Issues in the Linguistic Analysis of a Dead Language, with Particular Reference to Ancient Hebrew1
Robert Holmstedt
University of Toronto
Near and Middle Eastern Civilizations
Rm 328, 4 Bancroft Ave.
Toronto, ON M5S 1C1


With the increasing maturation of the linguistic analysis of ancient Hebrew, it becomes increasingly important that we keep in mind the inherent challenges of analyzing no-longer-spoken languages, like ancient Hebrew. In this article I address a number of such issues in the hopes of provoking some fruitful discussion. First, I address the distinction between linguistic analysis and philological analysis. Then I address some of the major methodological and theoretical challenges facing those who bring modern linguistic theories to bear upon a ‘dead’ language such as ancient Hebrew, including the lack of native speakers, the limited corpus, and the relationship of ancient Hebrew to modern Israeli Hebrew.

1. Introduction

The modern linguistic study of ancient Hebrew is transitioning out of adolescence. This is quite clear from the variety of linguistically-oriented offerings in the primary Hebrew language journals, as well as from the diversity of linguistic theories represented at biblical studies or Semitic language conferences. It is thus an opportune time for those of us engaged in this endeavor to step back and ask (or remind) ourselves what distinguishes the linguistic analysis of ancient Hebrew from philological analysis. A correlative to what distinguishes the linguistic study of ancient Hebrew is what limits there are to the analysis of no-longer-spoken languages, preserved solely in writing. In addressing these issues in this essay, my intent in casting such a wide net is not to offer any final word, but rather to address the general lack of methodological and theoretical reflection in ancient Hebrew linguistics, and to correct a few misunderstandings along the way.

That a methodological and theoretical conversation is conspicuously absent is highlighted by a recent presentation by Jens Bruun Kofoed, “Using Linguistic Difference in Relative Text Dating: Insights from other Historical Linguistic Case Studies.”2 One of Kofoed’s principle motivations is a concern for methodological rigour in the arguments about the history of ancient Hebrew:

[...] I believe it is of equal importance to bring the methodological presuppositions of historical Hebrew linguistics “out into the open and has other people looking at it” as Pooh phrases it. Biblical scholars too often use methods that are either dated or taken out of context in comparison with the way they are used in their “domestic” environments of historiography, anthropology, sociology, linguistics etc., and bringing them out in the open often reveals that “a Thing which seemed very Thingish” inside the perimeters of biblical studies, is less straightforward than you expected when it is brought to test in the broader society of scholars.3

I wholly agree with this critique, and Kofoed proceeds to use case studies from unrelated dead languages, Old English and Eddic (Old Norse), to illustrate the disconnect between the modi operandi within ancient Hebrew studies and the modi operandi of linguists in studying similar issues of language change and dialects in Old English and Eddic. What should add a particular sting to Kofoed’s challenge for greater awareness is that he is primarily a historian within biblical studies, not a linguist. I have stated elsewhere the need for greater methodological and theoretical clarity, and have addressed this desideratum within my own work,4 but it is now time to move beyond simply stating my own positions and provoke a conversation among ancient Hebrew linguists that I hope is a lengthy and productive one.

I will proceed in four parts in this essay. I will first address the various ways to define and distinguish linguistics and philology. Then I will consider the following issues from the standpoint of recent research: ancient Hebrew as a no-longer-spoken linguistic artifact; ancient Hebrew as representative of a ‘real’ language; and ancient Hebrew in light of Spoken Israeli Hebrew.

2. Linguistics and Philology

First, we must attend to terminology. Konrad Koerner reminds us that the original meaning of philologia was “love of learning and literature,” and that both this general meaning, as well as the more specialized meaning of “the study of literary texts,” have been retained in French and German usage; however, the English term “philology” mostly refers to the more narrow “historical study of literary texts” only.5 It is worth considering the history of this association, since it highlights the polemic that is quite often involved in the use of the labels “philology” and “linguistics.” Beginning with the full bloom of the historical-comparative study of languages in the early nineteenth century, the first generation of scholars involved in this type of study (e.g., Friedrich von Schlegel, Jacob Grimm, Rasmus Rask, Franz Bopp) saw their activity as “neue Philologie.”6 It was not until the next generation that August Schleicher (1821–1868) argued for a sharp distinction between Philologie and Linguistik.

For Schleicher, philology was an historical inquiry, using language as a vehicle in the study of culture; in contrast, he asserted that linguistics was the scientific investigation of language itself.7 While Schleicher agreed that quite often the linguist and philologist need each other, he envisioned the two to have two distinct objects of investigation: “In contrast to a philologist, who could work on the basis of the knowledge of only one language (e.g. Greek), a linguist, in Schleicher’s view (1850:4) needs to know many languages, to the extent that ‘Linguistik’ becomes synonymous with ‘Sprachvergleichung’ (p.5).”8

The third generation of historical-comparative scholars, the Junggrammatiker, or “neogrammarians,” were Schleicher’s conceptual heirs; the “jung” did not have the sense of “young Turk” for nothing, though.

It therefore seems strange to us that Berthold Delbrück in his 1880 Einleitung in das Sprachstudium [...] presents Schleicher ‘in the essence of his being’ as a philologist, since thirty years before it had been Schleicher (and no one else) who had clearly set off his work from those of the (classical) philologists. However, if one remembers the ‘eclipsing stance’ which the Young Turks at the University of Leipzig and elsewhere in Germany took vis-à-vis their elders from the mid-1870s to the mid-1880s, one might not be surprised that Delbrück (1842–1922) distorted the facts to suit his argument, namely that the junggrammatische Richtung represented ‘new endeavors’ [...] rather than a continuation of research along established lines.9

Thus the association of philology with “older” historical-comparative study and “linguistics” with “newer” methods does not reflect the earliest uses of the terms (i.e., Schleicher’s), but it was a stance adopted soon thereafter, and it was often taken with a dismissive attitude towards philology.

That the eclipsing stance towards philology continued well into the twentieth century is represented well by Geoffrey Sampson’s brief definition in his history of linguistics: “in modern English usage linguistics normally means linguistics in the twentieth-century style—therefore primarily synchronic linguistics—while philology, if used at all refers . . . to historical linguistics as practised in the nineteenth century.”10 Unfortunately this approach has been appropriated into ancient Hebrew studies. James Barr, in his 1969 article on the interaction between the two disciplines in our field, defines the two terms in just this way.11

Thankfully, every couple generations of language research the conversation arises about what philology and linguistics concern and how the two relate, and since the 1970s there appears to have been a slow shift back to the view that the two activities may have a productive relationship. In this spirit, I think there are two equally tenable approaches that we can take, one that takes philology as a tool in the use of linguistics, and one that sets the two on equal footing but with different objects of study.

The first approach takes philology as the study of written records in order to retrieve linguistic information.12 Thus, Lyle Campbell writes in his introduction to historical linguistics,

In the use of philology for historical linguistic purposes, we are concerned with what linguistic information can be obtained from written documents, with how we can get it, and with what we can make of the information once we have it. The philological investigation of older written attestations can contribute in several ways, for example, by documenting sound changes, distinguishing inherited from borrowed material, dating changes and borrowings, and helping to understand the development and change in writing systems and orthographic conventions.13

Thus it is that Konrad Ehlich can assert philology to be a subsidiary discipline: it is the sometimes necessary handmaiden of linguistics.14 When an old or difficult text must be read, call in the philologist!

