The only readily seen difffer-ences between the two constructions are semantic. This variance is based on the distribution of the specifĳic preposition element where the PREP+S pattern is typically temporal and the PREP+REL+S pattern is comparative.

8 Gen 38:11; Exod 15:16 (2x); Lev 14:43; 25:48; Num 23:24; Deut 33:11; Josh 2:22; 10:13; 1Sam 1:22; 5:9; 14:19; 2Sam 10:5; 21:10; 1Kgs 11:16; 15:29; 2Kgs 3:25; 7:3; 10:17; 1Chron 15:12; 2Chron 21:15; 29:34; 36:21; Ezra 3:11; 8:29; Job 8:21; 14:6; 27:5; 32:11; 34:32; 42:7; Ps 57:2; 58:8; 71:18; 73:17; 94:13; 110:1; 132:5; 141:10; Prov 7:23; 12:19; 23:7; Is 22:14; 26:18, 20 (2x); 32:15; 42:4; 59:18; 62:1, 7 (2x); Jer 41:16; Lam 3:50; Ezek 28:15; 39:15; Dan 11:36; Hos 10:12; Zech 10:8.



CHART 5 Prepositions found in Adverbializer Constructions

PREP+S PREP+REL+S

ʔɛl ʔɛl

ʔɛt

ʔaḥar ʔaḥar

ʔaḥă re ʔaḥă re

b-

bilʕă le

kmo k-

kʕal

l-

min me-

ʕad ʕad

ʕal ʕal

taḥat

The linguistic variation seen with these patterns is not static. And adverbializa-tion did not cease developing but continued to evolve in later strata of Hebrew. The evidence from the Hebrew of the Dead Sea Scrolls (3rd century BCE–1st century CE) indicates a preference for the PREP+REL+S pattern to the exclu-sion of the other. These constructions include many of the same prepositions as in BH (ʔḥr ʔšr, ʔl ʔšr, ʔt ʔšr, bʔšr, bʕbwr ʔšr, lʔšr, mʔšr, mš, kʔšr, and ʕd ʔšr). The semantics of the most common construction kʔšr, however, have expanded from the BH usage as a comparative to include the temporal relation (e.g. CD-A 15: 12 wkʔšr yqym ʔwtw . . . ‘now when he sets himself up . . .’). In contrast, no example of the well-attested BH construction ʕad+S is found in these docu-ments. Instead one fĳinds ʕad ʔšr ‘until’ as its equivalent at Qumran.9 It is rea-sonable to conclude that the general pattern with the relative has expanded to be the sole adverbialization strategy in these documents. This change appears to be the result of an internal shift toward pattern regularizing in the linguistic strata subsequent to BH.

With the early Rabbinic period (70 CE–3rd century CE), the situation had evolved further. The adverbializer pattern remained PREP+REL+S, but the

9 Five examples are extant in the Temple Scroll (11Q19 45:8, 17, 18: 50:3; 58:20).

40 Boyd and Hardy


lexeme functioning as the relative shifted from ʔšr to the monosyllabic pro-clitic š-.10 The former lexeme is known from biblical quotations and a handful of what appear to be static idioms, such as kʔšr—which itself is found only about a dozen times. In Rabbinic Hebrew, the latter pattern PREP+š(REL)+S is the productive means of marking adverbialization (e.g., ʔḥr š-, byn š-, kmw

š-, kš-, lpy š-, mš-, ʕl š-, etc.).11 For instance, only three examples of the adverbi-alizer ʕd ʔšr are evidenced in the Tosefta, whereas several hundred examples consist of ʕd š- ‘until’ throughout the corpus of Rabbinic Hebrew.

In light of the origin of adverbialization and the distribution of the construc-tions, a plausible picture of the development may be sketched. The BH poly-morphous adverbializer patterns, PREP+REL+S and PREP+S, were reduced to the single construction with the relative ʔšr by the early centuries BCE. This form appears to be a conservative feature of the Hebrew known from the Dead Sea Scrolls.12 Rabbinic Hebrew perpetuated the PREP+REL+S adverbialization pattern but exchanged the relative š- in place of ʔšr.13

3 Language Contact and Adverbializers in BH

In the previous sections, we explored the emergence of the two syntactic patterns of adverbializers in BH and examined the changes to these patterns in later Hebrew. In the following, we analyze the role of language contact of Aramaic in the evolution of Hebrew adverbialization. Doing so allows for a linguistic basis in approaching a particularly difffĳicult issue in Exod 19:18, which will be examined in the fĳinal section.

