Hebrew Linguistics and Biblical Criticism: A Minimalist Programme
Vincent De Caën
Near and Middle Eastern Civilizations, University of Toronto

Abstract

This paper introduces a programme in historical linguistics with important implications for biblical criticism. Traditionally, grammatical variation in the Bible has been interpreted in light of nineteenth-century historical-literary criticism. In this light, such variation appears erratic and random. To date we have developed a somewhat vague distinction between “early” and “late Biblical Hebrew” (EBH vs LBH). The programme outlined here proposes to let the Hebrew language speak for itself, to let natural diachronic processes explain the distributions independent of the literary paradigm. The results should suggest a new alignment of texts and sources. The paper has two parts. The first, polemical part situates the programme within recent, indeed controversial, departures in biblical studies. The second part works through a problem that has hitherto resisted explanation to showcase the methodology and to indicate the anticipated results. As a first approximation a fivefold stratification is proposed, considerably refining the traditional taxon “early Biblical Hebrew” (EBH). The most interesting conclusion is the priority of Deuteronomy within the five books of Moses. Another result is the sorting of composite books like Psalms by linguistic criteria. The programme is expected to yield a three-volume study: morphology, syntax, lexicon (in that order).

We can be bound up in the “scientific method.” … However, the principles of logic and research that may be applicable to the study of science, or even to the humanities, often are not applicable in the biblical sphere. But this is hard to accept … .

—A. B. & A. M. Mickelsen, Understanding Scripture (1992) 9

If either the historicity of the biblical construct or the actual date of composition of its literature were verified independently of each other, the circle could be broken. But since the methodological need for this procedure is overlooked, the circularity has continued to characterize an entire discipline [biblical studies]—and render it invalid.

P. R. Davies, In Search of “Ancient Israel” (1995) 37

Typically, when questions are sharply formulated, it is learned that even elementary phenomena had escaped notice, and that intuitive accounts that seemed simple and persuasive are entirely inadequate.

N. Chomsky, The Minimalist Program (1995) 4

The sources for these different synchronic states would have to be distinguished. Synchrony would be achieved by separating out perhaps ten or a dozen synchronic states within the corpus of biblical texts.

J. Barr, “The Synchronic, the Diachronic and the Historical” (1995) 3

1. Invitation to a Minimalist Programme

1.1 This paper briefly introduces the relatively new paradigm that has been emerging in biblical studies, especially in the last decade or two—no doubt the paradigm for my generation.1 As Philip Davies asserts, “Anyone familiar with the range of current scholarship will know that the gap between the biblical Israel and the history of Palestine is widening, and that new scholarly constructs are in the process of emerging.” Further, “a search for the nature and the source of the biblical Israel might provide a valuable working agenda for the next generation of historical biblical research” (Davies 1995, 46; italics mine in both cases).

1.2 This agenda for my generation I am here renaming the Minimalist Programme for biblical studies, to which I intend to contribute the missing keystone of Hebrew linguistics. I will invite you to wander with me in that forbidding Wilderness of Hebrew Grammar, but only for a few short minutes—not years; and I promise to bring you within sight of the Judean highlands. I will invite you, to switch metaphors, to enter the margins of Hebrew grammars in search of oppositions and paradoxes that, when properly understood, subvert and deconstruct “Biblical Hebrew”. I will challenge you, for the very first time, to the let the Hebrew language speak for itself from the periphery of biblical studies, for it has a great deal to say.

1.3 Accordingly, the paper is divided into two parts. The shorter first part is necessarily polemical in nature: a Minimalist Manifesto, as it were. The extended second part, because of time constraints, can only work through a simple but nevertheless telling and representative diachronic problem to give a concrete sense of my actual investigation and its expected results. This is supplemented by a further example of the counterintuitive behaviour of Hebrew, suggested by Schniedewind (1999).2

PART I

2. Defining “Minimalism”

2.1 “Minimalism” is a state of mind. Minimalism is, in the first instance, an aesthetic (and so moral) imperative. Minimalism is the sublime austerity of theoretical economy, the quest for an elegance and simplicity—the very inspiration of metaphysics. Minimalism is the most radical rethinking of foundations (from the apt title to Van SetersFestschrift edited by McKenzie et al. 2000), straining at the limits to the point of intellectual vertigo.

2.2 Minimalism also just happens to be Noam Chomsky’s latest programme in linguistics (Chomsky 1995; see further Epstein & Horstein 1999). This Minimalist Programme is his radical re-examination of foundations and assumptions, the salvaging and synthesizing of the successes of a half-century of generative grammar, obeying the imperative of theoretical economy and pursuing the empirical consequences.

2.3 Alas! Minimalism in this narrow sense is also a term of abuse from reactionary Hebraists, perceiving in it the latest salvo of the dreaded “sentence-grammar”3 from the Great Satan himself and his minions in the academies, and also correctly perceiving in it the underlying threat from interdisciplinary cognitive-scientific materialism.

2.4 Minimalism is also a term of abuse in biblical studies generally, signifying the Enlightenment horrors of hypercriticism, cynicism and nihilism: this way be dragons (Davies 1995, 25). Minimalism, in this broader context, is a fighting word. So be it. As any good iconoclast would, I will adopt this term of abuse for my own overall programme, explicitly playing on this double entendre. The opprobrium is reasonably expected to be of short duration.