Undoubtedly this is a legitimate approach, since philology is closely associated with each of the tasks that Campbell describes above. But there is a framework in which the relationship between philology and linguistics is neither chronological nor hierarchical, viz. they differ only in their specific objects of investigation. Consider how H. A. Gleason delineates the two activities: “Linguistics, at least potentially, deals with those things which are common to all texts in a given language, whereas philology deals with those things which are peculiar to specific texts.”15 Writing for an ancient Hebrew studies audience, Bodine appropriates Gleason’s distinction and elaborates:

Philology gives attention to particular texts (usually of a literary nature and written), seeks to elucidate features of these texts which are more-or-less language specific, emphasizes the content of the texts, and draws implications that are related to the culture in which the texts were produced. Linguistics, on the other hand, studies speech with an eye to language qua language, attends more to features of its texts and other sources of information which are shared among languages rather than language specific, is concerned more with the structure of language than the content of texts, and is more theoretically than culturally oriented.16

To summarize, then, linguists have as their goal the system of language, whereas philologists have as their goal a better understanding of the meaning of the text being observed, and language is simply the primary means to that end.

But is “text versus system” all there is to the distinction between philology and linguistics? Would that it were so simple. There exists yet another important axis by which we can distinguish the two disciplines—by their primary (but not sole) method of inquiry. Philologists primarily adopt an inductive approach in that they take a finite corpus and reconstruct the grammar of that corpus from within. In contrast, linguists, particularly within the generative approach,17 adopt a deductive approach in that they proceed from a small set of presuppositions about the human mind, “language,” and attested language systems and use the data to test and refine these hypotheses.18 That this is to some degree a legitimate distinction between the two approaches is supported by the common criticisms leveled by each against the other. On the one hand, philologists often claim that linguists impose theory on the data; on the other hand, linguists often describe philological activity as little more than listing and categorization of forms (i.e., simple, and therefore mostly un-insightful, taxonomy).

Perhaps it is personal bias on my part, since I have formal training in both philology and linguistics, but I refuse to think that there is no way around this animus. I prefer an approach that allows for a functional and productive working relationship between the two disciplines. In other words, let us allow that the tools may be the same for philologists and linguists, but that the goals differ. Whereas philologists study specific texts, linguists study linguistic systems and even the internal (mental) grammars of native speakers. Whereas philologists privilege the finite corpus and are reticent to hypothesize beyond the extant data (in good Bloomfieldian fashion), linguists recognize that no corpus represents the infinite set of sentences available to the native speaker. That is, linguists recognize that data from a corpus-bound study will “always underdetermine grammar.”19

Thus, I advocate a complementary approach, in that each discipline is able to address potential weaknesses in the other. As Jan Faarlund has recently stated, “A linguist working on historical material depends on a good philologist.”20 I agree. Philologists read the texts, sort through the data, establish what is available, and categorize it. Minimally, a philologist tells us what is there to study, and maximally, the reconstruction of the grammar that he provides may be accurate. But it is the role of linguist to check this reconstruction. She compares the given data to unrelated linguistic systems in order to determine possible correspondences as well as potential grammatical gaps or even mistakes. She also checks the reconstructed grammar against the accumulated evidence and hypotheses of modern theories, such as the principles and parameters of generative linguistics.

In this way, any Hebraist who investigates the linguistic features of a particular corpus, e.g., a passage or book of the Hebrew Bible, is engaging in philological analysis. In contrast, those who examine linguistic features in light of some linguistic theory in order to make sense of some dialect or stratum of ancient Hebrew as a system are engaging in linguistic analysis.21 And, those of us who examine specific texts or corpora as well as linguistic systems can identify ourselves as both philologists and linguists.

3. Analyzing No-Longer-Spoken Languages

The reconciliation of philology and linguistics does not address an issue that is of central importance to the study of ancient Hebrew, though. “How can we analyze a ‘dead’ language?” remains a troublesome question. Faarlund begins his handbook on the syntax of Old Norse by identifying the challenge of using modern linguistic theory to describe no-longer-spoken languages.

Describing the syntax of a dead language is rife with theoretical problems and methodological stumbling blocks. A major question is determining what the description should seem to describe. Traditional, philologically oriented grammars of dead languages are descriptions of finite corpora. Modern generative grammar, on the other hand, aims to account for speakers’ linguistic competence, their internalized grammar....In the absence of live speakers and their intuitions, and in the absence of contemporary syntactic descriptions, our sources of knowledge of the internalized grammar of the speakers are limited to extant texts, besides grammatical theory.22

Linguists, like Faarlund, are increasingly taking on the challenge of analyzing no-longer-spoken languages. While not yet wildly popular, the linguistic analysis of, e.g., Old English, Old Norse, Middle Dutch, Middle French, Early Modern English has a respectable place at the table. And with the recent addition of generative-oriented studies on Old and Coptic Egyptian, Sumerian, Akkadian, biblical Hebrew, Indo-European, classical Greek, Latin, and classical Sanskrit,23 this new trend is slowly but surely expanding in scope.

Yet, linguistic analysis of a ‘dead’ language is a formidable challenge, and discussions of the theoretical challenges lag behind the analysis of the data. It is well-known that, at least within formal linguistic approaches such as generative grammar, native speaker intuition24 is crucial. Why? If a language is represented by the infinite diversity of utterances that the grammar residing in the mental language faculty of the ideal listener-speaker is able to generate, it is impossible to describe the entire language or even to specify all possible constructions. However, the linguist can use the intuition of native speaker informants to discern the boundaries of what is possible.25 Hence, great importance is assigned to both counter-examples and the so-called “ungrammatical” examples (those marked with asterisks): they illustrate grammatical boundaries for the purposes of syntax and provide necessary clarification for semantic readings and pragmatic nuances. This allows whatever proposals are put forth to meet the criteria of observational and descriptive adequacy and to predict the acceptability of novel examples. Consider the contrast among the three English examples in (1).

(1a) I didn’t see that the red car went into the ditch

(1b) I didn’t see which car went into the ditch

(1c) *I didn’t see which car that went into the ditch

The first two examples illustrate a complement clause and a preposed wh-phrase, respectively. The third example, in (1c), illustrates that a wh-phrase + that sequence is not acceptable in modern standard English. By this process, we are able to establish a linguistic boundary between the grammatically “good” and “bad.” Unfortunately, we don’t have this type of evidence for ancient Hebrew. The lack of native speakers for ancient Hebrew puts us at a distinct disadvantage; it is impossible to elicit fresh data or to check the semantic reading or pragmatic nuance of the data we do have. To put a fine point on it, are we treading on thin ice to propose unattested counter-examples, as in (2) below, based on the extant data set, and perhaps a bit of intuition?26

(2a) כְּכֹל אֲשֶׁר צִוָּה אֹתוֹ אֱלֹהִים‎ (Gen 6:22)

(2b) unattested: אֱלֹהִים צִוָּה אֹתוֹ אֵת הַכֹּל

The clause in (2a) is a relative clause, with a verb-object-subject order within the relative clause; (2b), then, represents the non-relative version of the same clause, with a proposed subject-verb-object(s) order, even though the particular clause in (2b) is not attested within the ancient corpus.27 Is such reconstruction legitimate? Most philologists would consider this process of reconstructing unattested examples methodological thin ice, indeed, particularly if it were used to bolster some other novel or controversial claim, such as basic subject-verb-object word order for ancient Hebrew. So what recourse do we have?