As Pat-El argues extensively, Aramaic lost “independent subordinators” early in its existence, and, more specifĳically, “never developed independent adverbial subordinators.”14 Instead, various dialects of Aramaic formed adver-bial subordination through extensive use of the relative pronoun. According to Pat-El, this role for the relative particle increased over time in the language and any form of subordination required the use of the relative.

10 For the two primary derivational options for these relatives, see Heuhnergard 2006 and Holmstedt 2007. The replacement of ʔšr with š- may be seen as a broader change in Rabbinic Hebrew (Fernandez 1999: 50).

11 See further Segal 1927: 148.12 Qimron 1986: 82.13 Fernandez notes that while the origin of the relative/genitive particle של may be found

in Hebrew, “the extent of its use has been influenced by the Aramaic relative (6 :1999) ”די. This connection should perhaps be expanded further to include the proclitic relative šɛ-.

14 Pat-El 2008: 59.



With this mapping of the historical development of adverbial subordina-tion in Aramaic established, Pat-El proposes a contact-induced change in BH causal subordination due to Aramaic influence.15 This change in BH diverges from the standard formation of this type of subordination that marks relation-ships between clauses. Both languages shared kî as a causal conjunction; how-ever, Aramaic lost this conjunction early in its history.16 In order to compensate and formulate causal subordination, Aramaic developed a variety of expres-sions using the relative pronoun.17 Pat-El identifĳies in LBH two Aramaic loans on the basis of constructions with the relative pronoun. The fĳirst is Example (7) where Aramaic bdyl ‘because of ’ corresponds to BH ba-ʾăšɛr l- ‘on account of ’:

(7) Aramaic: bdylh

b-dy-l-h

PREP+REL+PREP+him ‘because of him’18

BH: באשר ל ba-ʾăšɛr l-

PREP+REL PREP ‘on account of . . .’ (Jon 1:8)

The second example relates Aramaic bdyl d ‘because’ to BH b-šɛ-l ʾăšɛr ‘though’:

(8) Aramaic: bdyl dy hwʾ ydʿ

b-dy-l dy hwʾ ydʿ

PREP+REL+PREP REL he VERB ‘because he knows’19

15 Pat-El 2012.16 We would like to thank A. M. Butts for drawing our attention to Syriac kay, which may be

a remnant of this particle, though in Syriac it does not function as a clausal conjunction.17 According to Pat-El (2012), while BH contains some instances of causal subordination

through the adverbialization pattern PREP+REL+S, this usage of the construction is rare. For example, Gen 39:9 contains b-ʾšr used causally (‘He has not kept back anything from me, except you, because [b-ʾšr] you are his wife’), an adverbializer that can also be used spatially (Judg 17:8).

18 This particular construction in Aramaic tends to occur with pronominal sufffĳixes. Pat-El (2012) cites Targum Onkelos Gen 3:17 (bdylk ‘because of you’) and 11QTgJob 38:3 (bdylh ‘because of him’).

19 4Q197 frg. 4 ii 4, as cited in Pat El 2012.

42 Boyd and Hardy


BH: בשל אשר b-šɛ-l ʾăšɛr

PREP+REL+PREP REL ‘because (though) . . .’ (Eccl 8:17)

Pat-El’s study of Aramaic’s influence on causal subordination in BH serves as a basis for further theorization of issues of language contact and for the devel-opment of adverbializers in BH. Pat-El makes the case for language contact in the development of causal subordination in BH and places the issue of syn-tactic borrowing within a larger framework of contact linguistics discussed in the work of Thomason and Kaufman.20 Yet the very notion of syntactic borrowing has become a controversial area in the study of language contact. Linguists such as Winford have challenged the assertion that structure can be independently borrowed from lexicon, suggesting that syntactic borrowing is instead the product of extensive calquing.21 In this fashion, the syntax from the source language (SL) is borrowed into the recipient language (RL) as the agents of change are the speakers of the RL modelling expressions through loan translations.