2.5 Minimalism will henceforth be a declaration of independence. Some archaeologists (e.g., Finkelstein & Silbermann 2001) and historians (e.g., Thompson 1994, Davies 1995, Lemche 1998; cf. Jenkins 1991) are already declaring their independence, reclaiming autonomy for their own biblical disciplines despite the vicious and vituperative howls from the religious conservatives. They are reasserting the scientific can(n)ons that define their disciplines, and following these dictates wherever they might lead.

2.6 Similarly, I will make to boldly declare the independence of linguistics and literary studies, and to follow the dictates of my specialization wherever they might lead.

2.7 “Maximalism” would be the appropriate, broad designation in this context for that modern paradigm which we are leaving behind. Maximalism is that exhausted and exhausting 19th-century exegetical discourse, primarily German and so Idealist and Romantic, primarily Protestant but in any case fundamentally theological—a subdiscipline of theology, as it were (cf. Oden 1987, ch. 1 and epilogue)—to be more specific, we might say specifically Pauline, following Jacob Neusner (2001, 3 et passim): a discourse that has hitherto dominated the study of the Bible.

2.8 Maximalism is mistaking an essentially literary construct for something that it is not and can never be. Maximalism is the continuation by other means of Christian discourse on the ancient Jew, the supposedly “self-righteous, simple-minded legalist” (McCann 1993, 34), part and parcel of that larger discourse called Orientalism (in the classic formulation of Edward Said)—the occidental study of the Jew and the Arab (the Others): their marginal histories; their inferior religions and cultures; their arrested languages; indeed, even their arrested, non-Aryan minds.

2.9 By definition, then, Biblical Minimalism must be a postmodern paradigm, still in the initial stages of its formation, still accumulating its critical mass (though founded a quarter of a century ago (see further McKenzie et al. 2000). Consequently, in its self-defining, polemical aspect, Biblical Minimalism entails the spade-work of Foucault’s archaeology, genealogy and problematization (see, e.g., Gutting 1994): an unflattering history of our field that we would rather not write (and so has hitherto not been written!). Further, and crucially in my opinion, it entails Derrida’s deconstruction of the dyads early/late, pre-exilic/post-exilic, Moses/Ezra, Israelite/Jew (on Derrida see, e.g., Howells 1999).

2.10 In summary, then, Biblical Minimalism is fundamentally about the right and obligation to pursue a methodologically sound, crucially non-theological, investigation of late Iron Age Palestine and the Persian province of Yehud, independently (at least in the first instance) of that derivative literary construct we call ancient Israel: be the approach archaeological, anthropological, historical, linguistic, etc.

3. The Problem with Maximalist Philology

3.1 I will show in the second part how to reassert the autonomy of linguistics in biblical studies, and in so doing, reassert the priority of Hebrew linguistics as the queen of the biblical disciplines. By walking you through one very simple example, I will indicate the sort of results that can be anticipated from this research programme. It will become clear that Hebrew linguistics provides the crucial, empirical grounding that the new programme requires, thereby breaking the circularity identified by Davies in the second motto above, and justifying the initial epithet missing keystone.

3.2 First, we must reassert that the Hebrew language is inherently worth studying qua natural language, a function of Universal Grammar.4 Maximalist Hebrew philology as the handmaiden of theological exegesis needs, therefore, to be radically transformed into theoretical Hebrew linguistics, employing the latest in theory and method, and obeying the canons of description and argumentation.

3.3 Second, we must insist on the methodological principle of internal reconstruction preceding comparative grammar. Hebrew is not Arabic with a bad accent. Hebrew is not some teratogenic Mischmaschsprache, combining east and west in a typologically implausible fashion. Further, the Tiberian reading tradition is certainly not a linguistic fall from grace, but rather the expression of many generations of the finest phonological description anywhere.

3.4 Third, we must take linguistic variation seriously, and not as a distraction—or noise in statistical lingo. It cannot be emphasized enough that there is no such thing as biblical Hebrew, unless we mean by that taxon, somewhat trivially, the family of dialects attested between the covers of the Bible.5 Any rigorous attempt to write a generative (and so crucially synchronic) grammar of soi-disant biblical Hebrew will founder on linguistic variation: morphological, syntactic, semantic, discourse-analytic, &c. (see the helpful discussion by Barr 1995; cf. Diest 1995). It might be claimed that God does not change (though he certainly changes his mind often enough!), but the language employed to encode his oracles certainly does change, as any natural (read human) language must.

For a great many Christian scholars, Hebrew was a one-dimensional language in which Scripture was written. Because Hebrew was God’s language and God never changes, it was easy and perhaps logical to assume that Hebrew had a uniform character and personality and was not subject to either internal or historical development and change (Jerome Friedman, The Most Ancient Testimony (1983), cited by Rooker 1990, 26).

4. The Maximalist Blindspot: Diachronics

4.1 It is quite striking, perhaps shocking, that after two centuries of intense modern study of Hebrew that we can show nothing better than a lame distinction between early and late biblical Hebrew (EBH vs LBH; for a concise history and a sense of the state of the art, see Rooker 1990). To the extent that variation is even noticed, it is not considered tractable—if not considered completely random.

