This post follows the custom of Prof. Werner Weinberg (cited below) and abbreviates ḥq = ḥaṭaf qamaṣ, qg = qamaṣ gadol, and qq = qamaṣ qaṭan.
Let me apologize for not dealing with the meteg. I believe that the meteg is often useful in determining the status of a qamaṣ, but has so many other uses that it isn’t always reliable. I remembered to write about the meteg at the last minute, and this post has been on my virtual desk for so long, so I’m putting the discussion off until another time (b”n). I apologize, because I know some of you are interested in the meteg. Why don’t you write about it, and I can just comment on your post?
So now to finally begin (or is finally beginning illogical? whatever).
In a post dated November 29, 2011, I announced that I disagreed with what I called the ṣohorayim rule (actually, at the time I called it the tzohorayim rule)--the rule that when a qamaṣ precedes a guttural with a ḥq, that qamaṣ is a qq. Although I disagreed with the rule then, I now believe that such a qamaṣ is in fact often, or even usually, a qq, as the rule predicts. But let’s discuss qamaṣim in general before we get to the question of those that are followed by a guttural with a ḥq.
As Gesenius notes in section 9u of his Grammar (downloadable here), “The grammatical origin of the words in question…is of course the surest guide” to the status of a qamaṣ. Makes perfect sense, and in fact the qamaṣ in צָהֳרַיִם is a qq under this guideline, since the word is the dual form of צֹהַר (the qq is the short vowel corresponding to the ḥolam). This, like Gesenius’s statement, is intuitive.
Without any knowledge of a word’s origin, we rely on two qualities of the qq: it must be in an unaccented closed syllable. There seems to be unanimous agreement that an accented qamaṣ is always a qg. But an unaccented qamaṣ in an open syllable may be a qq, and one in a closed syllable may be a qg. Everyone seems to agree that the first qamaṣ in שָׁרָשִׁים is a qq, but it is in an open syllable. This is consistent with Gesenius's generalization; שָׁרָשִׁים is the plural of שֹׁרֶשׁ.
There is at least one whole category of closed unaccented qamaṣim that are gedolim: the ⊃atei meraḥiq, discussed in Gesenius 20f. When one word that ends with an unaccented open qamaṣ or segol has a conjunctive cantillation and is followed by a word that begins with an accented syllable (this includes monosyllables), the first letter of the second word often takes a dageš ḥazaq. The open qamaṣ that ends the first word was and remains a qg, even after it’s closed by the dageš. An example of ⊃atei meraḥiq given by Gesenius is שׇׁבִיתׇ שֶּׁבִי (Psalms 68:19). The first word is accented on the bet, the second on the šin. The qamaṣ at the end of the first word is a qg, and it is closed by the dageš ḥazaq in the šin. Thus, the syllable is closed, and yet the qamaṣ remains a qg.
What about the inflected forms of רֹאשׁ that begin with -רָאשׁ? This seems like an obvious case where we need a qq. So why do people treat it like a qg? According to Gesenius (97, near the end [97 isn’t divided into subsections]) רֹאשׁ is “obscured from רָאשׁ"--that is, its ur-form (which doesn’t necessarily mean it’s from Aramaic) is רָאשׁ. There's no way of knowing this if you don't know anything about other Semitic languages; those who have never heard of Ras Tafari should probably pronounce these words with a qq.
Finally, let’s talk about Mordecai. Everybody seems to agree that the name מָרֽדֽכַי comes from the Persian name “Marduk.” If that is correct, and if “Marduk” is pronounced “Marduk,” and if the surest guide to the status of a qamaṣ is its origin, why is the qamaṣ in מָרֽדֽכַי not a qg?
Now, let’s return to the ṣohorayim rule. My former reasoning, which I now reject, was straightforward. A qq, I assumed, has to be in a closed unaccented syllable. More importantly, I assumed that a syllable that ends with a ḥaṭaf vowel is an open syllable. Why? Well, it sure doesn’t sound closed. And in the recorded ši⊂urim that first got me interested in this stuff, the lecturer said that you can identify open and closed syllables by their sound (assuming that one pronounces an ⊂ayin at the end of a syllable). Let’s use צָהֳרַיִם itself as an example. The first syllable can’t end with the hei; if it did, the first thing in the next syllable would be the ḥq. That won’t work, because the first thing in a syllable has to be a consonant. But if the first syllable is the ṣade and the qamaṣ, then the syllable is open, and the qamaṣ has to be gadol. (A reminder: I no longer accept this argument.)