Additionally, Winford has argued many of the cases claimed to be syntactic borrowings are actually examples of SL agentivity;22 if so, then the label “bor-rowing,” which usually denotes RL agentivity, is problematic in these examples between Aramaic and BH. Pat-El notes that “syntactic interference happens in a situation when the source language is considered dominant over the receiv-ing language for any number of reasons.”23 In this situation, Winford would demure the label “borrowing” and rather call such syntactic contact-induced change the result of shift or imposition.

In defense of syntactic borrowing, Heine and Kuteva have examined situ-ations in which syntax can be borrowed without language shift, showing that

20 This framework has been further developed in Thomason 2001. In more recent writings, Thomason has acknowledged the similarities between her typology developed with Kaufman and that derived independently in the work of Frans van Coetsem, a similarity that van Coetsem had also already acknowledged (Thomason 2003: 692; see also the brief comments in Thomason 2001: 95; van Coetsem 1990: 60–62). As Thomason and Kaufman note, they were not the fĳirst to notice the linguistic processes under examination, but rather they sought to refĳine the analysis of these linguistic outcomes through the accu-rate labeling and assessment of the linguistic processes underneath contact-induced phe-nomena (Thomason and Kaufman 1988: 42).

21 Winford 2003: 97; 2010: 175–77.22 Winford 2010: 179–81.23 Pat-El 2012: 251.



structural linguistic elements can be borrowed.24 While Pat-El is, therefore, on safe ground in her analysis of syntactic borrowing, another topic in the study of language contact is relevant for her discussion, particularly as it relates Aramaic influence on Hebrew. That topic is the matter (MAT) and pattern (PAT) distinction.

Sakel and Matras defĳine MAT and PAT borrowing in the following manner: MAT loans refer to the process in which any morphological element and “its phonological shape” is transferred from the SL to the RL; PAT loans involve the borrowing of patterns and shapes of constructions from the SL to the RL without transferring the morphemes themselves.25 (“Morphemes” in this defĳi-nition include lexemes.) MAT loans do not require a high level of bilingualism whereas PAT loans typically involve a higher degree of profĳiciency in both the SL and RL in the speech community. This feature of PAT loans is logical inas-much as speakers of the RL have to be bilingual enough in the SL to be able to identify the pattern and then produce it sensibly in the RL but using the mor-phemes and linguistic features of the RL.

The MAT/PAT distinction helps make sense of some of Pat-El’s observations. For example, in her discussion of Aramaic influence on causal subordination in Jon 1:8 and Qoh 8:17 she states that “all the elements of the preposition and conjunction shown above in Jonah and Qoheleth are original Hebrew features, but their new structural confĳiguration as well as the causal function are the

24 For this counterargument to Winford, see Heine and Kuteva 2005: 157–59. They use an example from Malcom Ross’ study of contact between Takia and Waskia. It is cer-tain that Proto-Western Asiatic, the language family of which Takia is a part, had a determiner (‘the’) that preceded the noun; however, Takia developed a postpositioned determiner, based on the model of Waskia. The process through which this change occurred in Takia involved patterning the Takia near deictic morpheme, which follows the noun when used attributively, on the SL (or, “model language” to use Heine and Kuteva’s terminology), namely Waskia. According to Heine and Kuteva, structural changes such as this example show that syntactic influence (in this case, through grammatical replica-tion, though Takia also borrowed the SOV syntax from Waskia as an independent devel-opment) can occur without language shift attested as being a broader part of the contact situation. Since this situation does not involve shift, the example of the postpositioned determiner in Takia shows how structural borrowing can occur even without a (or, as a separate process from) change in word order: the Proto-Western Asiatic determiner was not simply placed behind the noun in Takia; rather, another lexeme that already fĳit into this postpositioned syntactic slot (the near deictic pronoun) was desemanticized through grammaticalization and modeled on the defĳinite article in Waskia. This example shows, then, a clear case in which syntax is borrowed without language shift. For further argu-ments concerning the borrowing of structural linguistic elements, see Butts Forthcoming.

25 Matras and Sakel 2007: 15.

44 Boyd and Hardy


result of language contact.”26 This statement conforms to the defĳinition of the PAT loans mentioned above.

