Adapted, by permission, from Israel Yeivin, Introduction to the Tiberian Masorah, translated and edited by E. J. Revell. Copyright © 1980 by the Society of Biblical Literature.
The distinction of silent from vocal shewa was a great concern for the Masoretes for two reasons:
| E.g., in | וַיִּשְׁכְּנ֨וּ | |
| the silent shewa | וישְכנו | goes with what precedes, and |
| the vocal shewa | וישכְנו | goes with what follows. |
| So, | וַי יִּשְׁ כְּנ֨וּ | is the syllable structure of this word. |
| And, in | וְהָאֲחַשְׁדַּרְפְּנִ֣ים | |
| the silent shewa s | והאחשְדרְפנים | go with what precedes, and |
| the vocal shewa s | וְהאחשדרפְנים | go with what follows. |
| So, | וְהָ אֲחַשְׁ דַּרְ פְּנִ֣ים | is the syllable structure of this word. |
(From אֲחַשְׁ in the example above, we can see that ḥaṭef pataḥ, like simple vocal shewa, goes with the syllable of the next vowel.)
Phonetic gaʿya (#346) could change silent shewa to vocal, and so change the syllable structure of a word. The word וּזְהַב without gaʿya has syllable structure וּזְ הַב , but with gaʿya, וּֽזְהַ֛ב , it becomes וּֽ זְהַ֛ב. (This word only appears in the Bible with gaʿya.) In the same way at the end of a word, silent shewa (whether marked or only potential) marks the structure. In בְּרֵאשִׁית בָּרָא, if the potential shewa on tav were vocal, the structure would be בְּרֵא שִׁי תְבָּ רָא.
Biblical references in this section: וַיִּשְׁכְּנ֨וּ Gen 25:18, וְהָאֲחַשְׁדַּרְפְּנִ֣ים Est 9:3, וּֽזְהַ֛ב Gen 2:12, בְּרֵאשִׁית בָּרָא Gen 1:1.
At the start of a word, shewa is vocal. The only exception to this is found in the forms of שְׁתַּיִם . Here the initial shewa is considered silent. According to some sources (e.g. Levi, 1936, p. ט, translation p. 8-star), this pronunciation was made possible among the Tiberians by the use of a helping vowel, as אֶשְׁתַּיִם (the first syllable (highlighted) is a closed syllable). This helping vowel was evidently the source of a yetiv vs. pashṭa disagreement when the word has no preceding conjunctive, as in the following:
| תְּנוּפָ֗ה | (אֶ)שְׁ֚תַּיִם ? (אֶ)שְׁתַּ֙יִם֙ ? | |
| לְאִ֗ישׁ | ↑ | |
| לַדְּלָת֑וֹת | ↑ | 1st time in this verse |
| דְּלָת֔וֹת | ↑ | 2nd time in this verse |
Footnotes for this section:
φ1 Note that שְׁתַּיִם without an initial, implicit helping vowel has initial stress whether or not we consider its shewa to be vocal. (As usual, we use the masoretic notion of a syllable, so initial stress means stress on the first full vowel, whether or not there is an initial vocal shewa.) And, שְׁתַּיִם with an initial, implicit helping vowel still has initial stress if that helping vowel is not a full vowel, e.g. if the pronunciation is more like (אֱ)שְׁתַּיִם than (אֶ)שְׁתַּיִם (more like ḥaṭef segol than segol ).
φ2 Yetiv on an initial shewa occurs nowhere else but in these four cases of שתים. In this respect, yetiv differs from the other prose prepositive, telisha gedolah, because telisha gedolah often appears on initial shewa. It even appears in some cases somewhat analogous to שתים such as שְׁ֠מֹר and שְׁ֠נֵי . (Those examples look like שְׁ֠מֹ֠ר and שְׁ֠נֵ֠י when shown with “stress helpers” for those not used to the masoretic notion of a syllable.)
φ3 My guess is that bN put yetiv in this novel position to try to clarify that the stress was on the syllable whose main part is תַּ . I.e. my guess is that this position was meant to mimic the way yetiv would appear on a (fictitious) word תַּ֚יִם . Unfortunately this also must have looked quite like a mehuppak, i.e. must have looked quite like שְׁ֤תַּיִם .
Biblical references in this section: תְּנוּפָ֗ה Lev 23:17, לְאִ֗ישׁ Ez 1:11, לַדְּלָת֑וֹת Ez 41:24, דְּלָת֔וֹת Ez 41:24, שְׁ֠מֹר 1K 8:25 and 2C 6:16, שְׁ֠נֵי 2S 21:8.
At the end of a word, shewa is silent. Most sources also say that when two shewa s come together at the end of a word, as וַיֵּ֣בְךְּ עֲלֵהֶ֑ם , both are silent. However, some sources state that the first is silent and the second vocal, unless the word occurs at a major pause, as וַיֵּֽבְךְּ׃ , in which case all agree that both are silent.
Biblical references in this section: וַיֵּ֣בְךְּ עֲלֵהֶ֑ם Gen 45:15, וַיֵּֽבְךְּ׃ Gen 29:11.
Within a word, where shewa is marked on a pair of letters, the first is silent and the second vocal. Shewa on a letter marked with dagesh is vocal. Apart from these two clear cases, shewa within a word is consider silent, with the exception of several special categories of shewa which are noted in various masoretic sources.