The linguist of an ancient language like Hebrew must accept two principles of operation. First, nearly all of the extant ancient data is grammatical, interpretable, and pragmatically felicitous within its discourse context. Aside from some disagreements at the fringes,28 if most of the data were not grammatically acceptable, then we would have nothing upon which to reconstruct the grammar! Second, as the aforementioned quote by Faarlund makes clear, a linguistic theory that has been developed and tested on languages for which native speakers exist must provide the interpretive filter. So, as Kiss asserts,

The difficulties of reconstructing the grammar of ancient language resemble the difficulties that a child experiences when reconstructing the grammar of its mother tongue. A child acquiring its mother tongue, too, has access only to a limited—and sometimes defective—set of positive evidence (the correction of a child’s mistakes by adults is by no means a necessary element of language acquisition). If the two processes are similar, then the methodology adopted in the reconstruction of the grammar of an ancient language must also be similar to that employed by a child in the course of language acquisition. What the child does is interpret the data it has access to on the basis of the genetically coded Universal Grammar that it possesses. This is what the linguist setting out to reconstruct the grammar of a dead language must do, as well; he or she must interpret the data available as indications of how the open parameters of Universal Grammar are to be set.29

Being bound by a corpus that cannot be tested against native-speaker intuition also has implications in the opposite direction, for the contribution of ancient Hebrew study to linguistic theory. In other words, while it is unquestionable that we can use insights from other languages and linguistic theory to elucidate the structure of ancient Hebrew, is it possible to use ancient Hebrew to elucidate features of other languages, or more specifically, to contribute to the generative theory of Universal Grammar? The answer is a qualified “yes” to the first question and a “maybe” to the second. Given that the generative search for Universal Grammar is ultimately about the “language faculty” within the human mind, and also given that the corpus of data for an ancient language like Hebrew severely limits our access to the ancient Hebrew speaker’s mind, the question is whether we can legitimately make significant additions or modifications—if any at all—to the understanding of Universal Grammar? (I offer this as an open question, not a rhetorical one.)

Let me illustrate. It is often asserted that one of the features of left-dislocation in English is that it cannot be embedded, as in example (3).30

(3a) No Left-Dislocation: He loves Abigail (cp. Because he loves Abigail, he bought her a toy)

(3b) Left-Dislocation: Abigail—he loves her (cp. *Because Abigail—he loves her, he bought her a toy)

The examples in (3b) illustrate how left-dislocation is prohibited in embedded structures, and apparently this constraint is so well attested that it once was considered a fact of left-dislocation generally.31 But consider the biblical Hebrew examples in (4).

(4)עֵינֵיכֶם הָרֹאֹת אֵת אֲשֶׁר עָשָׂה יְהוָה בְּבַעַל פְּעוֹר כִּי כָל הָאִישׁ אֲשֶׁר הָלַךְ אַחֲרֵי בַעַל פְּעוֹר הִשְׁמִידוֹ יְהוָה אֱלֹהֶיךָ מִקִּרְבֶּךָ

Your eyes were those that saw how Yhwh acted at Baal Peor, that every man who followed Baal Peor—your God destroyed him from your midst (Deut 4.3)

What do we with such examples (and there are more)? We appear to have a case in (4) of left-dislocation embedded within a non-root clause, in this example, a complement clause. But is it really? Can we suggest a modification of a general linguistic principle, even though we can’t check the grammaticality or felicitousness by appealing to native speakers? Or must we read ancient linguistic data through the strict lens of principles, parameters, and various derivative conditions built upon modern data? I find no easy answer to this.

Moreover, I would argue that certain areas of grammar are more accessible, or to put it another way, we can be more confident of some reconstructions than of others. The written nature of the ancient Hebrew texts limits our access to the phonetics of ancient Hebrew (as opposed to Tiberian Masoretic Hebrew).32 As Cynthia Miller has recently noted, if we had a resurrected native speaker for ancient Hebrew, one “could confirm that the orthographic symbol ‘ayin (ע‎) was used to represent two sounds in ancient Hebrew—the pharyngeal /ʕ/ as well as the postvelar fricative /ɣ/.”33 Without the resurrected native speaker, though, we are limited to determining a phonetic range for each consonant, based on distribution with other consonants and representation in contact languages, such as Akkadian, Aramaic, and Greek.34

We are similarly deprived of most prosodic features of the language, and the possible connections or uses as signals for syntax (e.g., restrictive versus non-restrictive relative clauses) and pragmatics (e.g., intonation used to mark focus).35 And what about issues of meaning, such as verbal semantics and idioms? We can tell a good story on any of these topics, but no matter how good, coherent, logical, or empirically grounded, it doesn’t mean it’s accurate. Yet, piecing together such proposals is what we must continue to do, hopefully with a greater degree of linguistic refinement at each step, and perhaps every few decades with an additional piece of epigraphic evidence that sheds light on known cruces or adds support to tentative proposals.

Finally, I want to address whether there are any linguistic approaches that cannot be used for the study of ancient Hebrew. For instance, William Schniedewind has recently asserted that,

Generative Grammar is not appropriate for the study of ancient written languages, and especially for a specific ancient language like Classical Hebrew, because the assumptions and methodology of Generative Grammar are based on vernacular and on the premise of linguistic universals in spoken languages. Since Classical Hebrew is known to us only as a written language, the traditional and formal linguistic approaches that underlie most modern studies of Classical Hebrew seem especially inappropriate.36

There are many problems with this position, of which I will critique only a few.37 First, Schniedewind makes unsubstantiated assumptions about the nature of the extant data, for instance, that they preserve only the more literary written registers. Yet, some of the epigraphic letters that Schniedewind has published on arguably represent the “vernacular” form of the language; simply because these letters have been written down does not mean that the actual linguistic system or register differs significantly from if they had remained oral (i.e., one must distinguish between a genre, such as “letter,” and a register of the language or linguistic system, which Schniedewind does not). Moreover, even if for the sake of argument we were to grant him (and I am not convinced of this) that the conventions of writing in a primarily non-literate society are “particularly artificial” and “often remain unchanged even after speech-forms have undergone profound linguistic changes,”38 it does not at all follow from the supposed existence of solely literary forms (e.g., letters) or the conservative nature of writing versus speech (including the spelling conventions that he uses as illustration) that the morphology, syntax, and semantics of the oral and written texts39 necessarily differ. Such assertions must be supported, and Schniedewind does not do so.

Second, Schniedewind has ignored the history of generative linguistics in the last twenty years, and has misunderstood the generative enterprise as a whole. On the former point, generative linguistics has increasingly moved beyond its understandably narrow beginnings, in that where it once was focused primarily on spoken European languages it has now branched out to include languages from every corner of the globe as well as a wide selection of no-longer-spoken languages. Additionally, the study of Universal Grammar is now being applied to diachronic phenomena, not simply synchronic ones (it is worth noting that there was never any theoretical reason that prohibited generativists from examining diachronic issues; rather, it was a matter of prioritizing possible objects of study in order to construct and test a theory and the various frameworks by which it has evolved).

On the latter point, the assertion that Chomskyan linguistics “denies the importance of the study of particular languages” because it has as its primary goal Universal Grammar is sorely misinformed. It is true that Chomsky’s overriding interest is determining Universal Grammar as a means to understanding the human mind better. However, it is obvious that the means to determining Universal Grammar is the study of particular grammars. And it is quite possible that ancient Hebrew has something to offer the study of Universal Grammar (see above, the discussion surrounding examples [3]–[4]).