Finally, a sociolinguistic facet of the MAT/PAT distinction also undergirds and supports Pat-El’s thesis about language contact and causal subordination in Aramaic and Hebrew. In Matras and Sakel’s study, they catalogued over twenty languages and their various contact situations, noting the sociological backgrounds of each.27 In doing so, they found that the predominant sociolin-guistic context for PAT loans involved situations in which the SL was not only a socially dominant language generally, but more specifĳically functioned as a lin-

gua franca. This aspect of PAT loans, and the fact that causal subordination in a few examples in BH can be explained due to this contact-induced change with Aramaic, is consistent with a larger argument in Pat-El’s article. She claims that this feature in BH does not occur in SBH, but only in LBH. While her point was to display the fact that diachrony in BH is discernable, the relevance of her diachronic argument here is that the period when such PAT loans appear in BH aligns with the status of Aramaic as a lingua franca in the Persian period.28

In this section, we examined the role of contact-induced change from Aramaic causal subordination into BH. We afffĳirmed the insights of previous studies that contact-induced change is present; however, we proposed another process was at play in the concept of a PAT loan. Understanding this contact scenario as a PAT loan has the advantage of explaining why the grammatical elements remained Hebrew even as Aramaic influence is evident as well as elucidating the sociolinguistic background of Aramaic as a lingua franca.

4 Causal Subordination, PAT Loans, and a Literary Problem in the Pentateuch

The history of the two adverbializing patterns (1) PREP+S and (2) PREP+REL+S in Hebrew and the contact-induced change in causal subordination from

26 Pat El 2012: 258.27 Matras and Sakel 2007: 1–13 (especially 4–13).28 While Aramaic was certainly a lingua franca in the Assyrian heartland in the eighth

century BCE (and well attested in the seventh century BCE), it does not follow that it functioned similarly in the southern Levant in Judah at this time as well. For arguments concerning Assyrian, not Aramaic, influence in the speech of the Rav Shaqeh, see Oren 2013: 7–10. Moreover, the earliest epigraphic evidence of Aramaic being used administra-tively in Judah is the Aḥiab seal, dated to the late sixth century BCE (the early Persian/Achaemenid era) (Vanderhooft 2011: 538).



Aramaic is signifĳicant for more than purely linguistic reasons. Identifying patterns in BH that show evidence of internal development or Aramaic influ-ence can also have benefĳits for the interpretation of the Pentateuch, particu-larly from a documentary perspective. In this section, a pattern that matches the second adverbializer construction and that may also show evidence of contact-induced change from Aramaic is presented. Its appearance in Exod 19:18 is analyzed, revealing that the adverbializer and the clause it governs can-not be easily aligned to any of the traditional sources. Finally, we will offfer a suggestion concerning the nature of the causally subordinated clause in Exod 19:18.

The second class of adverbializers, PREP+REL+S, as has been shown above, has a long history in Hebrew. As also discussed above, Aramaic contact-induced change in causal subordination in BH manifests itself in LBH. This causal subordination in Aramaic can have a similar pattern as BH PREP+REL+S. In Aramaic, the relative particle /d/, according to Wertheimer, nominalizes the following clause, in efffect translating the clause into a nominal constituent.29 The addition of the preposition in front of the relative particle then adverbial-izes the relative phrase, which can denote any number of adverbial expres-sions. In Syriac, two constructions in this pattern are used to form adverbial expressions: ʿl(PREP)+d(REL)+S; and mṭl(PREP)+d(REL)+S.30 This particular pattern, being common in Aramaic and BH, cannot be ascribed to language contact. Moreover, the construction ʿl(PREP)+ REL +S appears in both BH (in a wide variety of literary strata) and Aramaic, and also cannot be attributed to contact-induced change. The shift toward standardization in later Hebrew may well have been influenced by the Aramaic pattern and certainly the even-tual shift to the mono-syllabic šɛ- evinces a likely PAT borrowing situation.

The occurrence of the adverbial marker mpny ʾšr (marking reason) in Exod 19:18 and Jer 44:23 is peculiar in BH. This fact alone is not sufffĳicient to pro-pose language contact. Given the literary problems in both verses, however, in which the adverbial clauses are likely later additions to each text, the unique-ness of the construction mpny ʾ šr in BH merits further examination. The excep-tional nature of Jer 44:23 has been discussed in detail. The content of the verse governed by mpny ʾšr is an explanation of Jer 44:22, and Jer 44:23 is peculiar inasmuch as it is the only instance in BH in which the causal subordinator mpny begins a sentence in which the verb is a sufffĳix-conjugation (SC).31 This