Shewa on resh. The Masorah gives rules on the subject of shewa on resh, but these are not the same in different sources. It is said that in nouns, if resh is the second letter, and has qameṣ or ṣere before it, then a shewa on the resh is vocal. E.g.:
| הָרְוָחָ֔ה | הָ | רְוָ | חָ֔ה |
| מֵרְפִידִ֗ים | מֵ | רְפִי | דִ֗ים |
If it has ḥireq or ḥolem before it, however, shewa on the resh is silent. E.g.:
| בִּרְחֹ֣ב | בִּרְ | חֹ֣ב |
| יוֹרְדִ֔ים | יוֹרְ | דִ֔ים |
(From יוֹרְדִ֔ים above, we now see that the real criteria is that resh is the second letter excluding any vowel letter (mater lectionis).)
Rules on shewa on resh in verb forms are given for the roots גרשׁ and ברך.
| Rules for the root גרשׁ | Example |
|---|---|
| If the shin has segol, then a shewa on the resh is vocal. | אֲגָרְשֶׁ֖נּוּ |
| Otherwise, the shewa is silent. | וַיְגָרְשׁ֣וּ |
| Except, in the following word, the shewa is considered vocal by bA even though the shin has ṣere rather than segol. Ben Naftali rejects this exception, considering the shewa silent. | וַ֝יְגָרְשֵׁ֗הוּ |
| Rules for the root ברך | Example |
|---|---|
| If the accent is on the bet, shewa on the resh is silent. | בָּ֥רְכוּ נָ֖א |
| If the accent is on the kaf, shewa on the resh is vocal. | בָּרְכֵ֥נִי |
| Except, in the following word, the shewa is considered silent even though the accent is on the kaf (Diqduqe ha-Ṭeʿamim (Dotan 1967, section 21)). | בָּרְכֵ֔ת |
Shewa on resh is considered vocal also in the following:
| sources and notes | |
|---|---|
| אֵֽרְדָה־נָּ֣א | Diqduqe ha-Ṭeʿamim. |
| נִ֣ירְשָׁה לָּ֑נוּ | Given as a case of agreement between bA and bN. |
| אֲקָֽרְבֶֽנּוּ׃ | According to bA but not bN. μA and μL have ḥaṭef pataḥ. |
Biblical references in this section: הָרְוָחָ֔ה Ex 8:11, מֵרְפִידִ֗ים Ex 19:2, בִּרְחֹ֣ב Jud 19:17, יוֹרְדִ֔ים Jud 9:37, אֲגָרְשֶׁ֖נּוּ Ex 23:30, וַיְגָרְשׁ֣וּ Jud 11:2, וַ֝יְגָרְשֵׁ֗הוּ Ps 34:1, בָּ֥רְכוּ נָ֖א 1C 29:20, בָּרְכֵ֥נִי Gen 27:34, בָּרְכֵ֔ת Dan 4:31, אֵֽרְדָה־נָּ֣א Gen 18:21, נִ֣ירְשָׁה לָּ֑נוּ Ps 83:13, אֲקָֽרְבֶֽנּוּ׃ Job 31:37.
Shewa on letters other than resh. The Masorah gives rules for the pronunciation of shewa in some other verb forms:
| Rules for the root אכל | Example |
|---|---|
| If the lamed has segol, then a shewa on the kaf is vocal. | תֹּאכְלֶ֑נּוּ |
| Otherwise, the shewa is silent. | תֹּאכְל֖וּן |
| Except, in the following word, the shewa is considered silent even though the lamed has a segol. | אֹכְלֶ֑יהָ |
The rules above for the root אכל are stated in Diqduqe ha-Ṭeʿamim (Dotan 1967, section 22), and are also given in the Sefer ha-Ḥillufim as the opinion of bA. However, bN regards the shewa on kaf as silent in all cases.
Rules for the root הלך – Shewa on the lamed is silent except in the long form of the imperfect where the stress is retracted to the first syllable, and the next word has deḥiq (conjunctive dagesh) (#405), as אֵ֣לְכָה נָּ֗א (Diqduqe ha-Ṭeʿamim (Dotan 1967, section 25)). (μL has ḥaṭef pataḥ.)
Biblical references in this section: תֹּאכְלֶ֑נּוּ Dt 12:24, תֹּאכְל֖וּן Nu 11:19, אֹכְלֶ֑יהָ Qoh 5:10, אֵ֣לְכָה נָּ֗א Ex 4:18.
Shewa after Phonetic Gaʿya. Shewa after a short vowel is usually silent, even where this vowel has musical gaʿya (#319). However, if the short vowel has phonetic gaʿya intended to show that the shewa that follows is vocal, the shewa is, of course, vocal. The situations in which phonetic gaʿya is used are described starting in #345. The Masorah gives rules covering some of these situations.
Shewa on Mem after initial He with Pataḥ. (cf. #347)
These rules apply not only to he, but also to bet, kaf, or lamed representing a preposition with the vowel of the definite article.
Where the he has a gaʿya, if it is a musical gaʿya, the shewa that follows is silent, but if it is a phonetic gaʿya, the shewa that follows is vocal. The Diqduqe ha-Ṭeʿamim (Dotan 1967, section 14) states that, as a rule, gaʿya in this situation is phonetic, so the shewa is vocal.
