The Verbal System of Ancient Hebrew: A Postscript

A few weeks ago, this blog hosted a lively discussion about tense, mood, and aspect in the verbal system of ancient Hebrew. A number of top researchers in the field participated in the debate, and clarified their points of view in the process. The exchange is informal in style, and would be hard for those concerned to revise according to the more formal conventions of peer-reviewed publications in the field. This became clear in follow-up correspondence. The exchange, therefore, will remain as is, but I thank those who suggested that it was worthy of publication elsewhere, and Ehud Ben-Zvi of JHS for his willingness to consider publication there. For the convenience of future readers, Here is the whole thing as a single downloadable file with an attached cumulative bibliography.

The debate, of course, will go on. Peter Bekins over at his great blog posted a brief summary of his take on the system in the course of reviewing an article by John Cook that has just appeared in JHS. A strength of his summary is the attempt to describe the system historically and cross-linguistically, against the background of features found in the verbal systems of cognate languages. Below the fold, I briefly describe the system “synchronically,” with a view to what it is becoming, rather than with a view to what it looked like – allegedly – “in origin.”

As I’ve thought through the issues again, I find myself agreeing with Joosten’s take on things more and more. Aside from a couple of idiosyncrasies, the following paragraphs, if I’m not mistaken, are compatible with his views.

I simplify matters by not trying to determine what a particular form “marks” as opposed to what it “expresses.” I limit myself to describing what a form can express.

Expressing Tense in Ancient Hebrew

If the communication requires future temporal reference: in a majority of cases, yiqtol or weqatal is used. The two often appear together in specific configurations.

If the communication requires past temporal reference: wayyiqtol or qatal is used. The two often appear together in specific configurations.

If the communication requires present temporal reference: qotel is used. In the case of stative and a few other verbs, and in some performative utterances, qatal is used. In questions, yiqtol may be used to express present temporal reference.

Expressing Modality in Ancient Hebrew

A variety of forms may be used to express modality: yiqtol, the cohortative-jussive-imperative subsystem, and energic forms. In contrast, qatal may be used to express “moodlessness” or omnitemporality: this usage is equivalent to the “gnomic” aorist in Greek.

Looked at from within the system, the use of yiqtol to express the indicative future (= “will” in English) and iterativity ( = “would” in English) are to be understood as examples of modality.

As Joosten argues, periphrastic haya/yihyeh + qotel is used to express durativity (= “was /will be . . . –ing” in English).

Expressing Aspect in Ancient Hebrew

The yiqtol / qatal contrast in ancient Hebrew does not express a contrast in aspect. As Joosten points out, yiqtol often has a nuance of incompleteness, but in other cases, it has a punctual, complete, or aoristic nuance. As for qatal, it regularly refers to events as complete, but that is what one typically does in narrating past events. On the other hand, it cannot be used to refer to future events which have a punctual or complete aspect.

In later Hebrew, wayyiqtol and weqatal tend to disappear. Concurrently, yiqtol / qatal / qotel come to express a full-blown three-way tense contrast, and periphrastic “tenses” are used more regularly.

Comments

[Vince DeCaen]

Dear John,

You wrote, “I find myself agreeing with Joosten’s take … ancient Hebrew does not express a contrast in aspect”—to which I would say, yes, of course.

Since I find nothing to critique in your summary, I will add some thoughts which weren’t highlighted in the original exchange.

(A) re tense vs. aspect:

I draw your attention to an amazing article by Matthew Anstey in JSS (2006), of the I-wish-had-thought-of-it variety. As background to his study of Hebrew clitics, he outlines a tagging scheme to bring Biblical Hebrew into the modern period (!). He tags the finite forms “past” and “nonpast”, and the qotel participle as “progressive” aspect.

When I saw this, I immediately thought two things: (i) how did he get that by the referees without at least having to offer a footnote? and (ii) has common sense finally prevailed in our benighted field? I use the term “benighted” advisedly in this context.