29 Wertheimer 2001: 289.30 Wertheimer 2001: 284.31 Goldstein states (2013: 100): ’בלשון המקרא איננו מוצאים בדרך כלל שימוש במילית ’מפני

בפתיחת משפט משועבר. יתר על כן, בלשון המקרא נפתחת פסוקית סיבה על ידי ”יען״ או

46 Boyd and Hardy


construction is unique in BH syntax, and could be explained as a later gloss on the meaning of Jer 44:22. In Aramaic, the adverbializer b-dyl d- expresses a causal meaning when governing a SC and a resultative meaning when govern-ing a PC. As Pat-El observes, the constructions b-ʾšr l- in Jon 1:8 and b-šl ʾšr in Qoh 8:17 are loans from the Aramaic causal subordinator b-dyl d-; however, in these verses in the Hebrew Bible the causal subordinators govern prefĳix-conjugation (PC) verbs, the reverse of the usage in Aramaic.32 Pat-El correctly notes that the RL does not have to have the “same distribution” of borrowed features as the SL. If, however, the adverbializer mpny ʾšr is patterned from a similar construction in Aramaic (see below), then the oddity that Jer 44:23 is the only sentence in the Hebrew Bible that begins with a causal conjunc-tion and governs a clause with the SC could fĳind an explanation as a contact-induced change.

Even more peculiar syntactically, then, is Exod 19:18, in which the adver-bial construction mpny ʾšr functions within the sentence as an explanation for the reason why Sinai was wrapped in smoke. The adverbializer mpny ʾšr marks an explanation of the fĳirst part of the verse, which has been identifĳied as the J source; however, in J the deity does not descend on the mountain until Exod 19:20, which contradicts the content of the clause governed by mpny ʾšr.33 In the E source, however, the deity states that he will imminently descend in a cloud (Exod 19:9a), as per Exod 19:16–17. As such, the content of the mpny ʾšr clause matches the narrative of E, and explains why, according to Exod 19:17, the people stand at the foot of the mountain (since the deity had descended on top of the mountain).34 Such an explanation in Exod 19:18, however, is hardly

”יען כי”, והצירוף ”מפני אשר”, כפותח משפט סיבה מופיע רק במקום אחר נוסף במקרא,במ’ יט, יח.

32 Pat-El 2012: 258.33 For a source discussion of J and E in Exod 19, see Baden 2012: 77 (J), 117–18 (E). See also page

128, in which he observes the fact that Deut 4:11 reflects only the narrative of Exod 19:16–17, the only E material in the otherwise exclusively J block of Exod 19:10–18, is consistent with how D uses E and J separately. Baden here builds on the work of August Dillmann, who states, “Aus dem Dt. kann ein Beweis für die Existenz eines BC vor dem ABC auch nicht gezogen werden. Denn zwar ist wohl richtig, dass D in seiner Geschichtsdarstellung sich an B u. C anschliesst; aber schon S. 609f. ist gezeigt, dass D den B u. C noch als besondere Schriften vor sich hatte, während ein sicherer Beweis dafür, dass er BC als Ganzes benüt-zte, nicht geführt werden kann” (Dillmann 1886, 679; for Dillmann’s exegesis leading to this conclusion, see 609–11). For more of a discussion of Dillmann in recent theories of the development of the Pentateuch, see Baden 2009: 37–40. For a more extended analysis of D’s use of E and J, see Baden 2011: 327–44.

34 The trembling of the people within the camp in Exod 19:16b is connected with the trembling of the mountain in Exod 19:18b. An argument could be made that Exod



necessary in E as the narrative connection between Exod 19:9a and Exod 19:16 already makes it clear that the deity has come down on top of Sinai.35 From a source critical perspective, the clause governed by mpny ʾšr in Exod 19:18 belongs neither to the J nor to the E source and can be plausibly identifĳied as a later harmonistic insertion. Moreover, the marking of the phrase in Exod 19:18 by mpny ʾšr is unparalleled as a causal subordinator in BH, excepting Jer 44:23, prompting further exploration as a basis for explanation of the adverbial mark-ing and the literary nature of the verses in their context.