I was about to reaffirm my earlier post but allow for exceptions and leave it at that until I came across a paper by Werner Weinberg, “The Qamāṣ Qāṭān Structures.” Weinberg identifies three ways of determining the status of a given qamaṣ--the etymological (he cites the statement of Gesenius that appears near the beginning of this post regarding the origin of a given qamaṣ), the phonological (a qq appears in a closed unaccented syllable), and the morphological. As its title suggests, the paper is largely devoted to morphology. Although Weinberg talks about these as three approaches, he doesn’t actually treat them as separate. At the end of the etymology section, he writes (paragraph 1.4), “Aside from obvious cases, one has to be an accomplished Semitist to use a purely etymological approach--and yet, even the expert cannot always decide between qq and qg.” The footnote attached to this statement lists fourteen footnotes that cite Semiticist minority reports on certain qamaṣim, and all are in the section on morphology, even though he had just said that these were disagreements on etymology.
(Weinberg’s paper has several illogical arguments around the edges. I mention this not out of disrespect for Weinberg, but because I’m implicitly sort of recommending the paper by citing it, and I believe it’s my duty to warn people.)
Anyhow, the insight. The main thing we’re taught about ḥaṭaf vowels, other than that they’re shorter than short, is that they appear under guttural letters when a ševa⊃ na⊂ would otherwise be called for; a ševa⊃ na⊂ can’t appear under a guttural. True enough. But what isn’t dwelled on as much is the fact that a ḥaṭaf vowel can also stand in for a ševa⊃ naḥ. We see examples of this all over the Hebrew language. Consider some hif⊂ils. Look, for example, at מַבְדִּיל. It seems noncontroversial that the ševa here is naḥ. Therefore, the ḥaṭaf pataḥ in מַחֲזִיק must be standing in for a ševa⊃ naḥ.
OK. Now. Here comes the insight I got from Weinberg (wait for it!). Weinberg has several lists, by category, of words that take qq. I quote the beginning of section 3.1.1 (one of several that make this point):
3.1.1. qoṭl-type nouns, singular inflected: אׇהֳלִי (Jer 10 20), etc. [incl. pr. nouns]; אׇזְנִי (I Sam 20 2), etc….And a few other sections of Weinberg contain similar lists: 3.1.2 (qoṭl-type dual nouns, including our old friend צָהֳרָיִם [Isaiah 16:3]) and 3.1.3 (qoṭl-type nouns plural construct, such as אׇהֳלֵי [Numbers 16:26]).
Don’t worry, the insight is in the next paragraph.
All of the examples I’ve cited (although not all that Weinberg lists) are consistent with Gesenius’s statement about the origin of the qamaṣ: all come from words whose uninflected singulars have a ḥolam. But in terms of the argument about the ṣohorayim rule, here’s the insight. All of these are, according to Weinberg, lists of “qoṭl-type nouns.” In other words, the second radical carries a ševa⊃ naḥ. The ḥq under the hei in oholi stands in for a ševa naḥ. The first syllable of the word isn’t the qamaṣ ⊃alef; the first syllable is the qamaṣ ⊃alef and the ḥq hei, and (my big insight from Weinberg) it is a closed syllable. The qamaṣ under the ⊃alef meets the criteria for a qq without invoking the ṣohorayim rule. My big insight should have been obvious--when a syllable ends with a ḥaṭaf vowel that stands in for a ševa⊃ naḥ, it’s a closed syllable. My problem, as I mentioned earlier, was that I assumed that closed syllables sound like they end with a consonant. [Note to myself: remember to include video of me smashing a rotten eggplant on my head in sheer embarrassment.] The first syllable of צָהֳרָיִם is צָהֳ, but its underlying form is צָהּ; it is a closed syllable. So now I conclude that the ṣohorayim rule often works.
But not always. R’ Mordecai Breuer writes, “A qamaṣ that precedes a ḥaṭaf qamaṣ whose source is a ševa⊃ naḥ is considered a short vowel.” Note the restriction--“whose source is a ševa⊃ naḥ.” This strongly suggests that he rejects the ṣohorayim rule as a universal generalization. So do I.