4. The Linguistic Status of Biblical or Ancient Hebrew

One area of investigation in which we are now seeing a slow but steady degree of refinement concerns the linguistic status of ancient Hebrew. The Semitist Edward Ullendorff famously asked almost thirty years ago whether biblical Hebrew presents us with the adequate data to consider it a ‘language.’ His conclusion to this question was negative: “In the sense in which I have been endeavoring to present the problem biblical Hebrew is clearly no more than a linguistic fragment.”40 But it is not clear that Ullendorff is correct.

The problem is that Ullendorff’s argument simply does not pertain to the linguistic status of biblical Hebrew as a “language” in the technical sense of the word. A language, in all of its complexity and unlimited expressive variability41 is not an object that is possible to describe fully. So on the issue of the ancient evidence as a “language” the answer is incontrovertibly “no.” But Ullendorff’s definition of language as “a system of communication” means that he was really considering the question, “is biblical Hebrew a linguistic system?” and Miller has cogently pointed out in her recent article, “language” and “linguistic system” are two very different concepts.42 If we take all of the ancient Hebrew data together, they arguably present us with a good representation of the linguistic system.43 Even so, in the spirit of Ullendorff’s essential objection, we should ask whether we are justified in treating it all as a single “linguistic system,” as is the practice of both teaching and reference grammars.44

Minimally, there has been a consensus for over two centuries that at least three historical stages of Hebrew are witnessed to by the biblical and epigraphic evidence, typically referred to as archaic, classical, and late biblical Hebrew. Thus, we are faced with three linguistic systems; moreover, three linguistic systems now appears to be a deficient position. Gary Rendsburg is well-known for advocating a northern dialect of Hebrew within the biblical material that can be distinguished form the dominant southern dialect.45 Rendsburg and Frank Polak each present evidence for a barely discernible distinction between oral/vernacular/colloquial Hebrew and written/literary/formal Hebrew.46 And Ian Young has tirelessly championed the diversity of Hebrew in the pre-exilic period while also challenging the traditional three-stage chronological model.47 Additionally, Rendsburg, Young, and, just recently, I have asserted that some texts reflect deliberately artificial language, whether in terms of the choice of lexical items or syntactic structure.48

For example, consider the case of the relative word שׁ‎. It is often identified as a remnant of both a northern dialect and a standard feature of Second Temple period Hebrew.49 Additionally, the cases of שׁ‎ have been explained as instances of a Hebrew vernacular (and אשׁר‎ would then represent the literary idiom), which increasingly exerted influence on the literary register in the later Second Temple period.50 Such an explanation, that שׁ‎ was originally the northern colloquial relative word, made its way south after 722 B.C.E, and infiltrated the literary register until it became the item of choice by the period of the Mishna, may account for many of the occurrences, but not all. In some cases, it appears that the distinction between אשׁר‎ and שׁ‎ was used as a literary device, specifically to create a northern Hebrew or “other” atmosphere. For instance, in 2 Kgs 6:11, given in (5), the שׁ‎ is placed in the mouth of an Aramean king.

(5) וַיִּסָּעֵר לֵב מֶלֶךְ אֲרָם עַל הַדָּבָר הַזֶּה וַיִּקְרָא אֶל עֲבָדָיו וַיֹּאמֶר אֲלֵיהֶם הֲלוֹא תַּגִּידוּ לִי מִי מִשֶּׁלָּנוּ אֶל מֶלֶךְ יִשְׂרָאֵל

And the heart of the king of Aram was moved concerning this matter and he called to his servants and said to them, “Will you not tell me who of those that are ours is for the king of Israel?” (2 Kgs 6:11)

Not only is this reported speech placed in the mouth of the Aramean king, it is also spoken to his Aramean subordinate. That the use of שׁ‎ in this verse is a technique by which to characterize the king of Aram as “foreign” has been noted by Avi Hurvitz and followed by Rendsburg.51 Similarly, the three examples of שׁ‎ in Jonah can be explained as intentional literary devices. The first time we encounter שׁ‎ is in 1:7 (6a); note that the clause presents us with reported speech and it is the foreign sailors who speak.

(6a) וַיֹּאמְרוּ אִישׁ אֶל־רֵעֵהוּ לְכוּ וְנַפִּילָה גוֹרָלוֹת וְנֵדְעָה בְּשֶׁלְּמִי הָרָעָה הַזֹּאת לָנוּ

And each man said to the other: Come, let us cast lots so that we may know on account of (the deed) that belongs to whom this catastrophe (has come) to us. (Jon 1:7)

Whereas (6a) presents the sailors speaking among themselves, or at least to a general audience on board, in (6b) they address Jonah directly. It is significant that these non-Israelites use אשׁר‎ instead of שׁ‎.

(6b) וַיֹּאמְרוּ אֵלָיו הַגִּידָה־נָּא לָנוּ בַּאֲשֶׁר לְמִי־הָרָעָה הַזֹּאת לָנוּ

They said to him: Tell us on account (of the deed) that belongs to whom this catastrophe (has come) to us? (Jon 1:8)

Yet just a few verses later we encounter Jonah speaking to the sailors and using שׁ‎.

(6c) כִּי יוֹדֵעַ אָנִי כִּי בְשֶׁלִּי הַסַּעַר הַגָּדוֹל הַזֶּה עֲלֵיכֶם

Because I know that on account (of the deed) that belongs to me this great storm (has come) upon you. (Jon 1:12)

In chapter 1, the characters are assigned speech patterns and thus interact with each other based on their constructed identities. In Jonah 1 the contrast between the Israelite Jonah and the non-Israelites sailors, and thus the characterization of each group, is a major component of the message, so we should expect that the sailors use their “foreign” speech when talking among themselves (6a) (just as we saw in the Aramean court in 2 Kgs 6:11), but speak to Jonah in his own language (6b), and similarly that when Jonah speaks to the sailors he emulates their tongue (6c). Crucially, this description is cogent only if שׁ‎ was not a standard lexeme within the audience’s grammar(s); שׁ‎ must have been perceived as “foreign,” i.e., non-standard, even within non-literary registers.

When we turn to the single occurrence of שׁ‎ in Jonah 4, given in (7), we should first note that it is almost immediately preceded by the use of אשׁר‎.

(7) וַיֹּאמֶר יְהוָה אַתָּה חַסְתָּ עַל־הַקִּיקָיוֹן אֲשֶׁר לֹא־עָמַלְתָּ בּוֹ וְלֹא גִדַּלְתּוֹ שֶׁבִּן־לַיְלָה הָיָה וּבִן־לַיְלָה אָבָד

Yhwh said: You had pity on the castor-oil plant, which you did not toil over or raise, which was one night old and perished (as) a one-day-old (plant). (Jon 4:10)

Certainly Yhwh was not perceived as a foreigner to the book’s Israelite audience, and, in any case the presence of אשׁר‎ attenuates any suggested foreign characterization. Why, then, does the author switch the relative words within a single utterance of Yhwh’s? Again, it fits the book’s overall rhetorical purpose, which builds strongly in the final chapter: Yhwh is not just the deity of the Israelites, his domain and care extends well beyond the borders of Israel.52 Thus, the author has used a subtle shift in style, viz., the switch in relative words from אשׁר‎ to שׁ‎, to reinforce his point linguistically.