Two explanations can be offfered for the origin of this construction in BH. First, given the similarities between mpny ʾšr and later mpny š in Tannaitic Hebrew, one could argue that it is a back-formation from this idiom into BH. The scribe who made the insertion could have recognized the dialectal distinc-tion and translated the later Hebrew form into the biblical form with ʾšr for the insertion. The lack of other clear examples of this process, however, may argue against this solution.

Second, the presence of mpny ʾšr in Exod 19:18 and Jer 44:23 could be explained as a result of a PAT loan from Aramaic. Wertheimer has explored a linguistic process in Syriac in which a relative pronoun substantivizes the linguistic elements it governs (a process called “translation,” borrowed from Tesnière).36 A further (or, second order) translation of PREP+REL converts the construction into an adverbial clause in a complex sentence, forming a subor-dinating conjunction marking reason and other adverbial relationships. For

19:18b should be changed to ‘all the people trembled very much’, following the reading in the Septuagint. A similar confusion between people and the mountain occurs in the Samaritan Pentateuch in Exod 19:12a. In this understanding of Exod 19:18b, the people are increasingly in awe and tremble as the theophany happens before their eyes, which explains their response in Exod 20:18–20. On the other hand, reading with the MT allows for another literary explanation. The people tremble in the camp in Exod 19:16b, and after Moses brings them out of the camp for the encounter with the deity he then stations them at the foot of the mountain. To communicate the sheer immensity of the divine encounter, the narrator claims that the mountain, the location of the theophany, had an even more severe response than the people, trembling very much.

35 An argument might be made that a preposition governing a SC can locate the event in the non-simple past, envisioning a completed event in the future leading to the harmon-istic translation in Exod 19:18 ‘because God will descend. . .’. In this understanding, Exod 19:18 could belong to J. For example, Exod 10:3 begins with the temporal marker ʿd mty, which sets the SC mʾnt as a completed future event, ‘How long will you refuse . . .?’ (Waltke and O’Connor 1990: §30.4.c). Against this understanding of Exod 19:18, this use of SC is marked with temporal adverbs, not causal. The SC in Exod 19:18 is therefore better trans-lated ‘because God had descended’.

36 Wertheimer 2001: 261–63.

48 Boyd and Hardy


example, the NP ‘the words of the king’ (or, ‘the king’s words’) can be converted into an adverbial phrase with the addition of mṭl as in ܐ mṭl mly

mlkʾ ‘because of the words of the king’. The PREP+REL forms a subordinate conjunction, “marking the beginning of a clause in a complex sentence”:

ܐ ܕ mṭl d-mlkʾ mll ‘because the king spoke’.37 In this case, the PREP+REL (mṭl d-) functions as an adverbial conjunction denoting reason.

The relative in this pattern PREP+REL is a translatif, or a morphological marking of the translation or transfer of a word from one grammatical cat-egory to another, according to Tesnière’s framework. In complex sentences, a double translation occurs in which the REL marks a fĳirst stage in which a ver-bal phrase is translated to a nominal element in the sentence and the PREP turns the new nominal element into an adverb. Wertheimer’s overall argument is to show that the REL in Aramaic in all its forms is a nominalizing particle. Regarding BH, the REL ʾšr nominalizes the following verbal clause, and the use of mpny turns the entire phrase into an adverbial element modifying the fĳirst part of Exod 19:18. One could argue that mpny ʾšr in Exod 19:18 is patterning after this construction in Aramaic given the literary and linguistic issues that mark the clause governed by mpny ʾšr as a later, post-redactional addition and given the time period (Persian era when Aramaic was a lingua franca) that such an addition would occur.

Against this explanation is the fact that Wertheimer’s study was based on Syriac, at a much later period than the biblical text, and the types of scribal additions proposed for Exod 19:18. Moreover, an internal explanation within Hebrew may lie in the possibility that a similar process may be behind the development of the preposition ʿl ‘upon’, which can be causal on its own,38 but also governs the REL ʿl ʾšr meaning ‘because’.39 In a similar manner, Hebrew mpny can have a causal meaning on its own,40 and mpny ʾšr could be under-stood as a more explicit marking of an adverbial translation, converting the

