Friday, February 20, 2009

Hebrew Consecutives, part 2

In the previous post I described the weird phenomenon of the so called consecutive tenses of Classical Hebrew - the strange fact that the word for "and" seems to make the two Hebrew quasi-tenses switch meaning with each other. At the end of that post I promised a historical explanation for this somewhat outlandish state of affairs, and one does not like to disappoint, so here goes.

In the linguistic stages before Classical Hebrew, the verbal system was somewhat different. We know this from older North West Semitic languages, most notably Ugaritic, which was spoken and written in modern day Syria during the latter half of the second millennium BC.

At this earlier time, the verbal system in North West Semitic (of which Hebrew is a part) seems to have looked something like the following. There were two tenses which expressed true verbal actions, both of them created using prefixes (and some suffixes). If one uses the standard example verb q-t-l (meaning "to kill" - that's a weird piece of grammatical morbidity for you), we get the forms yaqtul and yaqtulu in the 3rd person singular masculine. These two forms differed somewhat in meaning: the longer form yaqtulu seems to have been used for actions of duration, for present and future actions. The shorter one, yaqtul, was used for narrative purposes, for the past, but also for exhortations and orders.

There was also a suffix form, qatala, which originally had stative meaning (it marked a state rather than an action). After a while it also started to be used for actions, often (but not always) in the past, actions which were known to be true with high cerntainty, background facts, etc.

Then came a great reorganization of the system. The final vowels of Hebrew were dropped, so that the two prefix forms fell together: yaqtul and yaqtulu both fell together as yiqtol, the form mostly called "imperfect" in Hebrew grammar. The difference between present-future and past form became obscured in most (but not all) verbs. And the qatala form was increasingly used for past time in the form qatal, often called "perfect". Later on this temporal separation became full fledged, which led to the fact that the imperfect and perfect forms are usually regarded as "future" and "past" in the grammar of Modern Hebrew.

But in the classical language, the old narrative yaqtul-form lingered on. It remained as an old and fossilized relic - but normally only in one single environment: when preceded by the word for "and" (we- or wa-). This gave rise to the wayyiqtol form, the most common form of classical Hebrew narrative. A form of the "impefect" (present/future) was suddenly used for past time. It must have seemed strange indeed to the mediaeval Jewish scholars for whom this form was nothing but an old weirdness.

The reason for the perfect getting a "switch" of its own from past to future was completely different. That came to pass because of the original use of the perfect form: talking about certainties, emphatic utterances and states. This usage meant that the perfect often occurred in the second half of "if-then" type clauses (oy! I'm sounding like a programmer here!). The certain result of a condition would stand in the perfect form, often preceded by we-. which provided sequence. This usage can be found in Ugaritic, as well as in the letters from Bronze Age El Amarna. But again, this surviving construction appeared to be completely "backwards" to traditional Hebrew grammar. Perfect for the future? Weird.

And so both "tenses" had seemingly been switched. But that was never what really happened; it only seemed that way to those for whom the system was no longer living.


Martin said...

"Yaqtulu"? Iä! Cthulhu!

Ola Wikander said...

Cthulhu fhthagn!

Unknown said...

Hi Ola,

Based on what you wrote, I would like to propose a hypothesis concerning the application of Vav Consecutive Imperfect (VCI) and Vave Consecutive Perfect (VCP).

In English, the Present Tense is: I am writing. However English does not properly ascribe a tense to the verb: I write. The difference is that "I am writing" is what I am doing at this moment, but "I write" is a statement of truth or fact and it is timeless, because it is true of my adult abilities yesterday, today and tomorrow, until something happens to make it no long true. Hence "I write" is actually the Indefinite Tense -- indefinite with respect to time.

The Indefinite tense is also used in narratives -- and in telling jokes in particular: a man walks in to the bar and says ... . "Walks" and "says" are indefinite with respect to time -- they assert the action as true, irrespective of when it takes place.