4.2 In retrospect, it is not difficult to understand how this perspective could arise: after generations of Traditionsgeschichte und Religionsgeschichte und Redaktionsgeschichte und Soweiteronsgeschichte (for quick overviews, see Hayes & Holladay 1982, Rast 1972), the absolute dating of texts and sources is already chiselled on stone tablets. But if we have failed to properly distinguish materials, and if we have the relative dating of materials incorrect, then naturally variation will appear intractable or indeed random.

4.3 Of course, any language change would be in the direction of that putatively degenerate idiom that characterizes the admittedly latest books of the Bible, those decidedly Second Temple compositions lacking in christological interest, those decidedly insipid, Jewish texts lacking the thundering, universalist voice of the major prophets. But I claim that this degenerate Jewish idiom, this linguistic barbarism of Ecclesiastes and Qumran on the slippery slope to the vulgar idiom of the Sages and Rabbis, is in fact the key that unlocks the door to a rigorous historical dialectology. In short, we have been looking in all the wrong places: let us now look in the right place.

PART II

5. Methodology: Reverse Engineering Qumran Hebrew (QH)

5.1 I make the idealizing assumption, as a null hypothesis, that there is a continous development in the prestige, literary dialect of Jerusalem from the tailend of the late Iron Age through to the destruction of the Second Temple (an assumption that has been justified elsewhere, as well as by the results obtained here, pace Schniedewind 1999). (N. B. The onus would fall on those arguing for discontinous development.)

5.2 Further, as a heuristic, some concrete sense of directionality and teleology would be helpful. Methodologically, then, I will look to the terminus ad quem to bootstrap my investigations.6 Thankfully, Elisha Qimron has already done a superb job of description in his The Hebrew of the Dead Sea Scrolls (1986). There are, fortunately, enough puzzles in the margins of these pages alone to sustain several careers.

5.3 There is one puzzle that particularly tickles my fancy, because it has never been identified explicitly as a puzzle (which may speak volumes about Hebrew grammatical investigations),7 and because it is initially such a teaser, requiring the computational and statistical artillery of corpus linguistics (see below §10) to resolve it. The puzzle that I have selected is a verbal form, otherwise a quite marginal verbal form, that is apparently impervious to diachronic processes against all expectations, as I explain in the next section.

6. QH Sequential Forms of the Verb

6.1 My own doctoral studies tentatively sketched a generative grammar of “standard Biblical Hebrew”, focussing on the morphology, syntax and semantics of the verbal system—which is the heart of any such formal grammar (DeCaen 1995). I have since been obsessed with the initially bizarre sequential or consecutive phenomenon which dominates the ancient Hebrew system (statistically, at least).

6.2 In the past few years I have been able to identify and analyze the phenemon crosslinguistically (in unpublished work), and now have a more or less complete account of the generative syntax and semantics of modal coordination for Universal Grammar. And yet, I have been stymied by the increasingly obvious diachronic variation in the biblical phenomenon itself.

6.3 Starting from the perspective of QH we can easily identify the two diachronic processes at work on the morphology of the principal sequential form—the so-called wayyiqtol: first, the process of apocopation applies to forms derived from glide-final roots, as shown in (1); second, a reanalysis of the modal, so-called paragogic heh (Shulman 1996) as the single, general ending for the first-person modal forms, shown in (2).

6.4 Both processes combined have the effect of collapsing the distinction between modal coordination and modal forms in general, hence the very natural diachronic explanation of levelling by analogy (Qimron 1986, §310.122; cf. Joüon 1996, §48d; Waltke & O'Connor 1990, §33.1.1b).

(1) Apocopation (or, loss of stem-final vowel)

[e]# > Ø

ויבכה‎ [wayyivke] > ויבך‎ [wayyevk]

(2) Suffixation (or, reanalysis of first-person modal paradigm)

Ø > [a]#

ואכתב‎ [wa′extov] > ואתבה‎ [wa′extóva]/[wa′extvá]

6.5 There is really no room for controversy here. However, curiously enough, the processes apparently make conflicting demands of the first-person forms (both to drop stem-final vowel and to add the vowel of the general suffix). Even worse, the actual form consistently found in QH is that shown in (3).

(3) Attested final-weak first-person forms (10x: Qimron 1986, §310.129(2c))

*! ואבכה‎ [wa′evke]

6.6 In (3), we observe a form that apparently is impervious to both processes, in defiance of all common sense. Both apocopation and suffixation have failed to apply. Curiouser and curiouser.

7. Maximalist Analysis (Such as it is)

7.1 The general phenomenon registers, of course, in the standard references, but as a nuisance to be ignored—or at best, as an intrusion from the later, degenerate idiom: certainly not worth further investigation. Here is an opportunity, therefore, to begin a site survey for an “archaeology” of Hebrew grammatical analysis.

7.2 The current standard is Muraoka’s translated revision of Paul Joüon’s 1923 grammar (Joüon 1996). Notes on the anomalous paragogic heh can be found at §§45, 48d, 114 and 116b. However, there is a very telling note buried in §118v8 in which we are asked to consider two examples: first, from Genesis 41:11; then, from Ezra 7:28. Why is this telling? Because Hebraists can decode the note as follows. The first example is located in the so-called J source, traditionally considered the earliest composition dating from the tenth or even eleventh century bce (give or take). The second example is located in the latest stratum of the Chronicles-Ezra-Nehemiah complex. Given such a distribution, therefore, the phenomenon can be of no diachronic import: quod est demonstrandum.