Let me give you a little autobiographical background by way of explanation:

I came to Hebrew after East European Studies and Slavic Linguistics. My Hebrew instructor was an authoritarian who insisted on chapter and verse of received dogma as embedded in his homemade textbook. When he said Hebrew aspect worked just like Russian, I pointed out that that simply was not true (Russian, Polish and Hungarian, e.g., do not work the way they are generally described in Semitic studies). When he said the tenselessness of Hebrew reflected the “Hebrew mind” which worked differently from the (implicitly superior, rational) “Greek mind”, I gasped in horror. When he said Hebrew was VSO, I also pointed out that that clearly could not be the case. Alas, I was wasting my time, and almost gave up on Hebrew studies at the time.

I have since explored the history of Biblical Hebrew’s tenselessness, and have come to the following conclusions. The notion is derived from questionable speculations on the prehistory of the Indo-European verbal system: specifically, that there was a “primitive” stage before the Aryans somehow stumbled upon tense. This “primitive tenselessness” is then posited as the model for the “arrested” non-Aryan languages and their “arrested” minds. I have since come to understand that this is simply part of the larger discourse of Orientalism, explored famously by Edward Said.

I reviewed the basics in “Reconsidering the Aspectual Analysis of Biblical Hebrew,” Society of Biblical Literature (SBL 1995), Linguistics and Biblical Hebrew Section, Philadelphia, and “Ewald and Driver on Biblical Hebrew ‘Aspect’: Anteriority and the Orientalist Framework,” Zeitschrift für Althebraistik (1996) 9.2: 129-151.

I subsequently studied this history in more detail, and reported my findings in “The Invention of Hebrew Tenselessness, Berlin 1826: Race, Mind and Language,” Symposium paper, University of Toronto, Jewish Studies Faculty Symposia, 10 February 2000.

(Joosten has explored the work of the towering genius Kurylowicz with similar results.)

In short, I consider formal tenselessness to be an extension of Orientalist discourse that would best be left behind. (I have a collection of quotations on the Japanese, Chinese, Arab and Hebrew “mind” that I have culled from grammars that are in shockingly bad taste. Alas.) I consider the proposal of a +/-tense parameter among the world’s languages to be an updating of this original speculation of the 1800s. (I think a review of current elementary grammars in this regard would be in order.)

I believe in strong claims regarding Universal Grammar. I believe that differences between languages, like those between races, are exceedingly superficial. In this spirit, I think we would be better off adopting a well-motivated theory of an “aspectual default” as the basic parameter instead. On this view, languages differ only in which aspect they mark grammatically: imperfective or perfective. The unmarked forms are interpreted by default, which requires a robust pragmatics component in tense-aspect systems. On this view, it is this defaulting that gives rise to Buth’s “mixed” tense-aspect.

My view makes the differences between Biblical Hebrew and later forms relatively minor: just a tweaking of the system, with some “bleaching” of the progressive participle.

(B) Let me address two points that you raise in your summary.

(1) The business about periphrasis is easy to lose sight of because of the dominance of narrative wayyiqtol. I explored the problem in “Where’s the Rest of the Biblical Hebrew Verbal System?” Toronto School of Theology Colloquium Series, 30 November 1995. I compared the situation to a grammar of English that had only two forms, walk(s) and walked: where is the rest of verbal system? In a word, periphrasis.

I raised the point again in “A Unified Analysis of Verbal and Verbless Clauses within Government-Binding Theory,” pp. 109-131 in The Verbless Clause in Biblical Hebrew: Linguistic Approaches, edited by Cynthia L. Miller, Linguistic Studies in Ancient West Semitic, vol. 1 (Winona Lake IN: Eisenbrauns, 1999). The preferred analysis of haya/yihye + qotel is a dummy tense-marking auxiliary plus the progressive participle, not unlike the English construction, as you yourself point out. The difficulty for a tenseless analysis here seems to me insuperable IMHO.

In passing, I note that crosslinguistically the progressive participle is paired with a perfect participle in such systems, and I would suggest that qatul as the passive-perfect should accordingly be added to round out the system. (If you’re interested in grammaticalization, btw, the history of the periphrastic passive-perfect turning into a past tense in Aramaic/Syriac systems is fascinating, and is a nifty analogy for the prehistory of the Hebrew qatal.)