Let’s look at a few other cases. What about כָּאֳנִיּוֹת (Proverbs 31:14), which we sing Friday nights as part of “⊃Ešet ḥayil”? Here we have a qamaṣ preceding a guttural with a ḥq; how is it pronounced? According to Gesenius (102d), a prefixed prepositional bet, kaf, or lamed that is attached to a ḥaṭaf letter takes the full vowel associated with that ḥaṭaf. Based on some words that are familiar from worship, this makes intuitive sense. Consider בֶּאֱמֶת and לַאֲדוֹנֵי. Using this reasoning, it’s clear that the qamaṣ in כָּאֳנִיּוֹת is a qq. And obviously, since אֳנִיּוֹת begins with a ševa⊃ na⊂--words don’t begin with a ševa⊃ naḥ--this qq is an open syllable.
Although it is obviously open, it may be nonobviously closed. And at this point I’m speculating even more cluelessly than usual. Clearly, an initial ḥaṭaf is standing in for a ševa⊃ na⊂, since there is no initial ševa⊃ naḥ. Nevertheless, let’s look at what happens in a word that begins with a plain old simple ševa⊃ na⊂, and imagine what would happen if the gutturals could take a ševa⊃ na⊂. An asterisk indicates a hypothetical form that doesn’t really exist.
1b. אְנִיּוֹת *
Both of these ševa⊃s are na⊂, since they go with the first letter. Now let’s add a prepositional kaf to both. Without making any of the necessary changes, the words become
2a. כְּצְדָקָה *
2b. כְּאְנִיּוֹת *
I don’t know whether these hypothetical ševa⊃s would be na⊂ or naḥ. I suspect they’d be na⊂, and for our present purposes it doesn’t matter. Let’s go one further step, again treating word b like word a.
3b. כִּאְנִיּוֹת *
The real ševa⊃ in 3a becomes naḥ in this situation, and so does the hypothetical one in 3b. Is it possible that כָּאֳנִיּוֹת is derived from 3b? If so, once again we have a closed syllable.
What happens when the article is prefixed to a ḥaṭaf? Gesenius notes that before a guttural, depending on the details, the vowel in the article either remains a pataḥ or is “modified to a Seghôl or fully lengthened to Qameṣ” (35e) (emphasis added; Gesenius’s italics on the names of the vowels deleted to emphasize my emphasis). In other words, it’s a qg. Before an ⊂ayin with a ḥq, the article has a qamaṣ (35k), as in הׇעֳמׇרִים (Ruth 2:15). This is a qg preceding a guttural that has a ḥq: exactly what the ṣohorayim rule doesn't allow.
The last instance or noninstance of the ṣohorayim rule that I discuss concerns a postbiblical form. Near the end of the ⊂Al ḥeṭ, we call God
סׇלְחׇן לישראל ומׇחֳלׇן לשבטי ישורון
I’ve been assuming (maybe incorrectly) that סׇלְחׇן and מׇחֳלׇן have the same vowels. If so, what are they? Is the first qamaṣ a qg (with ševa⊃ na⊂ and a ḥaṭaf qamaṣ coming from a na⊂), or is it a qq (with naḥs)?
The dictionaries I’ve looked at all have מׇחֳלׇן. They also all have an entry for סׇלְחׇן, סַלֽחָן, with the pataḥ version appearing first. I surmise that these are two different pointings for the same word, rather than synonyms, because if they were two different related and synonymous words, they would have separate entries. Since the pataḥ is the short vowel that corresponds to the qg, I gather that the vowel in סׇלְחׇן is a qg--it just seems less Occamly plausible for the two pointings of the same word to differ by a pataḥ and a qq. The mishnaic grammar of M. H. Segal lists some basic forms of nouns formed by adding a final nun:
The correct vowels for סׇלְחׇן may be irrelevant. It’s irresponsible to guess authorial intention, especially for me--I mean, well, I’m clueless about my own intentions most of the time. Nevertheless, I’m guessing that the author of the prayer intended that סׇלְחׇן and מׇחֳלׇן be pronounced with the same vowels, so that’s what I’m going to do.
Cluelessness, uninformed speculation, weak assumptions, wild guesses, and more cluelessness, always cluelessness--what blogging is all about. Please comment and correct.
 קמץ הבא לפני חטף קמץ, שמקורו שוא נח, נחשב תנועה קטנה. R. Mordecai Breuer, Ṭa⊂amei ha-miqra⊃ (Jerusalem: Ḥorev, 1989), p. טז (frontmatter).
 M. H. Segal, A Grammar of Mishnaic Hebrew (Oxford: Clarendon, 1927), 119; downloadable here.
 Seligmann Baer, Seder ⊂Avodat Israel (Rödelheim: Lehrberger, 1901), 100. Hebrewbooks.org has two scans of it--here and here. Some pages have clearer scans in one version, some in the other.