New John Cook article in JHS
John Cook has a new article in the on-line Journal of Hebrew Scriptures: Cook, John A., “The vav-prefixed verb forms in elementary Hebrew grammar,” JHS Volume 8: Article 3 (2008). I won’t formally summarize the article for my bibliography since it is easily accessed (I love the idea of an on-line, peer-reviewed journal, and I am very happy that JHS has become a successful endeavor).
This article was originally presented as a paper for the National Association of Professors of Hebrew, and it was occasioned by the great increase in the number of elementary Hebrew grammars that have perpetuated the enigmatic description of vav-prefixed verbs as “vav-conversive” or “vav-consecutive”. Cook argues that such a description leaves students with the feeling that Biblical Hebrew is a “strange beast without any parallel among human languages.” The value of a better, linguistically informed, description is that students can appreciate how BH works as a human language rather then seeing it as some ancient puzzle.
The first part of the paper surveys some of these descriptions, but the second part is the most interesting as he gives some examples from his grammar (co-authored with Robert Holmstedt and available in a preliminary form here).
In previous discussions of the BH Verbal System I have tended to agree with John’s typological/grammaticalization approach. The core opposition of the BH system is the binary morphological opposition of a prefixed verbal form (yiqtol) and a suffixed verbal form (qatal). This opposition originally marked aspect (imperfect/perfect), thus typologically it is best to call it an aspectual system (of course, this can be a misleading label as Randall Buth pointed out that BH does not appear to be an “aspectually sensitive” language, especially in comparison to Greek whose morphology allows the variation of tense, mood, and aspect somewhat independently of each other). From a typological comparison of other such languages, a general trend can be seen of the perfect form also defaulting as a simple past tense, and the imperfect form defaulting as present/future. Over time, as new aspectual forms develop, the older forms begin to be used solely to express tense, and eventually they fall out of use altogether.
At this point, I think it is pretty widely accepted that the prefixed yiqtol form in wayyiqtol is not the same as the imperfect, but rather preserves an older prefixed preterite form that has fallen out of general use. This is supported by the existence of the prefixed preterite iprus in Akkadian, the apparent Canaanite reflexes in the Amarna letters, the negative past tense construction in Arabic, la yaktub, etc. In biblical narrative, the preterite is preserved as the special narrative tense wayyiqtol, and it also appears without the prefixed vav as a preterite yiqtol in classical poetry.
What was interesting to me in this article was the description of waw + suffix conjugation, weqatal. It is common among Semitic languages for the perfect form to also be used to express non-past/modal statements. (Not to confuse things at this point, but if indeed there is a trend for perfect forms to move to preterite forms over time, then perhaps the prefix “jussive” forms which occur both in Hebrew and Akkadian, ie the precative liprus, spun off from the old prefixed preterite at an earlier stage when it was actually a perfect?) Nevertheless, returning to weqatal, EJ Revell has shown that in the prefixed verbal forms, indicative and modal expressions can be distinguished by word order – in modal statements, the verbs consistently appear at the head of the clause. Vincent DeCaen has thus suggested that the vav-prefixed forms, being intrinsically verb-first, are in fact modal conjugations. This explanation does not seem to fit wayyiqtol, but the distribution of weqatal suggests that it is in fact “modal”. Cook suggests this non-indicative use most often marks subordinate clauses.
The pedagogical insight is that wayyiqtol and weqatal should not be treated as a single class of “vav-prefixed” verbal forms. Rather, they are quite distinct from each other. It is easiest to understand wayyiqtol as a special narrative tense, preserving an older preterite form. On the other hand, weqatal is a non-indicative/modal form which derives quite expectedly from the perfect qatal and which is differentiated from indicative statements by its clause-initial position.