Shmuel asked whether metegs are consistent across xumashim, and Jeremy answered that it shouldn't even be expected.
But shouldn't we expect consistency
within any given xumash/Tana"kh across multiple occurrence of the same word (and of words that are similar enough to seem analogous in this respect)?
This is a good opportunity for me to report on some "research" that I did two (or perhaps three) years ago, which revealed a shocking (to me) inconsistency in the Qoren Tana"kh in the sense that it treats certain words in Torah much differently from the way it treats those same words in Na"kh. If I didn't know any better, I would suspect that the Torah section and the Na"kh section had separate editors, and they were getting their texts from two very different sources. I'd welcome anyone's comments.
Ironically, I embarked on this "research" due to something I observed in Parshat Eiqev, which we just read last Shabbat, Specifically, I noticed (in both Tiqun Simanim and the Qoren Tana"kh) that there is a meteg under the tav in tixyun (8:1) but no meteg under the yud in either occurrence of yixye (8:3).
Before that, I was under the impression that a meteg always (or almost always) appears in any form of the verbs hei-yud-hei (to be) and xet-yud-hei (to live) that satisfies the following two conditions:
(a) The first letter of the shoresh (hei or xet) has a sh"va under it.
(b) The previous letter (which is not necessarily the first letter of the word) has a xiriq or segol under it. It is this letter which has the presumed meteg.
In the example cited above, this was true for tixyun but not for yixye --- and I wondered whether yixye was a rare exception or an example of a more widely occurring phenomenon that I was unaware of.
This intrigued me enough to spend a fair amount of time (perhaps 8-10 hours) using a Concordance to find every occurrence of every word that satisfies (a) and (b) in Tana"kh and look it up in the Qoren Tana"kh to see whether or not it has the expected meteg. There are 1210 such words in Tana"kh --- 1124 for hei-yud-hei and 86 for xet-yud-hei. In a small number (perhaps 3 or 4) of these cases, the letter in question has an actual trope (due to nasog axor) rather than a meteg. For obvious reasons, I counted the meteg as being present rather than absent in these cases. I emphasize that I was using only the Qoren Tana"kh and made no attempt to compare it to any other published Tana"kh.
Before I talk about the results of this endeavor, I'll mention a trivial tidbit that I discovered as part of the process and found noteworthy. There are 8 total occurrence in Tana"kh of the words sheyihye (5) and sheyihyu (3) --- and all 8 of them are in Qohelet.
The attached Excel file contains the results of my "research" in all of its gory details for anyone who is interested, but here is an executive summary.
1. In Na"kh, with very few exceptions, the expected meteg always appeared.
(a) For the verb hei-yud-hei, there are only 5 cases out of 683 (0.7%) where the meteg fails to appear. Three out of the 5 exceptions are in the word nihy"ta, and there might be a reason to suspect that they are not exceptions at all --- the attached file has the details. One out of the remaining two exceptions actually has a meteg, but under the letter preceding the one where I expected it --- again, details are in the attached file. Since this word appears 9 other times, all with the meteg in the "right" place, there's at least some possibility that this exception is just a misprint.
(b) For the verb xet-yud-hei, there were 3 cases out of 69 (4%) where the meteg failed to appear.
2. In Torah, on the other hand, it wouldn't be too much of an exaggeration to say that the meteg seems to be present or absent almost randomly, even from one occurrence to the other of the same word. In other words, the word yixye in D"varim 8:3 (which motivated this whole exercise) is not at all unusual.
(a) For the verb hei-yud-hei, the meteg is absent in 159 out of 441 cases (36%). The word yihye, where the meteg "fails" in 69 out of 206 cases (33%) accounts for almost half of the 441 total cases.
(b) For the verb xet-yud-hei, there are only 17 cases altogether (so not very much data), and the meteg is absent in 3 of them (18%).
Needless to say, the contrast between Torah and Na"kh is striking in both cases --- 0.7% vs. 36% for hei-yud-hei and 4% vs. 18% for xet-yud-hei. Although the comparison is slightly less lopsided for xet-yud-hei, the discrepancy is still large enough to jump right out at you.
Art