What this means for out study of Hebrew is clear: in addition to the remnants of real dialects, historical stages, and registers, we may also be dealing with artificial Hebrew, used for rhetorical purposes; crucially, this latter perspective may remove certain features, like שׁ‎, from consideration from the former perspectives. Once again, in light of all these issues, it is a significant linguistic concern whether it is responsible and accurate to speak of “biblical” Hebrew as a single linguistic system. Perhaps it has come to the point at which a new bottom-up approach is needed, in which separate descriptions are constructed for each “bibliolect,” that is, the grammar of each text.53

We can no longer work with three broad biblical stages—archaic, classical, and late biblical Hebrew—and then a sweeping post-biblical stage. Instead, we have to reckon with, in the least, archaic, archaizing, pre-exilic, exilic, Persian, Hellenistic, early Roman, and Tannaitic, along with Israelian, Judahite, faux Aramean, and oddly sectarian (i.e., certain Qumran texts like the Temple Scroll) forms of Hebrew, some in chronological sequence, some coexistent.54 While many of these issues have been discussed, as of yet there has been no synthesis that presents a plausible description with this level of sophistication of the variety of ancient Hebrew data to which we have access.

5. Ancient and Modern Hebrew

Finally, let us turn our attention to the comparative value of Spoken Israeli Hebrew—can we and should we use it to help us understand ancient Hebrew? Even if we conclude that the lexical inventory of Spoken Israeli Hebrew has undergone too much change, what about the grammatical structure? And if we do utilize Spoken Israeli Hebrew for comparison, do we restrict the data to the more formal registers of literature, as the best analogue to the Bible, or do we admit modern colloquial data? Clearly there are many issues to address. But an even larger one looms and casts a long shadow over the relationship between the ancient and modern stages of Hebrew: there is no consensus on the linguistic origins of Spoken Israeli Hebrew—is it really a Semitic language, or should it be classified instead as an Indo-European linguistic system with a Hebrew lexicon?

A great deal of debate in the late nineteenth and early twentieth centuries centered on whether the modern form should be based primarily on the form and lexicon of biblical Hebrew or that of rabbinic Hebrew,55 or both.56 This was especially the concern of prescriptivists and the official language academy. But among linguists concerned more with description than prescription, the debate hinged, and in some circles continues to hinge, on the genetic and typological status of Spoken Israeli Hebrew: was it really a Semitic language, given the undeniable influence of Yiddish, Slavic, German, English, French, and Spanish, or should it be classified instead as an Indo-European language, as some continue to claim? A rather pointed challenge to the Semitic background of Spoken Israeli Hebrew was advanced by Paul Wexler in The Schizoid Nature of Modern Hebrew: A Slavic Language in Search of a Semitic Past, the title of which leaves no doubt about his thesis.57

This lack of consensus, along with a complex web of ideological motivations, has resulted in the aforementioned shadow cast over the fruitful study of Hebrew as single language with ancient to modern stages. Rarely are Spoken Israeli Hebrew data used as comparative evidence in the study of biblical Hebrew due to the skepticism over the ‘sufficiently’ Semitic nature of Spoken Israeli Hebrew.58 While this is more understandable for phonetics, phonology, morphology, and lexical semantics, it cannot be assumed a priori with regard to syntax and the semantics of non-lexical issues (e.g., the verbal system).

Furthermore, the rejection or willful ignorance of Spoken Israeli Hebrew data does not take seriously the continued existence and development of literary Hebrew within numerous pre-twentieth century Jewish communities nor the use of Hebrew as a “Jewish lingua franca” both within the Diaspora and Palestine well before Ben-Yehuda.59 And perhaps “revival” is conceptually inaccurate for Spoken Israeli Hebrew, but rather the language underwent “revernacularization” as Shlomo Izre’el suggests.60 That is, as a language whose literary register had become the sole stratum and whose vernacular had ceased to exist, at least as a first language, “death” and “revival” or “resurrection” are not the correct metaphors. If this is so, and it is also the case that the literary registers of the modern language are the most appropriate with which to compare biblical Hebrew, then the precise status of colloquial Israeli Hebrew and the process by which it emerged are mostly irrelevant for comparative purposes.

In any case, since almost all of the data used to argue for or against the Semitic nature of Spoken Israeli Hebrew have been phonological, morphological, and lexical in nature, the uselessness of Spoken Israeli Hebrew syntax and pragmatics for elucidating ancient Hebrew has not yet been compellingly argued. Additionally, the genetic status of Spoken Israeli Hebrew is hardly relevant; its typological nature is more important, and even Wexler allows at least that Spoken Israeli Hebrew has become typologically Semitic. And if Spoken Israeli Hebrew is minimally typologically Semitic, then more studies that include all stages of Hebrew, like Yitzak Peretz’s volume on the relative clause,61 are necessary for us to discern just how many differences and what type exist between the stages.

Let me provide one brief but potentially significant comparison, and then conclude with a word of caution. In the formal registers of Spoken Israeli Hebrew that one witnesses in literature, we find a basic variation between the normal SV word order as in (8a) and VS order with a clause-initial constituent, like a subordinating function word as in (8b) or a fronted phrase such as the temporal modifier in (8c).

(8a) דני כתב את המכתב

(8b) אני ראיתי שׁכתב דני את המכתב

(8c) עכשׁיו כתב דני את המכתב

Recently, I have argued that biblical Hebrew contains similar variation, illustrated in (9a) with basic SV order and inverted VS order with an initial function word like the relative clause example in (9b).62 This also explains the frozen VS order in the complex wayyiqtol, illustrated in (9c), which is also VS and appears to have a reduced function word manifested only in the gemination of the prefix consonant.

(9a) פֶּרֶץ הוֹלִיד אֶת־חֶצְרוֹן ‎(Ruth 4:18)

(9b) כְּרָחֵל וּכְלֵאָה אֲשֶׁר בָּנוּ שְׁתֵּיהֶם אֶת־בֵּית יִשְׂרָאֵל‎ (Ruth 4:11)

(9c) וַיָּמָת אֱלִימֶלֶךְ אִישׁ נָעֳמִי ‎(Ruth 1:3)

Any significant similarities between biblical Hebrew and Spoken Israeli Hebrew should not obscure real linguistic differences among the stages of Hebrew, though. The language has clearly changed in certain aspects, such as the semantics of the verbal system, and those of us who use Spoken Israeli Hebrew, or those for whom it is their native language, must resist the temptation to impose Spoken Israeli Hebrew features back on to earlier stages of the language. For instance, the tense-system of Spoken Israeli Hebrew should not be assumed for any of the earlier stages, such as biblical Hebrew, but must be shown to explain the data adequately. And at present, the most convincing proposals for biblical Hebrew are aspectual,63 although the shift toward a tense-based system by rabbinic Hebrew suggests that Hebrew as a whole was moving towards a tense system. A significant desideratum in the study of Hebrew linguistic history is determining whether the verbal system in Spoken Israeli Hebrew represents the typologically expected result of the evolution leading up to that point.

6. Conclusion

So, how do linguistics and philology relate, and what limits are there to the linguistic analysis of ancient Hebrew? The “system” orientation native to linguistics contra philology, along with the attention to relevant cross-linguistic patterns (including those from Spoken Israeli Hebrew), suggests that cautious and theoretically-informed linguistic analysis holds immense potential for clarifying numerous long-standing grammatical cruces.