My hypothesis is that the VCI and the VCP mark the change to the Indefinite tense in Hebrew, with the VCI relating events that are the temporal or logical SEQUEL to what has preceded and the VCP relating events that are the temporal or logical CONSEQUENCE of what has preceded. Hence the VCI should be translated using expressions such as:

temporal: and then
temporal: and next
temporal: and at that time
logical: and subsequently.

The VCP should be translated using expressions such as:

temporal: and consequently
temporal: and the time comes (when/that)
temporal: and thus
temporal: and it happens that
logical: and therefore
logical: and hence
logical: and it follows that
logical: and wherefore.

This hypothesis removes the common ambiguity of "switching tenses" and it addresses some of the "aspectual (sort of)" aspects of your comments about the Hebrew tenses. It appears to align somewhat with your comments about qatala having a stative meaning that became to be used of actions which were known to be true with high certainty, background facts, etc. And it fits with your comments about the wayyiqtol form being the most common form of classical Hebrew narrative (which is a common role of the Indefinite Tense in English). It seems to me your comments that the perfect often occurred in the second half of "if-then" type clauses, which preceded by the we- provided sequence, ties it together. Hence my hypothesis.

Does it seem reasonable to you?

Unknown said...
This comment has been removed by the author.
Unknown said...

Hi Ola,

(updated - previous version had the VCI and VCP back-to-front).

Following the last post, I have formulated a postulate that I will try to test over the next few months, as time permits and update this blog is due course. Any feedback and suggestions are most welcome.

To test the postulate, I will use the parsing provided by Beall and Banks, "Old Testament Parsing Guide", starting at Gen 1:1. The plan is to use an Interlinear translation and translate the VCPs and VCIs in the following manner.

1. If the verb that introduces a VCI/VCP is followed by a VCI, the VCI is expected to express the temporal or logical SEQUAL of that action.

2. If the verb that introduces a VCI/VCP is followed by a VCP, the VCP is expected to express the temporal or logical CONSEQUENCE of that action.

3. The VCI/VCP verb will translate in the Indefinite Tense (that is, timeless, such as: I write, he says, it introduces, it presents).

It is apparent from Beall and Banks that the VCIs and VCPs can be intermingled and while generally the VCI follows a Perfect and a VCP follows an Imperfect this is not always the case. Hence the expectation is that it will always depend on the context of the discussion and the context of the introductory verb as to whether the context of the following verb is a temporal/logical sequential event or a temporal/logical consequence.

Irrespective of whether the clause under examination presents a sequence or a consequence, the next clause can contain another VCP or VCI, and the same rules will apply.

Beall and Banks show that sometimes a number of sequential clauses (VCP verbs) can follow an introductory verb and then a consequential clause (VCI verb) will appear, or vice versa. When this happens, the process will be to compare the keywords in the new clause with the keywords in the introductory clause to determine whether the new clause has introduced a new thread that stems from the introductory clause or the new clause follows on from the preceding VCI/VCP.

(If this discussion is off-topic for your blog, I won't be offended if you remove it, but I would value any feedback or guidance to rob at lasotell dot com dot au. Thank you.)

Ola Wikander said...

Rob - I hadn't seen your comments before (which was very negligent of me). I will return to you with some (hopefully intelligent) comments later this week!

Unknown said...

Hi Ola,

The preliminary Vav Consecutive analysis work below arose because I was asked to justify a statement I had made that Isaac was Abraham's eighth son, given that Gen 25:1 shows sons born after Sarah had died. To answer the question objectively, I wanted to show first of all that the validity of the detailed genealogies in 1 Chronicles 1 could be shown independently of my statements concerning Abraham descendants based those same genealogies. This meant I had to show the correct order of the birth of Noah's sons. Once that was established, then the Vav Consecutive analysis of the text leading up to Gen 25:1 is seen to be supported by the 1 Chronicles 1 genealogies, and thereby giving me two proofs that Isaac is Abraham's eight son.

The results of that work are in the following link:


Unknown said...

Hi Ola,

I have updated the document in the last post -- it had some dopey WCI instead of VCI typos. (Yes, I am mildly dyslexic.) Sorry about that.