7.3 That we have the correct decryption of that note is confirmed on the second stone tablet of Hebrew grammar (Waltke & O'Connor 1990). The distribution of first-person variants is irredeemably “erratic” (§34.5.3b). How so? The anomaly “also occurs in some pre-exilic texts but not in some post-exilic books … “(§34.5.3b). The Maximalist logic here is that we know, in advance, beyond a reasonable shadow of doubt, which material is late and which early. Additionally, the embarrassing argument ex silentio might be overlooked, were it not found repeatedly. “This is not necessarily a late feature of the language: it is not used by the post-exilic writer of Chronicles, but it is used by the earlier Deuteronomist … . Moreover, it is not found in Zechariah 1–8 or Esther, both of which are post-exilic” (§33.1.1c).

7.4 Cadit quaestio, to the extent there is an argument here at all. In passing, we might in fairness observe that Waltke & O'Connor are cribbing Robert Polzin’s Harvard dissertation (1976, 54–55), the only substantive work in diachronics to date, and that the fallacies are originally articulated there. Also in passing, it should be noticed that the same logic is at work in Hurvitz’s claim of priority for the Priestly Source (1982), and in the truly bizarre claim that Qoheleth can be contemporary with the reign of Solomon (Fredericks 1988).

8. The Solution with Biblical Confirmation

8.1 From a purely linguistic point of view, independent of the centuries of biblical criticism, the solution to this chestnut is childishly simple. The first step involves setting aside the canonical, medieval reading of Hebrew (a step that may be disorienting for some, if not radically unacceptable: so be it).

8.2 The second step, reading the consonantal text alone, is to recognize the common, orthographic h found in both diachronic rules. In this perspective, we find that the orthographic h is inherently ambiguous as to its value (either [e] or [a]).

8.3 And the third and crucial step is to invoke a cardinal principle from generative grammar: rule ordering. By resorting to this fundamental principle of linguistics, we are able to resolve this apparently paradoxical “blocking” of rule applications by the diachronic sequencing in (4).

(4) (a) (b)
Regular Root Glide-Final Root
Input ואכתב ואבכה
Apocopation ואבך
Suffixation (Reanalysis) ואכתבה ואבכה
_______________________
QH Output ואכתבה ואבכה

8.4 The diagram in (4) is read as follows. Before any rules apply, we have the first-person sequentials showing no inflectional endings; the final h of the glide-final form (column 4b) is a mater lectionis (vowel letter) indicating the stem-final [e]. The first rule to apply must be that of apocopation, the cutting away of the final vowels: this rule applies across the board to all such forms, but crucially catching the first-person sequential of glide-final roots (4b). The second rule of suffixation applies to all first-person forms, indicated by the orthographic h (both columns 4a and 4b). Note that this h now represents a final [a], not [e]. The elegance and simplicity of this account is that both rules apply without exceptions; the trick is that they apply serially.

8.5 We thus arrive, driven by the inherent linguistic logic, at a prima facie bizarre prediction: the diachronic distribution of the final h must show an apparent regression to norm. First, we should see the h of glide-final roots disappearing; but then we should see that same final h reappearing (hence regression). While the final h is progressively lost, it would represent stem-final [e]; while the final h is progressively regained, it would represent the suffix [a], not the original [e].

8.6 The bizarre prediction demanded by the inherent logic of the forms is represented graphically in Table §1 (the charts are grouped for convenience in the Appendix of Cross-Tabulations below). On the x-axis, reading from left to right, is represented the process of apocation: at the extreme left is the long, non-apocopated form (100% presence of final h); at the extreme right is the apocopated form (0% presence of final h). On the y-axis, reading top to bottom, is represented the process of suffixation to regular forms, i.e., excluding the glide-final roots: at the extreme top is the uninflected form (0% presence of final h); at the bottom is the inflected form (100% presence of final h).

8.7 The predicted diachronic behaviour is encoded by means of arrows superimposed on the cross-tabulation. First apocopation applies, represented by the arrow moving from left to right across the top of Table §1 (100–0% on the x-axis). Second, suffixation appears after apocopation has applied, represented by the arrow moving from top to bottom of the table (0–100% on the y-axis). The prediction must be that as suffixation applies to regular forms, the final h on glide-final forms must reappear in lock-step. The lock-step movement will also give the counterintuitive appearance of regression. The second arrow, then, represents the regression by moving from the top-right to the bottom-left.

8.8 As a first approximation, we will test the prediction against the biblical distribution of first-person sequential forms in reasonably uncontroversially dated books in Table §2 below. The distribution, it is predicted, should fall along the arrows again superimposed on the cross-tabulation. The books employed are prophetic in the first instance (Amos and First Isaiah, with Jeremiah and Ezekiel), and supplemented by Persian Zechariah and Job and Hellenistic Ecclesiastes. This initial distributional pattern strikingly confirms the prediction. So far, so good.