(2) You mention modal yiqtol without mentioning word order. Modal yiqtol cannot be primarily about morphology, however, since most verbs simply do not formally mark the modal/indicative distinction.

Rather, the primary distinction must be made via word order for the majority of verb forms, specifically verb movement, which is the very essence of my original syntactic analysis—together with generalized topicalization, with obvious discourse-analytic implications (“On the Placement and Interpretation of the Verb in Standard Biblical Hebrew Prose.” Ph.D. diss., University of Toronto, 1995).

I raised this observation to a strong claim about syntax (verb-initial vs verb-second) and modality (modal vs indicative, respectively), and this sent me down the road to “modal coordination” as the unified analysis of the so-called consecutives.

In this light, Cook is unforgivably inaccurate when he writes in passing, “Joosten (1992) and DeCaen (1995) also identify weqatal as modal; however, they do not posit the broader hypothesis developed by Holmstedt ... and Cook ..., namely, that the indicative-non-indicative modal distinction throughout Biblical Hebrew is consistently marked by word order: subject-verb versus verb-subject” (JSS 2004 p. 265). My mind still staggers at this footnote. Sigh.

At any rate, I have explored modal coordination in crosslinguistic perspective in “Distinctive Properties of the Biblical Hebrew Consecutives in Crosslinguistic Perspective: Modal Coordination in Ancient Egyptian, Fula, Swahili and Zulu,” Niagara Linguistic Society (NLS99), State University of New York at Buffalo, 26 September 1999.

I have developed the formal semantics of this proposal in my study “On the Semantics of Modal Coordination in Biblical Hebrew: A Strictly Compositional Approach to Truth-Functional Operators for Natural Language.” The idea is that wayyiqtol has an “and then” value in contrast to the “if then” version of weqatal; crucially, both have the operator “then” in their formal representations. I’m still working on the theoretical-literature review, but hope to have a submission draft this summer with the assistance of Bob Binnick, here in the linguistics department. (The rough draft is in my papers directory http://www.chass.utoronto.ca/~decaen/papers for comment.)

Shalom,
Vince

[JohnFH]

Hi Vince,

It's great to hear you are plowing ahead with the development of your theses. In a future post, I will draw attention to bibliography that will interest you.

Are all of the papers you mention available on your website? There are quite a few you mention I wasn't aware of. Furthermore, don't you have the rudiments of a grammar available in which your analysis is presented in a less allusive style, with examples and all? Here's hoping.

As for the agreements which exist between Revell and DeCaen (maybe I should add my name here; I am one of Revell's students after all), Joosten, Holmstedt, and Cook with respect to the importance of word order in the verbal system, it would be really nice if someone like Peter Bekins wrote up the history of research on this.

We always tend to concentrate on disagreements. It's just as helpful, I think, to identify areas of agreement.

[Vince DeCaen]

Dear John,

I think a third-party review and summary would be a good idea.

However, I think all parties would agree that the syntax insight should be attributed to Paul Jouon and his 1923 Grammaire de l'hébreu biblique (the insights were lost in translation, and revised out of existence by T Muraoka, A grammar of biblical Hebrew Rev. English ed. (2006) fyi).

All I really did was formalize that insight within the then-current generative framework.

Cheers!
V

[Doris]

The research of ancient Hebrew is particularly important for the study of Kabbalah and other areas of Judaism as part of the understanding is depended on the right understanding of the language. I even recommended the blog to my readers

[JohnFH]

Thanks, Doris. What an interesting blog you have. The graphics are fabulous.

[Hansen]

I'd like to ask a question of practical value: When doing word studies, is it always correct to assign the same definition to a word regardless of the form of the word in various passages?

For instance, if I am trying to understand the range of the Hebrew word for "justify", is it safe to let every passage, regardless of the form of the word, contribute to the definition of the word?

Another way of asking the question is: Does the form of a word change its meaning or just add nuances to a single definition?