Recently, I was privileged to be part of fun yet esoteric discussion on matters of Hebrew grammar. First some background: there is a grammatical phenomenon in Biblical Hebrew known as “nasog ahor,” literally, “stepped back.” In certain words that are accented on the last syllable but have an earlier syllable that is open, then sometimes the accented is shifted to that earlier syllable if the proceeding, grammatically connected word is accented on one of its earlier syllables. Examples that many are familiar with include the blessing on the Torah, the word is ba-HAR, but when connected to the next word, BA-nu, it becomes “a-sher BA-har BA-nu.” Or on the blessing on the bread, the word would normally be ham-mo-TZI, but when connected to LE-hem it becomes ham-MO-tzi L-hem.
The second is a phenomenon that is often a consequesnce of the first, and it is not well known at all. “athei me’rahiq,” lit. “coming from afar,” and is when a word ends with an open, unaccented syllable vowelized with a qamatz or segol is joined to the proceeding word that is accented on the first syllable, placing a dagesh in the first letter of that latter word. The fact that the first word’s last syllable is unaccented may be due to the “stepped back” phenomenon described above. Examples that come to mind from the recent Torah readings include Genesis 30:33, w’A-n’tha BI, in which the beth has a dagesh, and 31:12, O-seh LACH, in which the lamed has a dagesh.
While the nasog ahor phenomenon makes sense to me, and interestingly enough, has its parallels in spoken English, for instance, I do not understand the latter phenomenon, nor am I aware of any explanation among the various authorities. However, based on the theories I outline in my book, (see the tab above) I can tolerate why this phenomenon of basically closing the final syllable of the first word, would happen only with the segol or qamatz. The segol is a t’nu’a q’tana, a minor or short vowel, and the only t’nu’a q’tana that occurs in open, accented syllables that end words, making it more versatile than the patah, the only other minor vowel that occurs in open or accented syllables, and because it does not have a natural semivowel at its end (the Y sound at the end of the long E and A sounds, or the W at the end of long O or U sounds), closing its syllable does not result in an unaccented consonant cluster, which as explicated by Gesenius, is not allowed. As for the qamatz, it is the only t’nu’a g’dola, major vowel, that does not have have a natural semivowel conclusion, and once again closing its syllable does not result in the formation of a consonant cluster, although this would then require us to explain why an ordinary qamatz is treated like the other major vowels if it is lacking this essential feature.
Like every rule, athei mer’hiq has its exceptions. For example, this week we read A-sa LO in Genesis 37:3 , and in that case, the lamed should have a dagesh, but it does not, or in 1:5, QA-ra LAY-la, and once again the lamed should have a dagesh, but it does not.
Last week, Dr. Marc Shapiro, k’darko baqodesh and blogs, released another must-read article on the Seforim blog. In it he made the following point:
In the ArtScroll siddur, p. 86 it reads:
ועל מאורי אור שעשית, יפארוך, סלה
There is a dagesh in the ס of סלה. This means that the comma after יפארוך is a mistake, as you cannot place a dagesh in this ס if preceded by a comma.
Dr. Shapiro’s assumptions in this matter are that the samech of sela receives a dagesh because of the athei me’rhiq rule, meaning that the previous word, y’fa-a-RU-cha, must be connected to it, and therefore it would be wrong to have a comma between the words. If there were a comma, then the samech would not receive a dagesh. It is then that I took issue with his argument, and wrote the following to him:
Actually you can have a dagesh. For example, אַ֭שְׁרֵי יֽוֹשְׁבֵ֣י בֵיתֶ֑ךָ ע֝֗וֹד יְֽהַלְל֥וּךָ סֶּֽלָה.
This is a well-known verse from the psalms.
Now, you might be initially inclined to dismiss this example, as in this case, sela is connected to the previous word by the trop, but the truth is that in the Sifrei Emet, what would normally be a mercha tip’ha (pause) siluq succession (in the other 24 books), does not and cannot exist when the word with the siluq is less than three whole syllables (or when accented before the last syllable, four whole syllables). Instead, what would be the tip’ha, the mafsiq, becomes the m’shareth of the siluq. For example, in Chronicles we have this well-known verse
הוֹד֤וּ לַֽיהוָה֙ כִּ֣י ט֔וֹב כִּ֥י לְעוֹלָ֖ם חַסְדּֽוֹ׃
But in the Psalms, because the word hasdo only has two syllables, the last three words are all connected, and hasdo is connected to the previous word:
:הוֹד֣וּ לַֽיהוָ֣ה כִּי־ט֑וֹב כִּ֖י לְעוֹלָ֣ם חַסְדּֽוֹ
The trop of the word ki is not the tip’ha-mafsiq that exists in the other books, rather, it is also a m’shareth:
The happens to also hold true for the etnah in Sifrei Emet, which, according to R’ Breuer, and as you can see from this example, has the weight of a zaqef of the other books, and also converts its mafsiq mishneh into a m’shareth when the word with the etnah is “short.” I have yet to figure out why this is. So, for example, if the verse lha’alot ner (mafsik) tamid were to be in Psalms, it would just be lha’alot ner tamid. Or the verse nagila w’nism’ha vo. In the Torah, it would have been nagila w’nism’ha (pause) Bo, but because bo is a short word, it is automatically connected to the previous. Every time the word sela appears at the end of a verse, it must be connected to the previous word, even if in context, the mafsiq that was supposed to be right before it would indicate the highest grammatical disjunction. For example, in the above verse, if sela were not there, it would be read “ashrei (pause) yosh’vei veithecha; od, y’hal’lucha (full stop).” And the same is true for basically every verse in Psalms that ends with sela.