8.9 Things become even more interesting in a complete cross-tabulation of the books of the Bible set out in Table §3. Granted that using books is somewhat coarse-grained; nevertheless, we must start somewhere.9 The results are promising enough to hypothesize dialect profiles for a secondary, more fine-grained investigation of composite books by pericopes, based on the idealization obtained from Tables §§1–2.

8.10 Having whetted our diagnostic tool, the natural place to apply it is the Pentateuch or Torah (the five books of Moses). Adjusting for absence of forms, indicated in Table §4 by the secondary, horizontal arrows, the distribution offers a coherent picture: indeed, coherent enough to provide a neo-Wellhausian documentary hypothesis. In this case, however, JE crucially follows D, but still precedes P (on pentateuchal criticism and sources, see, e.g., Blenkinsopp 1992, especially chs. 1–2). Just as significant, the anchor points permit a tentative dating of these sources by century.10

8.11 The value of such dialect profiles (x,y: 100–0, 50–0, 0–0, 50–50, 100–100) can be demonstrated in other composite works as well. The very fact that such forms appear in some psalms and not others already suggests a distributional approach to psalm taxonomy, which may prove to be an interesting supplement to Gunkel’s Gattungen. Further, we observe in Table §5 that the Psalms, as a canonical whole (ps*), cluster at one end of the spectrum.

8.12 Similarly, the so-called history books display a heterogeneous patterning amenable to such micro-analysis, as indicated in Table §6. Exceedingly puzzling is the distribution in the books of Samuel, especially the unit containing the so-called Court or Succession Narrative (see, e.g., McCarter 1984, 9–11 et passim), often understood as the historical core. The radical separation of the material known as Nehemiah’s Memoirs from the larger Ezra-Nehemiah complex is also attention-grabbing (see, e.g., Myers 1965, xxxviii-lii et passim); however, such a separation would elegantly solve the distributional paradox articulated by Polzin (1976, 55).

8.13 In summary, if we can keep the literary tail from wagging the linguistic dog, our understanding of the history of late Judean literature should change dramatically; and a fruitful tension and interaction is expected with recent historical-literary criticism. A concrete sense of just how dramatically can be gained by comparing this tentative development sketched here in the appended Tables and summarized in (5) with a similar table in Peckham (1993, figure 1) based on literary argumentation. (In (5), the nomenclature is borrowed from the archaeological periodization: see further Mazar 1992. The symbol * indicates the bulk of the book in question.)

(5) Tentative Sixfold Stratification of Ancient Hebrew

Iron IIC

Assyrian

ca 700

am, is

Babylonian

ca 600

dt*

ez, je

Persian

Persian I

ca 500

gn*, ex*

zc

Persian II

ca 400

nu*, gn 12-18

jb

Graeco-Roman

Hellenistic

ca 300

dt 5, 31-32, ex 8–9

ez-ne*, ec

Hasmonean

ca 150

qumran

9. Questions and Extensions

9.1 Two supplementary questions are raised by the study of first-person sequentials. First, can we expect to approximate Barr’s ten to twelve dialectal strata (Barr 1995, 3)? We should come close. Second, can we find other counterintuitive regressions (apparent regressions!)? Yes. In fact, Schniedewind (1999) founds his proposal for a Qumran “antilanguage” in part on one such case, incorrectly analyzed as suggested below.

9.2 There is a reasonable diachronic pathway through the variations in third-person object suffixes (3ms /-hu/, 3fs /-ha/, 3mp /-hem/, 3fp /-hen/). The consonantal /h/ should assimilate11 or completely drop out over time. An interesting case is the alternation between [-athu] and [-attu] when the 3ms object /hu/ is suffixed to the 3fs past tense ending in inflectional /at/. We would correctly predict the movement [athu] > [attu]. The distribution of such forms is given in (6).

(6) Distribution of [athu] vs [attu]

[-athu]

[-attu]

Genesis (2x, chap. 37)

Judges (3x)

Isaiah (59:16)

Ezekiel (6x)

2 Chronicles (22:11)

Psalm 105:19

Job (5x)

Zechariah (5:4)

Ruth (4:15)

Proverbs 31:1212

1 Samuel 18:28

Proverbs 7:21 and 31:1

1 Samuel 1:24 and 16:14

9.3 Such a distribution should suggest a further refinement for the proposed Persian taxon. Crucially, we can drive a wedge between Genesis and Second Isaiah on the one hand and Zechariah and Job on the other, suggesting that Zechariah represents a transition from Persian I to Persian II. This is of course extremely tentative, based on one distinction, but it is expected that further study would bear out the transitional nature of Zechariah.

9.4 Given the inherent directionality observed in the third-person pronominal suffixes, we should expect that the variation between, e.g., [pihu] “his mouth” and syncopated [piw] should run in the same direction. In other words, tokens of [pihu] should represent a preclassical fossil. The same would be expected of [avihu] “his father”, [axihu] “his brother”, &c. and even when the object suffix is applied to past tense verbs ending in [i] (resulting in [-tihu] in 1cs and 2fs). The distribution is quite complicated, and is the basis of a detailed study (in progress). However, it is clear in taking all such forms together, that as a first approximation the forms in [ihu] cluster in the latest stratum.