37 Wertheimer 2001: 284.38 Ps 119:13639 Num 20:2440 It should be noted that to the best of our knowledge the causal use of mpny by itself never

governs a verb. A further subordinator is added when a verbal element is needed, as in Jer 51:64: מפני הרעה אשר אנכי מביא עליה mpny hrʿh ʾšr ʾnky mbyʾ ʿlyh ‘because of the evil which I am bringing upon her’, where ʾšr functions as a further subordinator that governs a verbal element. We are aware of the following causal uses of mpny by itself: Gen 6:13; 41:31; Exod 3:7; 8:20; Deut 28:20; Judg 6:6; 1Kgs 8:11; Job 37:19; Is 7:2; Jer 9:6; 23:10; 51:64; Ezek 14:15; Lam 5:10; Amos 2:9; Hos 10:15; Mic 1:4; Mal 3:14. None of these cases consists of mpny governing a verb. In this manner, the use of ʾšr in mpny ʾšr may allow for such verbal governance.



following verbal phrase into a substantive element by use of the REL and then transforming the whole construction into an adverb by use of the PREP.41

In this section, a peculiar adverbial construction in Exod 19:18 and Jer 44:23 was examined. The clauses governed by these adverbializers are literarily prob-lematic and in the case of Exod 19:18 the clause has been suggested to be a later scribal harmonization between two sources. Given the time period for such scribal additions (Persian era or later when Aramaic pattern loans occur else-where in BH), Aramaic influence is one possible explanation for this unique construction. Since an internal suggestion can tentatively also be made, how-ever, the suggestion that mpny ʾšr is a pattern loan is an admittedly hypotheti-cal, albeit intriguing, possibility to a problematic construction that deserves further thought.

Bibliography

Baden, Joel. 2009. J, E, and the Redaction of the Pentateuch (Forschungen zum Alten Testament 68). Tübingen.

———. 2011. “Deuteronomic Evidence for the Documentary Theory,” in Thomas B. Dozeman, Konrad Schmid, and Baruch J. Schwartz (eds.), The Pentateuch:

International Perspectives on Current Research (Forschungen zum Alten Testament 78). Tübingen. 327–44.

———. 2012. The Composition of the Pentateuch: Renewing the Documentary Hypothesis (Anchor Bible Reference Library). New Haven.

Blau, Joshua (ed. by Michael Patrick O’Connor and Cynthia L. Miller). 2010. Phonology

and Morphology of Biblical Hebrew. An Introduction (LSAWS 2). Winona Lake.Butts, A. M. Forthcoming. Language Change in the Wake of Empire: Syriac in its Greco-

Roman Context (LSAWS 11). Winona Lake.Coetsem, Frans van. 1990. “Review: Language Contact, Creolization, and Genetic

Linguistics by Sarah Thomason; Terrence Kaufman; Lectures on Language Contact

41 The preposition mn plus ʾšr is typically non-causal in BH. The following are instances in which a non-causal use of mn plus ʾšr is evident: Gen 31:1; Exod 5:11; 29:27 (x2); Lev 14:30; Josh 10:11; Judg 16:30; Ruth 2:9; 2Kgs 6:16; Esth 4:11; Eccl 3:22; Is 47:13; Jer 40:7. The causal use occurs twice for this construction: Num 6:11; Is 43:4. The second is clearly exilic or post-exilic, and the Septuagint did not render the construction in Num 6:11 as causal, but rather ‘concerning the things (in which) he sinned’. The Aramaic construction mn dy is clearly causal in Dan 3:22. The form mzy in Sefĳire 1:25 is adverbial, representing mn and zy, and is likely temporal; a causal translation (‘because Ashur reigns!’) is theoretically possible, but does not make sense in the context of the curse formulae.

50 Boyd and Hardy


by Ilse Lehiste; Languages in Competition, Dominance, Diversity, and Decline by Ronald Wardhaugh,” Language in Society 19: 260–62.

Dillmann, August. 1886. Die Bücher Numeri, Deuteronomium, und Joshua (Kurzgefasstes exegetisches Handbuch zum Alten Testament 13). Leipzig.

Ferná ndez, Miguel P (trans. John Elwolde). 1999. An Introductory Grammar of Rabbinic

Hebrew. Leiden.Gensler, Orin D. 2000. “Why Semitic Adverbializers (Akkadian -iš, Syriac -āʾīṯ ) should

not be Derived from Existential *ʾīṯ ,” JSS 45: 233–265.Gil, David. 2013. “Genitives, Adjectives and Relative Clauses,” in Matthew S. Dryer and

Martin Haspelmath (eds.), The World Atlas of Language Structures Online. Max Planck Institute for Evolutionary Anthropology. (Available online at http://wals.info/chapter/60, accessed on 2014–08–26.)