Various examples of phonetic gaʿya are given, mainly in two categories. These categories differ from each other in the reason that the gaʿya was unlikely to have been musical (#319):
In words in these two categories, then, the gaʿya is phonetic and therefore the shewa is vocal. This is followed by a list of words in which the gaʿya is musical and therefore the shewa is silent. For most of these words, the fact that the gaʿya is musical is not surprising, because these words meet both the structural and accent criteria for a musical gaʿya. E.g., הַֽמְשַׂחֲק֖וֹת (FR3) and הַֽמְאָרְרִ֖ים (AFR4) . (φ1)
But for some some of these words, the fact that the gaʿya is musical is surprising, because they fail to meet one or both of the criteria (structural and accent) for a musical gaʿya.
Some have a conjunctive accent, as הַֽמְקַטְּרִ֣ים .
And some have non-regular structure, as הַֽמְעֻשָּׁקָ֞ה .
Whether it is surprising or not, in all these cases the gaʿya is musical, and the shewa is silent.
Some sources give rules for determining whether shewa is vocal or silent in such cases by the number of letters in the word.
Some exceptions to these rules are also listed.
The rule given in the Diqduqe ha-Ṭeʿamim covers most of the cases, but not all. The other rules, however, cover an even smaller proportion. In some manuscripts, notably in μA, the mem in this situation is pointed with ḥaṭef pataḥ when the shewa is vocal, which gives a clear indication of the pronunciation.
Footnote for this section:
φ1 In ITM, הַֽמְאָרְרִ֖ים is given three locales, including two in Nu 5:24: “Nu 5:19, 24, 24.” But, the second instance in Nu 5:24 is המְאָֽרְרִ֑ים , which, at least in μL, differs from הַֽמְאָרְרִ֖ים in the following ways:
One might speculate that the difference in the position of the gaʿya is explained by the pausal vs. non-pausal difference. This idea is strengthened by another nearby example in μL, הַמְאָֽרְרִֽים׃ . But, yet another nearby example in μL, הַמְאָֽרְרִים֙ , spoils the pattern. The table below summarizes the (somewhat confusing) situation. This table includes an example not mentioned above, הַמְאָרְרִ֤ים , which, in μL at least, has no gaʿya in any position. (A gaʿya on the ה of that word would be unexpected since it would be on a closed syllable before a conjunctive accent. But, as far as I know, there are no firm expectations as to whether there should or should not be a gaʿya on the open syllable starting with א.)
| הַֽמארר֖ים | 2 cases |
| המאֽרר֑ים | |
| המאֽררֽים׃ | |
| המאֽררים֙ | |
| המארר֤ים |
For completness, and because it may be related, we note that the notation of the shewa on the resh in these words varies considerably from manuscript to manuscript. See the MAM documentation. I.e. some manuscripts (notably μL and μB) use ḥaṭef pataḥ where others use simple shewa. The only word for which there appears to be no variation is (coincidentally?) the one with a conjunctive accent, הַמְאָרְרִ֤ים . For this word, all manuscripts documented by MAM use simple shewa.
We have used simple shewa throughout this footnote, since, for one thing, that is what ITM uses for the three words at the locales “Nu 5:19, 24, 24.” So our claims about the contents of μL above are not quite right, because in all cases except הַמְאָרְרִ֤ים , μL uses ḥaṭef pataḥ on resh. Whereas, we have shown simple shewa in all cases.
Biblical references in this section: הַֽמְנַדִּ֖ים Amos 6:3, הַֽמְשַׁלְּחִ֤ים 2C 32:31, הַֽמְשַׂחֲק֖וֹת 1S 18:7, הַֽמְאָרְרִ֖ים Nu 5:19 and Nu 5:24, הַֽמְקַטְּרִ֣ים 2K 23:5, הַֽמְעֻשָּׁקָ֞ה Is 23:12, הַֽמְסֻכָּ֣ן Is 40:20, הַֽמְצַפְצְפִ֖ים Is 8:19, הַֽמְחַכִּ֣ים Job 3:21, הַמְאָֽרְרִֽים׃ Nu 5:18, הַמְאָֽרְרִים֙ Nu 5:27, הַמְאָרְרִ֤ים Nu 5:22, הַֽמארר֖ים Nu 5:19 and Nu 5:24, המאֽרר֑ים Nu 5:24, המאֽררֽים׃ Nu 5:18, המאֽררים֙ Nu 5:27, המארר֤ים Nu 5:22.
Shewa after Phonetic Gaʿya in Other Situations. Shewa on other letters is vocal when it follows phonetic gaʿya. E.g.:
| Type | |
|---|---|
| הַֽנקלה | הַֽנְקַללָּ֤ה |
| ↑ | בַּֽסְעָרָ֖ה |
| ↑ | בַּֽנְחֻשְׁתַּ֔יִם |
| וּֽשלח | וּֽשְׁלַ֥ח |
| ↑ | וּֽטְהׇר־יָ֝דַ֗יִם |
| התִֽמלך | הֲתִֽמְלֹ֔ךְ |
| ↑ | אַֽרְזֵי־אֵֽל׃ |
Types used in the table above:
| הַֽנקלה | On he with pataḥ (or bet, kaf, or lamed with the vowel of the definite article) at the start of a word (#348). |
| וּֽשלח | On conjunctive waw pointed as shureq (#349). |
| התִֽמלך | On short-vowelled syllables (#350). |
(In the case of אַֽרְזֵי־אֵֽל׃ , bA and bN agree that the shewa is vocal.)