9.5 It comes as no surprise, then, to find the [ihu] forms also employed in QH—or does it? Schniedewind (1999) founds his case for a special QH dialect (an “antilanguage”), in large measure, on these putatively unexpected [ihu] forms. He claims that there is no “extant Hebrew dialect to which such anomalies can be consistently traced back”; no basis is given for this claim. From here he argues that the QH “use of such linguistic anomalies seems like an attempt to reconstruct preclassical forms” (1999, 245, italics his).

9.6 Schniedewind’s best case for the supposedly “anomalous character of QH” is precisely this [ihu] phenomenon: a “parade example” (p. 237). Such forms must be “the outcome of ideological manipulation of linguistic form” (p. 238). Why? Such a strong claim is based on the diachronic puzzle (an apparent regression, in our terms): “it remains difficult to explain the reappearance [of the he] in QH” (p. 238).

9.7 It is not difficult to explain the reappearance of the h in QH, when the entire paradigm is considered. In this case, the apparent regression can be explained, again, in terms of rule ordering. In the latest period, the systematically anomalous forms in [iw], the result of an earlier (preclassical) process, viz. the elision or syncopation of /h/, were regularized by analogy with such forms as [uhu], [ohu], [iha], &c., &c. In other words, the very natural diachronic process at work in levelling the first-person modal forms (with the resulting apparent regression), i.e., regularizing and simplifying of paradigms, can be extended to this case as well. We conclude that the same sort of puzzles will lend themselves to the same sort of general principle (regularization of paradigms by analogy).

10. Towards an Exhaustive Corpus-Linguistic Database

10.1 We now have a choice. We can remain mired in the Maximalist rut, with our data intractably “erratic” if not completely random, our comparative studies trailing into a dead end, and inventing antilanguages, apparently a function of our predetermined literary and linguistic history of the Bible. Or, we can adopt the new perspective of corpus linguistics, not so much a theory or even a methodology, but a global, empirical approach that harnasses the power of the computer to crunch the huge biblical corpus and to rapidly execute statistical analyses to identify associative patterns (see further, Biber & al. 1998; and McEnery & Wilson 1996).

10.2 One form or one contrast yields precious little, but all possible variants statistically correlated should yield much. In our simple example, we employed a two-dimensional cross-tabulation. We supplemented this result with two further examples. But the goal is to obtain an n-dimensional cross-tabulation for the entire corpus. I propose to pursue this goal in a three-volume set, increasing progressively in difficulty of execution: morphological, then lexical and finally syntactic. My preliminary, unpublished morphological studies show a rapid convergence on the dialectology sketched above.

11. Conclusion

11.1 The rapid convergence suggested by my historical linguistics is striking enough. The convergence on the same general developments, almost point for point, with the new archaeology as memorably detailed by Israel Finkelstein and Neil Silbermann (2001) is even more striking. Furthermore, the priority of Deuteronomy, e.g., and the relative positioning of the Court History following a 6th-century Yahwist recalls the pioneering yet still marginalized work of John Van Seters (McKenzie et al. 2000). This convergence of critical history, anthropology and archaeology together with historical linguistics is so striking, even at this early stage, to suggest not just a Minimalist Programme but a tentative Minimalist Theory.

11.2 We should not be naïve, however. The battlelines are being drawn all around the world and throughout the universities, wherever what is at issue is the origin of an ancient people, the origin of its religion and culture. In many ways, it is the same question that is being posed over and over again: to what extent should critical scholars be able to pursue their disciplines independent of cultural, religious and literary constructs, whether embodied in tribal myths or collections of texts canonized as “Holy Scripture”?

11.3 Just to give one example closer to home, consider the heated controversy in Aboriginal Studies over the peopling of the Americas (see, e.g., Dewar 2001). The bitter duelling of historians and archaeologists is compounded by white guilt and traditional aboriginal counterclaims.

11.4 For scholars committed to the academic study of those Holy Scriptures produced by the ancient civilizations of southwestern Asia, the challenge may be even greater, given the importance of religious traditions based on those scriptures. At some point, we all have to ask whether Holy Scripture is a special category exempt from the rigours of academic investigation. Biblical Minimalism is fundamentally about rejecting such special categories.

11.5 The archaeologists, the vanguard of this programme, have been able to let the mute stones speak. How much more articulate is the Hebrew language itself, for itself, it we have ears to hear. The circularity can be broken. Now let the chips fall where they may.

12. Bibliography

Barr, James

1995 “The Synchronic, the Diachronic and the Historical: A Triangular Relationship?” Oudtestamentische Studi ën 34: 1–14.

Biber, Douglas, Susan Conrad and Randi Reppen

1998 Corpus Linguistics: Investigating Language Structure and Use. Cambridge Approaches to Linguistics. Cambridge: Cambridge University Press.

Blenkinsopp, Joseph

1992 The Pentateuch: An Introduction to the First Five Books of the Bible. The Anchor Bible Reference Library. New York: Doubleday.

de Caën, Vincent

1995 “On the Placement and Interpretation of the Verb in Standard Biblical Hebrew Prose.” Ph.D. diss., University of Toronto.

Chomsky, Noam

1995 The Minimalist Program. Current Studies in Linguistics, no. 28. Cambridge MA: MIT Press.

Davies, Philip R.