Goldenberg, Gideon. 1998. “Attribution in Semitic Languages,” in his Studies in Semitic

Linguistics. Jerusalem. 1–20.Goldstein, Ronnie. 2013. The Life of Jeremiah: Traditions about the Prophet and their

Evolution in Biblical Times. Jerusalem. (Hebrew)Hardy, Humphrey. 2014. Diachronic Change in Biblical Hebrew Prepositions: A Case

Study in Grammaticalization. Ph.D. Diss., The University of Chicago.Heine, Bernd and Tania Kuteva. 2005. Language Contact and Grammatical Change

(Cambridge Approaches to Language Contact). Cambridge.Holmstedt, Robert. 2001. “Headlessness and Extraposition: Another Look at the Syntax

of אשר,” Journal of Northwest Semitic Languages 27: 1–16.———. 2006. “The Story of Ancient Hebrew ʾăšer,” Ancient Near Eastern Studies 43:

7–26.———. 2007. “The Etymologies of Hebrew ʾă šer and šeC-,” JNES 66: 177–191.Huehnergard, John. 2006. “On the Etymology of the Hebrew Relative šɛ-,” in S. E.

Fassberg and A. Hurvitz (eds.), Biblical Hebrew in Its Norwest Semitic Setting.

Typological and Historical Perspectives. Jerusalem – Winona Lake. 103–125.Matras, Yaron, and Jeanette Sakel. 2007. Grammatical Borrowing in Cross-linguistic

Perspective (Empirical Approaches to Language Typology 38). New York.Oren, Mikhal. 2013. “Interference in Ancient Languages as Evidenced by Governed

Prepositions,” JSS 58: 1–11.Pat-El, Na’ama. 2008. “Historical Syntax of Aramaic: A Note on Subordination,” in

H. Gzella and M. L. Folmer (eds.), Aramaic in Its Historical and Linguistic Setting. Wiesbaden. 55–76.

———. 2012. “Syntactic Aramaisms as a Tool for the Internal Chronology of Biblical Hebrew,” in Cynthia L. Miller-Naudé and Ziony Zevit (eds.), Diachrony in Biblical

Hebrew (LSAWS 8). Winona Lake. 245–63.Qimron, Elisha. 1986. The Hebrew of the Dead Sea Scrolls (HSS 29). Winona Lake.



Segal, M. H. 1927. A Grammar of Mishnaic Hebrew. Oxford.Thomason, Sarah. 2001. Language Contact: An Introduction. Washington, D.C.———. 2003. “Contact as a Source for Language Change,” in Brian D. Joseph and

Richard D. Janda (eds.), The Handbook of Historical Linguistics. Malden, MA. 687–712.

Thomason, Sarah and Terence Kaufman. 1988. Language Contact, Creolization, and

Genetic Linguistics. Berkeley.Thompson, Sandra and Robert E. Longacre. 1985. “Adverbial Clauses,” in Timothy

Shopen (ed.), Language Typology and Syntactic Description. Cambridge. 171–234.Vanderhooft, David. 2011. “ eʾl-mĕdînâ ûmĕdînâ kiktābāh: Scribes and Scripts in Yehud

and in Achaemenid Transeuphratene,” in Oded Lipschits, Gary N. Knoppers, and Manfred Oeming (eds.), Judah and the Judeans in the Achaemenid Period: Negotiating

Identity in an International Context. Winona Lake. 529–44.Waltke, B. K. and M. O’Connor. 1990. An Introduction to Biblical Hebrew Syntax. Winona

Lake.Wertheimer, Ada. 2001. “The Functions of the Syriac Particle d-,” Le Muséon 114:

259–89.Winford, Donald. 2003. An Introduction to Contact Linguistics (Language in Society 33).

Malden, MA.———. 2010. “Contact and Borrowing,” in Raymond Hickey (ed.), The Handbook of

Language Contact (Blackwell Handbooks in Linguistics). Malden, MA. 170–187.

Hebrew Adverbialization, Aramaic Language Contact, and mpny ʾšr in Exodus 19:18

Documents

Transcript of Hebrew Adverbialization, Aramaic Language Contact, and mpny ʾšr in Exodus 19:18