However, we must operate with full awareness of the limits of our study, since the nature of the ancient Hebrew data impose certain constraints on the conclusions we might draw. We must reckon with the limited and varied corpus, distinguishing as best we can all of the discernible registers, dialects, and other strata. Since we lack native speakers, who could have provided us with further data as well as intuitive judgments about grammaticality, etc., we must admit that any and every proposal we make is at the mercy of new epigraphic tidbits, or any newly identified construction hiding in the biblical, Qumran, or mishnaic corpora. While this is so for all proposals for any spoken language, it is much more the case, and much closer to the surface for analyses of no-longer-spoken languages, for at least with spoken languages the potential for fully descriptive and explanatory analyses exists in the abstract; for ancient Hebrew, we’ll never know how close we’ve come. And so, we must take extra care in our analyses and write with considerable humility.

[1] This article is a revision of a paper, “What Linguistics Has to Offer Ancient Hebrew Studies (and What It Doesn’t),” presented at the annual meeting of the Association of Jewish Studies, Washington D.C., December 18, 2005. I with to thank John A. Cook and Cynthia L. Miller for reading and commenting on this essay. I alone am responsible for all opinions and errors contained within.

[2] This paper was presented at the National Association of Professors of Hebrew (NAPH) Session at the Society of Biblical Literature’s Annual Meeting in Philadelphia, November 2005. I am grateful to Prof. Kofoed for sharing the revised version (fc., Hebrew Studies) with me.

[3] Ibid, p. 2.

[4] R. D. Holmstedt, “Adjusting Our Focus (review of Katsuomi Shimasaki, Focus Structure in Biblical Hebrew: A Study of Word Order and Information Structure)” (Hebrew Studies 44 [2003]:203–15; idem, “Word Order in the Book of Proverbs” (Pp. 135–54 in Seeking Out the Wisdom of the Ancients: Essays Offered to Honor Michael V. Fox on the Occasion of His Sixty-Fifth Birthday, ed. K. G. Friebel, D. R. Magary and R. L. Troxel; Winona Lake, IN: Eisenbrauns, 2005).

[5] K. Koerner, “Linguistics vs Philology: Self-Definition of a Field or Rhetorical Stance?” (Language Sciences 19/2 [1997]:167–175), p. 168.

[6] Ibid, p. 169.

[7] Ibid, p. 170.

[8] Ibid. In this quote, Koerner references the introductory chapter, “Linguistik und Philologie,” in Schleicher’s 1850 work Die Sprachen Europas in systematischer Uebersicht (Bonn: König; reprinted with introduction by K. Koerner; Amsterdam: John Benjamins, 1983).

[9] Ibid, p. 171.

[10] G. Sampson, Schools of Linguistics (Stanford: Stanford University Press, 1980), p. 243, n. 1.

[11] So J. Barr, “The Ancient Semitic Languages—The Conflict between Philology and Linguistics” (in Transactions of the Philological Society 1968; Oxford: Basil Blackwell, 1969), pp. 37–55, especially p. 37.

[12] Although less common among linguists, there still exists the broader definition of philology as the study of texts in order to interpret them and their cultural setting (so P. Daniels, “Writing Systems” (pp. 43–80 in The Handbook of Linguistics, ed. M. Aronoff and J. Rees-Miller; Oxford: Blackwell, 2001), p. 76.

[13] L. Campbell, Historical Linguistics: An Introduction. 2nd Ed. (Edinburgh: Edinburgh University Press, 2004), p. 362.

[14] K. Ehlich, “Native Speaker’s Heritage: On Philology of ‘Dead’ Languages” (in A Festschrift for Native Speaker; ed. F. Coulmas; The Hague: Mouton Publishers, 1981), pp. 153–165.

[15] H. A. Gleason, “Linguistics and Philology” (in On Language, Culture, and Religion: In Honor of Eugene A. Nida, ed. M. Black and W. A. Smalley, Jr.; The Hague: Mouton, 1974), pp. 199–212, quote from p. 200.

[16] W. R. Bodine, “Linguistics and Philology in the Study of Ancient Near Eastern Languages” (in “Working with No Data”: Semitic and Egyptian Studies Presented to Thomas O. Lambdin; ed. D. M. Golomb. Winona Lake, IN: Eisenbrauns, 1987), pp. 39–54, quote from p. 40.

[17] Here is a point at which my goal in this article to initiate methodological and theoretical dialogue comes to the fore: it would be extremely useful to have a functionally-oriented Hebrew linguist comment on the inductive-deductive divide as well as the finite-corpus versus “mind” issues that I use here to distinguish philology from generative linguistics.

[18] See K. É. Kiss, “Introduction” (in Universal Grammar in the Reconstruction of Ancient Languages, Studies in Generative Grammar 83; ed. K. É. Kiss; New York: Mouton de Gruyter, 2005), pp. 1–30.

[19] Ibid., p. 2.

[20] J. T. Faarlund, The Syntax of Old Norse: With a Survey of the Inflectional Morphology and a Complete Bibliography (Oxford: Oxford University Press, 2004), p. xiii.

[21] If this distinction is accurate, it entails—and this is significant for our current discussion—that linguistics in the broadest sense has as its object of study any language or dialect of language, whether spoken or written and whether ancient or modern, and accordingly uses any evidence that contributes to the analysis. Thus, one can only hope that younger Hebraists ignore Barr’s statement in the same article in which he claims that “if philology has its centre in the study of classical texts, linguistics has its centre in the observation of spoken languages” (“The Conflict between Philology and Linguistics,” p. 37). On this issue, see further my critique of W. M. Schniedewind below.

[22] Faarlund, The Syntax of Old Norse, p. 1.

[23] These studies have been collected in Kiss, Universal Grammar in the Reconstruction of Ancient Languages. It is notable that the studies in this volume “aim to demonstrate that descriptive problems which proved to be unsolvable for the traditional, inductive approach to ancient languages can be reduced to the interaction of regular operations and constraints of the hypothetical Universal Grammar” (Kiss, “Introduction,” p. 3).

[24] We should not confuse the use of intuition in generative linguistics with the common use of intuition. Often when we use the word intuition, the connotation is that of guesses and luck, something very “unscientific.” And, sometimes generative grammar has been criticized this way. However, this reflects a misunderstanding of the use of intuition in generative linguistics, in which it refers to “tapping into our subconscious knowledge” (A. Carnie, Syntax: A Generative Introduction [Malden, MA: Blackwell, 2002], p. 12). In this way, linguistic intuition about issues like grammaticality are no different, and no less scientific, than visual judgments about colors or shapes.

[25] Note that while native speakers cannot provide direct, conscious information about their mental grammar(s) in the form of linguistic analytical propositions, their judgments and intuitions on the grammaticality/acceptability and felicitousness of data do provide access to their mental grammar(s), although admittedly a step removed from the ideal of competence (given that even judgments reflect performance). See N. Smith (Chomsky: Ideas and Ideals [Cambridge: Cambridge University Press, 1999], pp. 28–48) for a clear discussion of these issues.

[26] See Ehlich, “Native Speaker’s Heritage,” for the following (very intriguing) claim:

“...we find the philologist in a position that is parallel to that of the ‘linguist native speaker (LNS).’ The philologist develops a kind of linguistic knowledge of which he makes use by introspection. Since there is no native speaker..., he himself is the only one who can really “speak” the language, i.e., who develops a concrete, individualized competence in that language, a competence comprising all of the elements that make a language. This secondary competence of the philologist is open to introspection, and introspection is the main way how the philologist comes to systematic results on structures of his subject....The philologist is his own LNS” (161).