Biblical references in this section: הַֽנְקַלָּ֤ה 1S 18:23, בַּֽסְעָרָ֖ה 2K 2:1 and 2K 2:11, בַּֽנְחֻשְׁתַּ֔יִם Jud 16:21, וּֽשְׁלַ֥ח 2K 9:17, וּֽטְהׇר־יָ֝דַ֗יִם Job 17:9, הֲתִֽמְלֹ֔ךְ Jer 22:15, אַֽרְזֵי־אֵֽל׃ Ps 80:11.
Shewa in a Particular Poetic Sequence. Shewa is vocal in the following sequence particular to the three books:
This sequence occurs mainly among the servi of revia gadol (#363), revia qaṭan (#368), ṣinnor (#365), and legarmeh (#370). According to the Diqduqe ha-Ṭeʿamim (Dotan 1967, section 13), the shewa is vocal. E.g.:
| נִ֥בְהָֽל לַה֗וֹן |
| קִ֥רְבַ֥ת אֱלֹהִ֗ים |
| תִּ֥בְחַ֣ר וּתְקָרֵב֮ |
| שִׁ֥מְעָֽה תְפִלָּתִ֨י׀ |
| וְלִ֥שְׁכֵנַ֨י׀ |
There are four exceptions in which shewa in this sequence is silent:
| שָׁ֥בְרָ֥ה לִבִּ֗י |
| שׇׁ֥מְרָ֣ה נַפְשִׁי֮ |
| טָ֥מְנֽוּ־גֵאִ֨ים׀ |
| יִ֥רְאַ֣ת יְהֹוָה֮ |
Biblical references in this section: נִ֥בְהָֽל לַה֗וֹן Prov 28:22, קִ֥רְבַ֥ת אֱלֹהִ֗ים Ps 73:28, תִּ֥בְחַ֣ר וּתְקָרֵב֮ Ps 65:5, שִׁ֥מְעָֽה תְפִלָּתִ֨י׀ Ps 39:13, וְלִ֥שְׁכֵנַ֨י׀ Ps 31:12, שָׁ֥בְרָ֥ה לִבִּ֗י Ps 69:21, שׇׁ֥מְרָ֣ה נַפְשִׁי֮ Ps 86:2, טָ֥מְנֽוּ־גֵאִ֨ים׀ Ps 140:6, יִ֥רְאַ֣ת יְהֹוָה֮ Prov 8:13.
Shewa on the First of an Identical Pair of Letters. In the Diqduqe ha-Ṭeʿamim (Dotan 1967, section 5), and in other masoretic treatises, the following rule is given (we’ll call it D1):
| Example(s) | |
|---|---|
| A shewa on the first of an identical pair of letters is vocal if it follows gaʿya or the stress. (φ1) | יִֽלְלַ֣ת לָֽקְק֤וּ |
| Otherwise, such a shewa is silent. | הִנְנִ֨י הִנְנ֣וּ חִקְקֵי־אָ֑וֶן |
| Except, in six words (φ2) such a shewa is silent (φ3) even though it follows the stress. | יְ֝שַׁחֲרֻ֗נְנִי |
This rule makes no distinction between gaʿya on a short vowel, which is certainly phonetic, and gaʿya on a long vowel, as צָֽלְלוּ֙ , which could be musical.
This rule is not reflected in the manuscripts, not even in manuscripts in which vocal shewa is often made explicit by the use of ḥaṭef shewa on non-guttural letters. In such manuscripts, some words’ pointing is consistent with the rule: ḥaṭef shewa is used where gaʿya is marked, but not where gaʿya is not marked. E.g., in μA:
| וְהִֽלֲל֖וּ | ḥaṭef shewa | gaʿya present | |
| מִתּֽוֹכֲךָ֙ | ḥaṭef shewa | gaʿya present | |
| מֵרִבְבֹ֣ת ? מֵרִבְבֹ֣ת ? | מֵרִבְבֹ֣ת | simple shewa | gaʿya absent |
| הַשּׁוֹדְדִ֖ים ? הַשּׁוֹדְדִ֖ים ? | הַשּׁוֹדְדִ֖ים | simple shewa | gaʿya absent |
However, other words’ pointing is not consistent with the rule. E.g., in μA:
| לְשִֽׁמְמ֥וֹת ? לְשִֽׁמְמ֥וֹת ? | לְשִֽׁמְמ֥וֹת | simple shewa | gaʿya present | (cf. μL20 מֲמ) |
| נְגַשֲׁשָׁ֤ה | ḥaṭef shewa | gaʿya absent | ||
| שֹׁדֲדִ֛ים | ḥaṭef shewa | gaʿya absent |
The study of the pointing of μA in this situation, and comparison of it with other early manuscripts, reveals a somewhat different rule (we’ll call it D2). This rule distinguishes between gaʿya on a short vowel and gaʿya on a long vowel. This rule can be stated as follows:
With this new rule, we are now able to resolve the three vocal/silent questions left open above: (φ8)
| מֵרִבְבֹ֣ת | silent: follows a gaʿya-free short vowel |
| הַשּׁוֹדְדִ֖ים | vocal: follows a long vowel |
| לְשִֽׁמְמ֥וֹת | vocal: follows a gaʿya-marked short vowel |
Footnotes for this section:
φ1 I added the qualification “or the stress”; this qualification is not present in ITM. I added this qualification on the assumption that the shewa s in the six exceptions are silent, making them exceptions to this first part of the rule. In ITM it is not explicitly stated that the shewa s in these six exceptions are silent. I assume they are silent on the authority of Khan (The Tiberian Pronunciation Tradition of Biblical Hebrew, volume 1, page 353).