1995 In Search of “Ancient Israel”. Journal for the Study of the Old Testament, Supplement Series, no. 148. Sheffield: Sheffield Academic.

Dewar, Elaine

2001 Bones: Discovering the First Americans. Toronto: Random House.

Deist, Ferdinand E.

1995 “On “Synchronic” and “Diachronic”: Wie es eigentlich gewesen”. Journal of Northwest Semitic Languages 21.1: 37–48.

Epstein, Samuel David, and Norbert Hornstein

1999 Working Minimalism. Current Studies in Linguistics, no. 32. Cambridge MA: MIT Press.

Finkelstein, Israel, and Neil Asher Silberman

2001 The Bible Unearthed: Archaeology’s New Vision of Ancient Israel and the Origin of its Sacred Texts. New York: Free Press.

Fischer, David Hackett

1970 Historians’ Fallacies: Toward a Logic of Historical Thought. New York: Harper Colophon.

Fredericks, Daniel C.

1988 Qoheleth’s Language: Re-evaluating its Nature and Date. Ancient Near Eastern Texts and Studies, no. 3. Lewiston: Edwin Mellen.

Gogel, Sandra Landis

1998 A Grammar of Epigraphic Hebrew. Society of Biblical Literature, Resources for Biblical Study, no. 23. Atlanta: Scholars.

Gutting, Gary (ed.)

1994 The Cambridge Companion to Foucault. Cambridge: Cambridge University Press.

Halpern, Baruch

1988 The First Historians: The Hebrew Bible and History. San Francisco: Harper & Row.

Hayes, John H., and Carl R. Holladay

1982 Biblical Exegesis: A Beginner’s Handbook. Atlanta: John Knox.

Howells, Christina

1999 Derrida: Deconstruction from Phenomenology to Ethics. Key Contemporary Thinkers. Cambridge: Polity.

Hurvitz, Avi

1982 A Linguistic Study of the Relationship between the Priestly Source and the Book of Ezekiel. Cahiers de la Revue Biblique, no. 20. Paris: J. Gabalda.

Jenkins, Keith

1991 Re-thinking History. London: Routledge.

Joüon, Paul

1996 A Grammar of Biblical Hebrew. [2d corrected edition.] 2 vols. Subsidia biblica, no. 14/I-II. Translated and revised by T. Muraoka. Rome: Pontifical Biblical Institute.

Lemche, Niels Peter

1998 The Israelites in History and Tradition. Library of Ancient Israel. London: SPCK.

Mazar, Amihai

1992 Archaeology of the Land of the Bible: 10,000–586 bce The Anchor Bible Reference Library. New York: Doubleday.

McCann, J. Clinton, Jr.

1993 A Theological Introduction to the Book of Psalms: The Psalms as Torah. Nashville: Abingdon.

McCarter, P. Kyle, Jr.

1984 II Samuel. The Anchor Bible, vol. 9. Garden City NY: Doubleday.

McEnery, T., and A. Wilson

1996 Corpus Linguistics. Edinburgh Textbooks in Empirical Linguistics. Edinburgh.

McKenzie, Steven L., Thomas Römer with Hans Heinrich Schmid

2000 Rethinking the Foundations: Historiography in the Ancient World and in the Bible: Essays in Honour of John Van Seters. Beihefte zur Zeitschrift für die alttestamentliche Wissenschaft, no. 294. Berlin: Walter de Gruyter.

Mickelsen, A. Berkeley and Alvera M Mickelsen (2d ed.)

1992 Understanding Scripture: How to Read and Study the Bible. Peabody MA: Hendrikson.

Myers, Jacob M.

1965 Ezra-Nehemiah. The Anchor Bible, vol. 14. Garden City NY: Doubleday.

Neusner, Jacob

2001 Recovering Judaism: The Universal Dimension of Judaism. Minneapolis: Fortress.

Oden, Robert A., Jr.

1987 The Bible Without Theology: The Theological Tradition and Alternatives to It. New Voices in Biblical Studies. San Francisco: Harper & Row.

Peckham, Brian

1993 History and Prophecy: The Development of Late Judean Literary Traditions. The Anchor Bible Reference Library. New York: Doubleday.

Polzin, Robert

1976 Late Biblical Hebrew: Toward an Historical Typology of Biblical Hebrew Prose. Harvard Semitic Monographs, no. 12. Missoula: Scholars. [revised Ph.D. diss., Harvard 1971.]

Qimron, Elisha

1986 The Hebrew of the Dead Sea Scrolls. Harvard Semitic Studies. Atlanta: Scholars.

Rast, Walter E.

1972 Tradition History and the Old Testament. Guides to Biblical Scholarship: Old Testament Guides. Philadelphia: Fortress.

Rooker, Mark F.

1990 Biblical Hebrew in Transition: The Language of the Book Ezekiel. Journal for the Study of the Old Testament, Supplement Series, no. 90. Sheffield: Sheffield Academic.

Schniedewind, William M.

1999 “Qumran Hebrew as Antilanguage”. Journal of Biblical Literature 118.2: 235–252.

Shulman, Ahouva

1996 “The Use of the Modal Verb Forms in Biblical Hebrew Prose”. Ph.D. diss., University of Toronto.

Thompson, Thomas L.