I think that Ehlich overestimates the “competence” of even the most gifted philologist. Or at least Ehlich’s understanding of competence is not that of the technical sort discussed within generative linguistics, in which competence refers to the mental language faculty of the ideal native speaker-listener. No amount of expertise in a “dead” language, gained from textual remains, could allow one to achieve this technical kind of competence, and thus no philologist could be considered a suitable stand-in for a true native speaker (no matter how much we would like to flatter ourselves!).

[27] Note that I am highlighting in (2a–b) only the issue of the possible word order difference between the relative (2a) and non-relative (2b) clauses. In the reconstructed clause in (2b), the reader may notice that I have “lowered” the head of the relative in (2a), the quantifier כל‎, back into its position within the relative. Within the relative, כל‎ serves as the direct object (and the phrase אתו‎ is the indirect object). I have included the accusative את‎ on the front of כל‎ (which I also made definite) within the reconstructed clause simply to disambiguate its role as the accusative complement of צוה‎. This is almost certainly the grammatical option for the “case” of the כל‎. In other words, the כל‎ would not be preceded by the כ‎ that is on the head ככל‎ in the relative version, since that כ‎ is part of the matrix clause and not part of the relative; additionally, if צוה‎ required an object such as כל‎ to be in the oblique case, with some sort of preposition marking it as the object, like ב‎, the salient feature of ancient Hebrew relative clause formation is that the preposition required by the verb is included in the relative clause along with a resumptive pronoun, e.g., בו‎ (see Holmstedt 2002:90–107).

[28] For example, I suggest in “The Story of Ancient Hebrew ’ăšer” (f.c. ANES) that eleven cases of the relative word א‎שׁר‎ in the MT are not just infelicitous but actually ungrammatical (Gen. 11:7; 34:13; Deut. 4:10, 40; 6:3; 11:26–28; 1 Sam. 15:15; Ezek. 36:27; Qoh. 7:21; 33 Dan. 1:10; Neh. 2:3).

[29] Kiss, “Introduction,” pp. 2–3.

[30] See N. Chomsky, “On wh-movement” (Pp. 71–132 in Formal Syntax; ed. P. Culicover, T. Wasow, and A. Akmajian; New York: Academic Press, 1977); M. R. Baltin, “A Landing Site Theory of Movement Rules” (Linguistic Inquiry 13[1982]:1–38); H. Lasnik and M. Saito, Move α: Conditions on Its Application and Output (Cambridge: MIT Press, 1992); H. Maki, L. Kaiser, and M. Ochi, “Embedded Topicalisation in English and Japanese” (Lingua 107[1999]:1–14).

[31] On this issue, at least, subsequent research on left-dislocation in languages other than English demonstrated that it would have been accurate to modify what was known of the principles of left-dislocation based on biblical Hebrew: for instance, it has been shown that, like biblical Hebrew, other languages, such as Chicheŵa and Zulu, allow left-dislocation in non-root clauses. On Chicheŵa, see J. Bresnan and S. A. Mchombo, “Topic, Pronoun, and Agreement in Chicheŵa” (Language 63[1987]:741–782); on Zulu, see J. Zeller, “Left dislocation in Zulu” (unpublished ms.; [cited Sept. 20, 2006]).

[32] Analyzing Tiberian Hebrew phonology and then reconstructing behind it presents an entirely different set of challenges than reconstructing ancient Hebrew phonology. For Tiberian Hebrew, see J. L. Malone, Tiberian Hebrew Phonology (Winona Lake, IN: Eisenbrauns, 1993).

[33] “Methodological Issues,” p. 292.

[34] For example, see A. Sáenz-Badillos (A History of the Hebrew Language [Cambridge: Cambridge University Press, 1993], pp. 80–85) for an overview of the core research on the reconstruction of Hebrew phonology from the Greek and Latin transcriptions.

[35] This assumes that the Tiberian system of טְעָמִים‎ reflects the prosody of the biblical text as it was read from the ca. 5th century C.E. and after, and that it very well may not reflect anything close to the prosody of ancient Hebrew before ca. 200 C. E. For linguistic discussions of the Tiberian טְעָמִים‎, see M. Aronoff, “Orthography and Linguistic Theory: The Syntactic Basis of Masoretic Hebrew Punctuation” (Language 61 [1985]:28–72; B. E. Dresher, “The Prosodic Basis of the Tiberian Hebrew System of Accents” (Language 70/1[1994]:1–52).

[36] W. M. Schniedewind, “Prolegomena for the Sociolinguistics of Classical Hebrew” (JHS 5 [2004–2005]; article_36.pdf), §1.3; see also §2.11. On the point of writing versus vernacular, Schniedewind has recently repeated this claim: “By language, linguists refer to vernacular, not writing, and writing—especially in the ancient Near East—is certainly not primarily an attempt to transcribe vernacular” (“Aramaic, the Death of Written Hebrew, and Language Shift in the Persian Period” [Pp. 137–47 in Margins of Writing, Origins of Culture, The Oriental Institute Seminars 2, ed. S. Sanders; Chicago: The University of Chicago, 2006], p. 138).

[37] Particularly revealing in Schniedewind’s critique in “Prolegomena” is footnote 29, in which he suggests that R. P. Botha’s book, Challenging Chomsky: The Generative Garden Game (Oxford: Basil Blackwell, 1989) is a “general critique” of Chomskyan linguistics. However, this book is certainly not a critique of Chomsky’s framework, but a critical (and eminently enjoyable) defense of it. In fact, Botha’s contributions to the field of linguistics as a whole have been thoroughly Chomksyan.

[38] Ibid, §2.14–15.

[39] “[T]he term text [is used] to differentiate linguistic material (e.g., what is said, assuming a verbal channel) from the environment in which “sayings” (or other linguistic productions) occur (context). In terms of utterances, then, “text” is the linguistic content: the stable semantic meanings of words, expressions, and sentences, but not the inferences available to hearers depending upon the contexts in which words, expressions, and sentences are used” (D. Schiffrin, Approaches to Discourse [Blackwell Textbooks in Linguistics; Oxford: Blackwell, 1994], p. 363).

[40] E. Ullendorff, “Is Biblical Hebrew a Language?” (Pp. 3–17 in Is Biblical Hebrew a Language? Studies in Semitic Languages and Civilizations; Wiesbaden: Otto Harrassowitz, 1977), p. 16.

[41] “[A] language is a state of the faculty of language, an I-language, in technical usage” (N. Chomsky, “Three Factors in Language Design” [Linguistic Inquiry 36/1(2005):1–22], p. 2).

[42] “Methodological Issues.”

[43] Putting aside the numerous lexical items that Ullendorff expects but does not find in the lexical inventory of biblical Hebrew, more salient to the issue of a linguistic system are the supposed “grammatical gaps” that he adduces: the lack of many 2nd person feminine forms, of certain Hof’al forms, and of certain types of clauses, and a “dearth of genuine dialogue features, of anacoluthon, and especially of non-literary...sentence structure” (“Is Biblical Hebrew a Language?,” p. 14). See Miller (“Methodological Issues,” pp. 287 n. 24, and pp. 293–96) for a response to Ullendorff.

[44] The caveats notwithstanding, the fact is that grammars present data that might reflect alternate linguistic systems of ancient Hebrew as marginalia. Even if we refer to ancient or biblical Hebrew as a bona fide “language” in the abstract, if we do have compelling evidence of historical stages, dialects, and/or registers, each deserves its own full-scale grammatical description.