φ2 The six exceptions are as follows:
| יְ֝שַׁחֲרֻ֗נְנִי |
| יִ֭קְרָאֻנְנִי |
| מְצָאֻֽנְנִי׃ |
| מְצָאֻֽנְנִי׃ |
| יְֽכַ֫בְּדָ֥נְנִי |
| שַׁחֲרֻֽנְנִי׃ |
(The first of this list, יְ֝שַׁחֲרֻ֗נְנִי , is the one given as an example in ITM.) (Only 5 of the 6 exceptions are unique since מְצָאֻֽנְנִי׃ is in two verses.) (Also note that 3 of the 6 exceptions are in the same verse: Prov 1:28!)
φ3 In ITM it is not explicitly stated that the shewa s in these six exceptions are silent. I assume they are silent on the authority of Khan (The Tiberian Pronunciation Tradition of Biblical Hebrew, volume 1, page 353).
φ4 How did Yeivin determine that Rule D1 is not obeyed in manuscripts in which vocal shewa is rarely made explicit? There are two ways I can think of that he might have done this.
φ5 How did Yeivin determine that Rule D1 is not obeyed in manuscripts in which vocal shewa is often made explicit? To answer that question, we first need to know: what does it even mean for this rule to be obeyed in such manuscripts? I ask this second, more basic question because in my interpretation at least, this rule’s purpose is to resolve the ambiguity surrounding the “vocality” of simple shewa, but in many cases these manuscripts have no such ambiguity, due to their use of ḥaṭef.
In words where a vocal shewa sound is already explicitly called for, i.e. in words where ḥaṭef is used, it is in some sense “unfair” to expect a gaʿya to redundantly call for a vocal shewa sound. So it is not surprising that, as Yeivin documents here and elsewhere, we don’t consistently find such redundancy.
The more interesting words are ones in which only simple shewa is used. In such words, the question becomes: which of the following is true?
As Yeivin documents here and elsewhere, unfortunately the answer is number 2: although in these manuscripts ḥaṭef is often used to unambiguously notate a vocal shewa sound, simple shewa is still used ambiguously in these manuscripts in the circumstances under discussion: on the FIP.
Let’s return to the basic question of what Yeivin means when he says that Rule D1 is not obeyed in manuscripts in which vocal shewa is often made explicit. What he means, initially at least, is the following:
So Yeivin’s examples merely show that manuscripts like μA don’t obey D1 “naively,” by which I mean that they don’t obey D1 assuming that simple shewa means silent shewa. But, the real question is whether Rule D1 is obeyed in these words without assuming that simple shewa means silent shewa. The answer is that Rule D1 is not obeyed in that way either. Yeivin implicitly provides that answer when he goes on to give Rule D2.
φ6 Presumably, I could add the qualification “or the stress” as I did in Rule D1, but it is not clear whether that ever happens, in practice.
φ7 Although ITM does not state it explicitly, I assume that the same six exceptions of the Rule D1 apply here in Rule D2 as well.
φ8 I added this sentence and the table below it.
Biblical references in this section: יִֽלְלַ֣ת Zech 11:3, לָֽקְק֤וּ 1K 21:19, הִנְנִ֨י Ex 10:4, הִנְנ֣וּ Jos 9:25, חִקְקֵי־אָ֑וֶן Is 10:1, יְ֝שַׁחֲרֻ֗נְנִי Prov 1:28, צָֽלְלוּ֙ Ex 15:10, וְהִֽלֲל֖וּ Is 62:9, מִתּֽוֹכֲךָ֙ Is 58:9, מֵרִבְבֹ֣ת Dt 33:2, הַשּׁוֹדְדִ֖ים Jer 51:48, לְשִֽׁמְמ֥וֹת Jer 25:12, נְגַשֲׁשָׁ֤ה Is 59:10, שֹׁדֲדִ֛ים Jer 51:53, יִ֭קְרָאֻנְנִי Prov 1:28, מְצָאֻֽנְנִי׃ Prov 1:28, מְצָאֻֽנְנִי׃ Prov 8:17, יְֽכַ֫בְּדָ֥נְנִי Ps 50:23, שַׁחֲרֻֽנְנִי׃ Hos 5:15.
In addition to these rules, a number of cases of ḥillufim also concern the pronunciation of shewa within the word. E.g.:
| וּקְטׇרְתִּ֔י | bA |
| וּֽקֲטׇרְתִּ֔י | bN and “Tiberias” |
And: (φ1)
| כִּקְסׇום־שָׁוְא֙ | bA |
| כִּֽקֲסׇום־שָׁוְא֙ | bN |
| כִּֽקֳסׇום־שָׁוְא֙ | R. Pinḥas |
The manuscripts also show some other words, not mentioned in the rules given above, in which a ḥaṭef shewa is used where a silent shewa is expected. E.g.:
On resh:
| וּבָרֲקַ֖ת | μS1 |
| דֹּ֣רֲשָׁ֑יו | μS1 |
After gaʿya or an accent sign on a long vowel:
| וְתִֽימֲר֖וֹת | μA |
| אָ֣זֲלַת | μA, μS1 |
| ה֥וּסֲפַת לִֽי׃ | μS1 |
After a short vowel with no phonetic gaʿya:
| קׇסֳומִי־נָ֥א | μA | cf. μL, μC, μS1 קָֽסֳומִי־ | φ2 |
| בִּסֲבׇךְ־עֵ֝֗ץ | μA | cf. μL בִּֽסֲבׇךְ־ |
It seems probable that, although the rules given above cover most of the cases in which shewa within a word was considered vocal, they do not cover them all. Apart from those cases mentioned in the rules, however, and the few exceptional cases, shewa within a word was considered silent, whether it came after either of the following:
Footnotes for this section:
φ1 Most manuscripts and printed texts note this word’s superfluous waw, either compactly, e.g. ‹ יתיר ו̇ ›, or fully, as a ketiv of קסומי corresponding a qere of קׇסֳמִי־ .