1994 [2d impression] Early History of the Israelite People: From the Written and Archaeological Sources. Studies in the History of the Ancient Near East, no. 4. Leiden: E. J. Brill.

Van Seters, John

1983 In Search of History: Historiography in the Ancient World and the Origins of Biblical History. New Haven: Yale University Press.

Waltke, Bruce K., and M. O'Connor

1990 An Introduction to Biblical Hebrew Syntax. Winona Lake IN: Eisenbrauns.

§13. APPENDIX OF CROSS-TABULATIONS

§1. Predicted Regression to Norm (x-axis final-weak roots; y-axis strong roots)

§1. Predicted Regression to Norm (x-axis final-weak roots; y-axis strong roots)

§2. First Approximation of Biblical Distribution

§2. First Approximation of Biblical Distribution

§3. General Distribution by Biblical Books

§3. General Distribution by Biblical Books

§4. Torah Pericopes (N.B. bulk of book marked *)

§4. Torah Pericopes (N.B. bulk of book marked *)

§5. Distribution of the Psalms: a first approximation

§5. Distribution of the Psalms: a first approximation

§6. So-called History Books: a first approximation

§6. So-called History Books: a first approximation

Endnotes

[1] This paper is a substantially revised version of the paper delivered at the University of Toronto, 23 March 2001. It has been revised in part to incorporate insightful criticisms by Ehud Ben Zvi, Harry Fox, Albert Friedberg, Giuseppe Regalzi, John Van Seters. My scholarship is made possible in part by a generous donation from the nonprofit GRAMCORD Institute (www.gramcord.org) and by continued support by Albert Friedberg.

[2] To address properly Schniedewind’s proposal on a Qumran “antilanguage” would require another paper, which I hope to submit shortly to the Journal of Biblical Literature.

[3] On a related note, the reactionary trend is to various degrees explicit in the growing interest in discourse analysis and textlinguistics. Scholars have also noted an increasing abuse of the Saussurean notion of “synchrony” in such studies as an explicit reaction against historical criticism (e.g., Barr 1995, 11–14; Diest 1995, 46). My programme outlined here is meant to supplement Literaturkritik in a revival of theoretically grounded historical criticism (on the problem of theoretical grounding, see Barr 1995, 9).

[4] Universal Grammar is a technical term in Chomskyan generative grammar. A “grammar” in this generative paradigm is a mathematical model of a given language; at the same time, it is assumed that there is a psychological reality which is being modelled, that a grammar is instantiated in the brain. A “universal grammar” would be a generalized model that accounts for just that typological variation observed across “natural languages” (languages actually spoken by real people, vs computer languages, &c.). Such a universal grammar gains explanatory adequacy to the extent that it can show how language acquisition works (the logical problem of language acquisition), i.e., to the extent that it can show how a speaker moves from an initial state (the child as language-acquisition device or LAD) to a full-blown adult competence in a given language.

[5] James Barr suggested ten to a dozen dialects as a ballpark number: in the light of this study, that number is probably the upper range (Barr 1995, 3).

[6] I am well aware of the distinction in the history and philosophy of science between the context of discovery versus the context of justification. To emphasize, my method of “reverse engineering” is heuristic only. Argumentation regarding the results of the biblical data must ultimately stand or fall independently.

[7] Shulman flags the problem on the last page of her excellent study of the morphosyntax of modality. She concludes by stating, “In order to properly describe and understand the process of change and development which took place throughout the different periods, a comprehensive study of the early material is needed. This study provides the basis for further description of the process of change, and the differences in usage between standard Biblical Hebrew and the Hebrew of later periods” (Shulman 1996, 257). In a sense, the brief study reported on here extends that valuable study to the “later periods.”

[8] Giuseppe Regalzi notes (pc) that this note is absent from the original 1923 French grammar.

[9] John Van Seters (pc) noted the weakness at this point. Working with whole books independent of literary analysis “is not very helpful”. No doubt the combination of Literaturkritik with the method employed here is the ultimate goal, and should provide a powerful exegetical tool; but methodologically the two must proceed independently. However, to counter the problem, I have extended the study in the following section to look for localized patterns: in effect, a linguisitc source-criticism.

[10] It is true that the anchor points are controversially dated and that technically no dating is proved by these distributions. There can be no “proof”, just a coherent picture that suggests hypotheses for further investigation. There is no “proof” in biblical criticism. However, the burden of argumentation should now shift to those who would take issue with this cogent and natural historical-linguistic explanation.

[11] Such assimilation is still marked in the Tiberian reading by the doubling daghesh.

[12] The phenomenon of books, known to be composite, straddling the dividing line might reasonably be expected. The case of this sort of straddling in Proverbs is very instructive in this respect. Certainly, we would expect [-attu] to be found in the late Proverbs. Yet there is a token of the putatively early [-athu] in Proverbs 31:12 within a few verses of the expected form [-attu] in 31:1. Upon closer inspection, we find a clearly identifiable block of material, ascribed to King Lemuel: prima facie, therefore, from an earlier anthology. It is in this block that we find the ex hypothesi earlier [-athu]; however, the later [-attu] is in the editorial superscription, undoubtedly from the latest stratum of the book. Hence, we have a perfect, tailor-made example of the usefulness of this work for source criticism.

The variation in Samuel should not be surprising in light of results obtained above.