[45] Among other of Rendsburg’s works, see “The Strata of Biblical Hebrew” (JNSL 17[1991]:81–99); “Morphological Evidence for Regional Dialects in Ancient Hebrew” (Pp. 65–88 in Linguistics and Biblical Hebrew, ed. W. R. Bodine; Winona Lake, IN: Eisenbrauns, 1992); “Notes on Israelian Hebrew (I)” (Pp. 255–58 in Michael: Historical, Epigraphical and Biblical Studies in Honor of Prof. Michael Heltzer, ed. Y. Avishur and R. Deutsch. Tel Aviv: Archaeological Center Publications, 1999); “Notes on Israelian Hebrew (II)” (JNSL 26[2000]:33–45).

[46] G. A. Rendsburg, Diglossia in Ancient Hebrew (New Haven, CN: American Oriental Society, 1990); F. Polak, “Style is More Important than the Person: Sociolinguistics, Literary Culture, and the Distinction between Written and Oral Narrative” (Pp. 38–103 in Biblical Hebrew: Studies in Chronology and Typology, ed. I. Young; London: T&T Clark, 2003); idem, “Sociolinguistics and the Judean Speech Community in the Achaemenid Empire” (Pp. 589–628 in Judah and the Judeans in the Persian Period, ed. O. Lipschits and M. Oeming; Winona Lake, IN: Eisenbrauns, 2006).

[47] I. Young, “The Style of the Gezer Calendar and Some ‘Archaic Biblical Hebrew’ Passages” (VT 42/3[1992]:362–375); idem, Diversity in Pre-Exilic Hebrew (Tübingen: J. C. B. Mohr, 1993); idem, “The ‘Northernisms’ of the Israelite Narratives in Kings” (ZAH 8/1[1995]:63–70); idem, “The ‘Archaic’ Poetry of the Pentateuch in the MT, Samaritan Pentateuch and 4QExodc.” (Abr-Nahrain 35[1998]:74–75); idem, “Late Biblical Hebrew and Hebrew Inscriptions” (Pp. 276–311 in Biblical Hebrew: Studies in Chronology and Typology, ed. I. Young; London: T&T Clark, 2003); idem, “Biblical Texts Cannot be Dated Linguistically” (HS 46[2005]:341–51); idem, “Late Biblical Hebrew and the Qumran Pesher Habakkuk” (unpublished ms).

[48] G. A. Rendsburg, “Confused Language as a Deliberate Literary Device in Biblical Hebrew Narrative” (JHS 2 [1998–99]; article_12.pdf); Young, Diversity, for instance, on Qoheleth, p. 157; R. D. Holmstedt, “The Distribution of še and ăšer in Qoheleth” (Paper presented at the annual meeting of the SBL, Washington, D.C., Nov. 19, 2006). Similarly, see W. M. Schniedewind, “Qumran Hebrew as an Antilanguage” (JBL 118/2[1999]:235–252).

[49] Representative of this near consensus view is P. Joüon and T. Muraoka, A Grammar of Biblical Hebrew (Rome: Editrice Pontificio Istituto Biblico, 1993), pp. 118–19.

[50] Rendsburg summarizes (and adopts a modification of) this position in Diglossia (pp. 116–17):

During the period of the monarchy, 1000–586 B.C.E., a standard literary Hebrew was utilized in which רשׁא‎ was the sole relative pronoun. The colloquial form, which existed side-by-side with the classical form, was שׁ‎, which in a very few instances infiltrated literary composition. The upheaval of 586 B.C.E., with the resultant exile and restoration, effected changes in the Hebrew language, and one of these was the further penetration of שׁ‎ into written records. (1990:116–17)

[51] Rendsburg, Diglossia, p. 123, n. 29. Note Young’s cogent objection to the acceptance of the שׁ‎ in 2 Kgs 6.11 as an actual Aramaism (“Northernisms,” pp. 65–66):

We do not know of this word in our Aramaic sources at all. We must therefore raise the possibility that beside genuine foreign and dialectal forms, the Hebrew author could also draw on a body of cliched “non-standard” forms. The draw some modern analogies, while parodies of foreign or dialectal speech will utilize certain language features which are felt to be absolutely characteristic of the target of the parody..., other accent features used will be from the general category of “funny speech,” which is built from a mishmash of many different varieties of “non-standard” language.

[52] I am indebted to W. Dennis Tucker, Jr. in my understanding of Jon 4.10.

[53] Jackie Naudé’s comments on“idiolects” and “the notion of language” are particularly relevant here (“The Transitions of Biblical Hebrew in the Perspective of Language Change and Diffusion” [Pp. 189–214 in Biblical Hebrew: Studies in Chronology and Typology, ed I. Young; London: T&T Clark, 2003], pp. pp. 196–97).

[54] See E. A. Knauf, “War ‚Biblisch-Hebräisch‛ eine Sprache?—Empirische Gesichtspunkte zur linguistischen Annäherung an die Sprache der althebräischen Literatur.” ZAH 3/1(1990):11–23.

[55] J. Klausner, דקדוק קצר שׁל העברית החדשׁה‎ [A Short Grammar of Modern Hebrew] (2nd ed.; Tel Aviv: Dvir, 1938).

[56] A. Bendavid, ?ון חכמיםשׁון המקרא או לשׁל‎ [Biblical or Rabbinic Hebrew?] (Tel Aviv: Dvir, 1951); idem, ון חכמיםשׁון מקרא ולשׁל‎ [Biblical and Rabbinic Hebrew] (Tel-Aviv: Dvir, 1967). Note the change of conjunctions in the title, as well as the change from interrogative to indicative; this reflects Bendavid’s shift from a prescriptive stance to a descriptive one.

[57] P. Wexler, On the Schizoid Nature of Modern Hebrew: A Slavic Language in Search of a Semitic Past (Wiesbaden: Harrassowitz, 1990).

[58] N. H. Tur-Sinai, ון העברית ובעיותיהשׁמתחית הל‎ [The Revival of Hebrew and its Problems] (Leshonenu 17 [1951]:29–36).

[59] S. Izre’el, “The Emergence of Spoken Israeli Hebrew” (Pp. 85–104 in Corpus Linguistics and Modern Hebrew: Towards the Compilation of the Corpus of Spoken Israeli Hebrew (CoSIH), ed. B. H. Hary; Tel Aviv: Tel Aviv University, 2003), pp. 85–87.

[60] Ibid, 86.

[61] Y. Peretz, ון העברי לכל תקופותיהשׁפת הזיקה בלשׁם‎ [The Relative Clause in Hebrew in All its Stages] (Tel Aviv: Dvir, 1967).

[62] R. D. Holmstedt, The Relative Clause in Biblical Hebrew: A Linguistic Analysis (Ph.D. dissertation, University of Wisconsin-Madison, 2002); idem, “Word Order in the Book of Proverbs”; idem, “Topic and Focus in Biblical Hebrew if it is an SV Language” (Paper presented at the annual meeting of the Society of Biblical Literature, Philadelphia, Nov. 20, 2005).

[63] J. A. Cook, The Biblical Hebrew Verbal System: A Grammaticalization Approach (Ph.D. dissertation, University of Wisconsin-Madison, 2002;) idem, “The Finite Verbal Forms in Biblical Hebrew Do Express Aspect” (f.c. JANES).