φ2 Most manuscripts and printed texts note this word’s superfluous waw, either compactly, e.g. ‹ יתיר ו̇ ›, or fully, as a ketiv of כקסום corresponding a qere of כִּקְסׇם־ .
Biblical references in this section: וּקְטׇרְתִּ֔י Ez 16:18, כִּקְסׇום־שָׁוְא֙ Ez 21:28, וּבָרֲקַ֖ת Ez 28:13, דֹּ֣רֲשָׁ֑יו Ps 22:27, וְתִֽימֲר֖וֹת Joel 3:3, אָ֣זֲלַת Dt 32:36, ה֥וּסֲפַת לִֽי׃ Dan 4:33, קׇסֳומִי־נָ֥א 1S 28:8, בִּסֲבׇךְ־עֵ֝֗ץ Ps 74:5, וַיִּשְׁלַ֤ח Gen 8:9, שָֽׁמְע֔וּ Gen 43:25.
From the point of view of length, shewa represented an ultra-short vowel — even shorter than a short vowel. From the point of view of quality, shewa usually had the same sound as pataḥ. If a guttural came next, however, the shewa had the same quality as the vowel sound after the guttural. Thus:
(For the Tiberians, the qameṣ sign always represented the same vowel quality, /ɔ/ (open-mid back rounded), so ḥaṭef qameṣ and qameṣ ḥaṭuf (qameṣ qaṭan) differed from qameṣ gadol only in length.) If the shewa was itself on a guttural, however, it was pronounced with the quality of pataḥ even where a guttural came next. E.g.:
| שָׁ֝חֲח֗וּ |
| יִמְחֲאוּ־ |
| אֲחִימַן֙ |
Before a consonantal yod, vocal shewa was pronounced as a very short ḥireq, as in בְּיוֹם and לְיָקִ֖ים. However, if shewa before yod was on a guttural, it was not pronounced as ḥireq, but as ḥaṭef pataḥ or some other ḥaṭef. E.g.:
| חֲיִ֔י |
| הֱיוֹתְךָ֣ |
| אֱיָֽל׃ |
The pronunciation of shewa with gaʿya is described in #336.
Biblical references in this section: לֻֽקְחָה־זֹּֽאת׃ Gen 2:23, שָׁ֝חֲח֗וּ Job 9:13, יִמְחֲאוּ־ Ps 98:8, אֲחִימַן֙ Nu 13:22, לְיָקִ֖ים 1C 24:12, חֲיִ֔י Ez 16:6, הֱיוֹתְךָ֣ Jud 18:19, אֱיָֽל׃ Ps 88:5.
Only on gutturals does Tiberian pointing regularly distinguish vocal shewa from silent shewa. In many cases, however, shewa on other letters is represented by ḥaṭef to indicate that it is vocal, in some — sometimes in most — of the manuscripts and printed texts. This may occur either for morphological or for phonetic reasons. This section and the next few will cover these two reasons:
(i) Ḥaṭef Shewa Used for Morphological Reasons. This occurs mostly where the shewa derives from an /o/ or /u/ vowel, in which case the sign used is ḥaṭef qameṣ. This is especially common on qof, gimel, the other begad-kefat letters, ṭet, and the sibilants ṣade and shin. Presumably the tendency to preserve the original /o/ or /u/ sound was greater with them than with other letters. E.g.:
At the start of a word, with ḥaṭef :
| דֳּמִי | |
| צֳרִי | |
| קֳבֵל | |
| קֳדָם | |
| אֶל־קֳבָתָ֑הּ | φ1 |
| וְאֶת־קֳדָשָׁ֔יו |
At the start of a word, with a full vowel sign (φ2) rather than ḥaṭef (see Dotan, 1972, p. 241–247):
| קׇדָשִׁים | |
| קׇדָמוֹהִי | |
| קׇטׇנִּ֥י | |
| קׇבׇלּ֔וֹ | |
| קׇבׇל־עָ֖ם | |
| שׇׁרָשָׁיו | |
| סׇובָאִ֖ים | φ3 |
Within a word — on a letter with dagesh:
| אֶצֳּרֶֽנָּה׃ | ||
| תִקֳּבֶ֑נּוּ | ||
| כֻתֳּנֹ֔ת | ||
| הַגֳּרָנ֖וֹת | ||
| שִׁבֳּלִים | but | שִׁבֲּלֵ֣י |
| צִפֳּרִים | but Aramaic | צִפֲּרֵ֣י |
| סֻבֳּל֗וֹ |
On the second of a pair of letters with shewa:
| יֶהְדֳּפֶ֑נּוּ | |||
| φ4 | וָאֶשְׁקֳולָ֣ה | but | וָאֶשְׁקֲלָ֨ה |
| φ5 | קׇדְקֳד֔וֹ | ||
| הַבַּרְקֳנִֽים׃ | |||
| מׇרְדֳּכַי |
Footnotes for this section:
φ1 In ITM, אֶל־קֳבָתָ֑הּ has no accent. (I show atnaḥ.)
