You've asked a good and interesting question of an extremely rare phenomenon and so a very difficult verse.
Overwhelmingly (and I mean, extremely) when a ויהי stands before a prepositional phrase, the ויהי is what we call a "discourse" ויהי and the prepositional phrase is actually part of the following clause, as a fronted scene-setting Topic. The main verb to which the fronted PP is an adjunct is typically another wayiqtol (e.g., Gen 4:3 with ויבא) or sometimes a Qatal or Yiqtol (e.g., Gen 7:10 with היו).
Here in Exod 12:41 is an almost unique example (the only somewhat similar example in the entire Hebrew Bible is an "extra" והיו in Exod 4:9, though I didn't tag that one as "unknown"). In this case there is no finite verbal clause / main verb coming after the PP. Rather, the PP is followed by a second ויהי, after which a second PP and then (!) the finite verb יצאו appears. That is why I have *for now* marked the second ויהי as "unknown". Hopefully I will sort out a grammatical solution.
Unfortunately, I do not think parallelism contributes to a grammatical solution. First, "parallelism" is not a grammatical concept. It is a literary notion that I don't actually consider all that useful (I have a forthcoming article on this in Vetus Testamentum; I've posted the draft on my Academia.edu site for anyone who is interested). Whatever one's view of parallelism, what you're suggesting would only fit one of two grammatical phenomena: 1) anacoluthon or right-node raising.
The former would mean a complete interruption of the syntax in the first line, whereas the second line is complete. I would like to see examples of this from poetry, because I am suspicious about its existence (not anacoluthon in general, but its use like this in poetry).
The right-node raising option would mean that the complement ("object") of the first verb is moved rightward out of the clause to a position attached to the second clause from which it satisfies the valency of both verbs (an example in English is "Jack saw and Jill heard the burglar
downstairs"). The problem with the RNR analysis is the complexity of the verse. If we take the ויהי verbs as discourse, as we have done, then the second ויהי makes too many clausal structures for RNR to work. If we don't take the ויהי as discourse level but as the main verbs in each clause, then there are three distinct clauses, unless we downgrade the Qatal
יצאו to an unmarked subordinate clause, which seems arbitrary and doesn't really make any sense in the context.
And if there are three main clauses, we must ask: what are the complements of each and what is "missing"? The verb היה can take a PP complement, which means that the natural interpretation of the first clause would be "And it was at the end of 430 years". Ok, that's grammatical, though actually not very common. The second clause becomes "And it was on that very day". Same thing. Now, interestingly, this could be a case of clausal apposition, by which the second clause reformulates the first clause for clarity. But then notice that the third clause begins with the Qatal verb, which is highly unusual syntax, though grammatical and when it occurs is almost always the irrealis use of the Qatal
, often as a future potential "should/shall," but that certainly doesn't fit the context here: "all the hosts of the Lord went out [past time reference] from the land of Egypt".
So, I can get you to your "parallelism" with a grammatical analysis, but I am left with a use of the Qatal
verb in the third clause, which, when I look through the rest of Exodus for realis past time uses of a Qatal
in the first position of the clause (i.e., before the subject), becomes clearly anomalous. I found 11 occurrences in Exodus: Ex 14:3; 15:8, 9, 10, 14, 15; 22:23; 27:7 (2x); 35:30; 37:9. Of these, 22:23, 27:7 (2x), 37:9 are clearly irrealis Qatal
("X shall ...") and arguably this is also the case fo 14:3 and 35:30 (the latter being the performative use of an irrealis Qatal
). It is fascinating that the other five are in the poetry of Exod 15 (which contrary to some I still take to be very old poetry and the VS order would support this, since I have argued that Hebrew began shifting from VS to SV during the monarchic period, much earlier than Late BH).
I hope this at least explains why I used the "unknown" tag temporarily at this particular point. Normally I can at least see a grammatical solution; I can't see one here yet and remain open to a textual corruption.
Edited by Robert Holmstedt, 30 April 2018 - 09:18 AM.