φ2 It is no coincidence that in all the examples below, this full vowel would be interpreted as qameṣ ḥaṭuf (qameṣ qaṭan), and indeed that is how we show it below (with the characteristic slightly taller shape used in some recent printed texts but of course not present in any manuscript tradition).
φ3 Most manuscripts and printed texts note this word’s superfluous waw, either compactly, e.g. ‹ יתיר ו̇ ›, or fully, as a ketiv of סובאים corresponding a qere of סׇבָאִ֖ים .
φ4 Most manuscripts and printed texts note this word’s superfluous waw, either compactly, e.g. ‹ יתיר ו̇ ›, or fully, as a ketiv of ואשקולה corresponding a qere of וָאֶשְׁקֳלָ֣ה .
φ5 In ITM no locale is given for קׇדְקֳד֔וֹ . I find it (only) at 2S 14:25.
Biblical references in this section: אֶל־קֳבָתָ֑הּ Nu 25:8, וְאֶת־קֳדָשָׁ֔יו 2K 12:19, קׇטׇנִּ֥י 1K 12:10, קׇבׇלּ֔וֹ Ez 26:9, קׇבׇל־עָ֖ם 2K 15:10, סׇובָאִ֖ים Ez 23:42, אֶצֳּרֶֽנָּה׃ Is 27:3, תִקֳּבֶ֑נּוּ Nu 23:25, כֻתֳּנֹ֔ת Ex 28:40, הַגֳּרָנ֖וֹת Joel 2:24, שִׁבֲּלֵ֣י Zech 4:12, צִפֲּרֵ֣י Dan 4:9, סֻבֳּל֗וֹ Is 9:3, יֶהְדֳּפֶ֑נּוּ Nu 35:20, וָאֶשְׁקֳולָ֣ה Ezra 8:25, וָאֶשְׁקֲלָ֨ה Ezra 8:26, קׇדְקֳד֔וֹ 2S 14:25, הַבַּרְקֳנִֽים׃ Jud 8:7.
In some cases (as noted above) ḥaṭef pataḥ occurs where this “morphological” ḥaṭef qameṣ is expected. Presumably this indicates that the original /o/ or /u/ coloring of the vowel was no longer audible, so it was treated as a ḥaṭef pataḥ (the normal sound of shewa). E.g.:
At the start of a word:
| קֲשִׁי֙ | μL1 |
On a letter with dagesh:
| מִֽסֻּבֲּכ֔וֹ |
On the second of a pair of letters with shewa:
| יִרְדֲּפֶ֑ךָ | |
| יִרְדֲּפֶֽךָ׃ | |
| אֶשְׁפֲּטֶֽךָ׃ | |
| אֶכְתֲּבֶ֑נָּה | μL |
The “morphological” use of ḥaṭef segol is rare in Hebrew — one example is וּֽבְצֶלְצֱלִֽים׃ (in μA and μC) but it is more common in Aramaic words. E.g.:
| בֱנַיְתַהּ֙ | ||
| קֱרִ֖י | ||
| גֱּלִ֣י | but | גְלִ֑י |
| וּמַדֱּקָ֔ה | ||
| מְמַלֱּלָ֑ה |
Biblical references in this section: קֲשִׁי֙ Dt 9:27, מִֽסֻּבֲּכ֔וֹ Jer 4:7, יִרְדֲּפֶ֑ךָ Ez 35:6, יִרְדֲּפֶֽךָ׃ Ez 35:6, אֶשְׁפֲּטֶֽךָ׃ Ez 35:11, אֶכְתֲּבֶ֑נָּה Jer 31:33 in some editions; others have it as 31:32, וּֽבְצֶלְצֱלִֽים׃ 2S 6:5, בֱנַיְתַהּ֙ Dan 4:27, קֱרִ֖י Ezra 4:18, גֱּלִ֣י Dan 2:30, גְלִ֑י Dan 2:19, וּמַדֱּקָ֔ה Dan 7:7, מְמַלֱּלָ֑ה Dan 7:11.
(ii) Ḥaṭef Shewa Used for Phonetic Reasons. Ḥaṭef qameṣ is the most common type of ḥaṭef shewa used for phonetic reasons. (Recall that it is also the most common type of ḥaṭef shewa used for morphological reasons.) It occurs mostly before a guttural with qameṣ. However, ḥaṭef qameṣ in these cases replaces a simple shewa which is vocal according to the standard rules, so that error in pronunciation would not be likely. Presumably the ḥaṭef sign was intended to prevent incorrect pronunciation resulting from lack of attention. E.g.:
At the start of a word:
| קֳהָת |
| וְכׇל־קֳהָלָ֔הּ |
On a letter with dagesh:
| הַקֳּהָתִי | |
| הַקֳּעָרֹ֤ת | φ1 |
| שִׂמֳּחָֽהוּ׃ | μL |
| פִתֳּחָ֣ה | μC |
| בְּסַאסֳּאָ֖ה |
On the second of a pair of letters with shewa:
| לְיׇקְמֳעָֽם׃ | |
| הִֽתְמַהְמֳהָ֑ם | |
| וְנִשְׁקֳעָ֖ה | φ2 |
Ḥaṭef pataḥ is used similarly, for phonetic reasons, in the following:
On a letter with dagesh:
| וַתְּצַדֲּקִי֙ |
On the second of a pair of letters with shewa:
| מִגְרֲשֵׁ֥י | |
| אַשְׁכֲּנַ֥ז | μL |
Footnotes for this section:
φ1 In ITM the locale Nu 4:6 is given for הַקֳּעָרֹ֤ת, but I only find it in the next verse, i.e. at Nu 4:7. Does Nu 4:6 reflect an alternate verse numbering tradition, or is it just a typo?
φ2 The ketiv of וְנִשְׁקֳעָ֖ה is ונשקה. We note this merely to be thorough; it is irrelevant to the point at hand.
Biblical references in this section: וְכׇל־קֳהָלָ֔הּ Ez 32:22, הַקֳּעָרֹ֤ת Nu 4:7, שִׂמֳּחָֽהוּ׃ Jer 20:15, פִתֳּחָ֣ה Is 48:8, בְּסַאסֳּאָ֖ה Is 27:8, לְיׇקְמֳעָֽם׃ 1K 4:12, הִֽתְמַהְמֳהָ֑ם Jud 3:26, וְנִשְׁקֳעָ֖ה Amos 8:8, וַתְּצַדֲּקִי֙ Ez 16:51, מִגְרֲשֵׁ֥י 1C 5:16, אַשְׁכֲּנַ֥ז Gen 10:3.
In some manuscripts, such as μA, ḥaṭef shewa is often used on non-guttural letters in all the categories described. In others, like μS, this is rare. Clearly the Masoretes considered the use of a ḥaṭef shewa sign to mark vocal shewa on a guttural as necessary, but on other letters as optional. Thus, for instance, the Diqduqe ha-Ṭeʿamim (Dotan 1967, section 19) says:
Some scribes, following a valid tradition, read ḥaṭef qameṣ in many places ... while others, also following a valid tradition, do not, but there is no (authoritative) source but the preference of the scribes.
Similarly the Horayat ha-Qore p. 64 (Dérenbourg, 1870, p. 372) says:
If one argues that the dalet of “Mordecai” (and other letters in other words) has ḥaṭef qameṣ, tell him, “but this sign is only a device used by some scribes to warn that the letters should be pronounced fully, and not slurred over.” Ḥaṭef qameṣ is written in some texts. It is not used in others, but the reader nevertheless pronounces the word in the same way when he comes to read it.
According to the rules given above, initial shewa before yod with ḥireq should sound like a very short ḥireq as /lĭyiśrɔʾel/ for לְיִשְׂרָאֵל . However, the pair of /i/ vowels separated by yod is likely to develop into a single long /i/ vowel, /līśrɔʾel/, and this gave rise to a systematic variation between ben Asher and ben Naftali. Ben Asher admits the long /i/ in only three words:
| וִֽילֲלַ֖ת | |
| לִֽיקְּהַ֫ת־אֵ֥ם | |
| כִּֽיתֲר֥וֹן | μS1 |
In these cases the shewa after the yod is vocal, so that the ḥireq occurs in an open syllable. In all other cases, bA marks the two vowels (shewa and ḥireq) separately. In contrast, bN requires the long /i/ vowel in these cases:
In all other cases, Ben Naftali does not require the long /i/ vowel, and so agrees with bA. Notable cases of agreement include words that start in the following two ways:
Footnotes for this section:
φ1 E.g. בִישְׂרָאֵ֗ל , כִּישְׂרָאֵ֔ל , לִישְׂרָאֵ֗ל .
φ2 In ITM the locale Jud 19:16 is given for וִיטַ֥ב but I do not find it (or anything close to it) there. I found וְיִטַ֥ב in 2K 25:24 so that is what I use above (with vowel adjustments).
Biblical references in this section: וִֽילֲלַ֖ת Jer 25:36, לִֽיקְּהַ֫ת־אֵ֥ם Prov 30:17, כִּֽיתֲר֥וֹן Qoh 2:13, וִיטַ֥ב 2K 25:24, לִירְאָתֶֽךָ׃ Ps 119:38, וִיחֵ֑לּוּ Job 29:21, בִישְׂרָאֵ֗ל Gen 34:7, כִּישְׂרָאֵ֔ל 2S 7:23, לִישְׂרָאֵ֗ל Hos 7:1.
Most of the early manuscripts conform to bA in the pointing of these words, but μC, and in the main μS, conform to bN.
There are a few words in the Bible pointed with shewa before yod with ḥireq where long /i/ would be expected. E.g.:
| וַיְּיִ֣ף |
| לְיִסּ֑וֹד |
(The expected pointings would be וַיִּ֣יף and לִיסּ֑וֹד.) This is possibly the result of over-correction: perhaps the long /i/ was avoided even where it should have been used.
In these situations, manuscripts with expanded Tiberian pointing use a system related to the system of bN. But these manuscripts use the pointing for long /i/ when any prefix precedes yod with ḥireq. Thus, in contrast to the real system of bN, in μR we find the following:
| וִיתֶּן־עֹ֣ז |
| וִירָ֖שׁ |
Biblical references in this section: וַיְּיִ֣ף Ez 31:7, לְיִסּ֑וֹד 2C 31:7, וִיתֶּן־עֹ֣ז 1S 2:10, וִירָ֖שׁ Is 57:13.