:יְהוָ֣ה צְבָא֣וֹת עִמָּ֑נוּ מִשְׂגָּֽב־לָ֨נוּ אֱלֹהֵ֖י יַֽעֲקֹ֣ב סֶֽלָה
Logically, according to our accepted use of commas, all of those verses should have a comma, or perhaps even a period, right after the penultimate word. “The Lord of Hosts is with us; Our stronghold is the God of Jacob. Sela!” Yet, here sela is once again connected to the previous word.
So yes, if m’orei or she’asita, y’faarucha, sela, were a verse in Psalms, the last two words would be connected (because y’fa’arucha is accented on an early syllable and ends with a qamatz) and the samech would have a dagesh.
Dr. Shapiro had some follow up questions:
I see that you are assuming that a tipcha equals a comma (and let’s assume we are dealing with Tehillim). Leaving aside the issue of since when do siddurim insert commas before words like they did before סלה? I have not seen that anywhere. But is a tipcha really a comma?
He also complimented me, and I wrote the following response to him:
The answer is that tip’ha is sometimes a comma. The rule in the 21 ordinary books of scripture is that tip’ha is the mafsiq before the siluq. Every verse in those books has at the very least a siluq and and tip’ha.
This needs to be clarified. There are probably only a dozen or so verses wherein the last mafsiq before the siluq is an etnah, but in those cases, the etnah is preceded by a tip’ha.
However, the objective value of the tip’ha depends on the entire context of the particular verse. In a short verse, like “Adam Sheth Enosh” (I Chron. 1:1) it corresponds to absolutely no punctuation. In וַיָּ֥זֶד יַֽעֲקֹ֖ב נָזִ֑יד וַיָּבֹ֥א עֵשָׂ֛ו מִן־הַשָּׂדֶ֖ה וְה֥וּא עָיֵֽף׃, the second tip’ha has the value of what we would call a comma, while the first tip’ha (under Yaakov) [does not.] If the word sela would ever end any verse in the 21 books, it would be preceded by a tip’ha, by definition,
or perhaps the even stronger etnah,
and then you would have to examine the verse’s context and meaning to determine what ever mafsiqim are featured. As you know, in some verses, there aren’t even zaqefim, let alone etnahim, before the tip’ha, e.g. וְהָי֞וּ הַדְּבָרִ֣ים הָאֵ֗לֶּה אֲשֶׁ֨ר אָֽנֹכִ֧י מְצַוְּךָ֛ הַיּ֖וֹם עַל־לְבָבֶֽךָ׃
In contrast, the verses in Sifrei Emet have almost half a dozen ways of ending! Some have a revia mugrash where the tip’ha would be, some turn the final revia into a m’shareth (like the examples I showed you earlier), and some have a string of m’shar’thim with no mafsiqim, once again due to the shortness of the all the words: עֵֽינֵי־כֹ֭ל אֵלֶ֣יךָ יְשַׂבֵּ֑רוּ וְאַתָּ֤ה נֽוֹתֵן־לָהֶ֖ם אֶת־אָכְלָ֣ם בְּעִתּֽוֹ׃
My guess is that this has something to do with certain musical rules.
The sages noted that the verses of Sifrei Emet tend to be shorter than the verses in the rest of the Bible. I would also add that after careful analysis, including a thorough comparison of the verses and passages that appear in both (David’s victory song, for example), the verses of Emet tend to have less mafsiqim. I would really like to find someone with whom to work to understand these phenomena, but alas, I immigrated to Israel after R’ Breuer and R’ Kappah left us.
Therefore, in answer to your question, I insert the comma because that is how the verse is to be understood when translated, and that is how the English language and modern Hebrew work, but in the scriptural form, the laws of grammar/music dictate that the pause be subsumed due to the shortness of the word. My belief is that the siddur makers should always leave in the trop where ever possible, but certain liturgies are not biblical, and therefore the siddur makers try to help the reader by adding our Western conventions.
(75:4 is also a perfect example, by the way: נְֽמֹגִ֗ים אֶ֥רֶץ וְכׇל־יֹשְׁבֶ֑יהָ
אָנֹכִ֨י תִכַּ֖נְתִּי עַמּוּדֶ֣יהָ סֶּֽלָה
If we were to parse this verse using modern commas and periods, the period would be before the word sela. And that is how, for example, the JPS 1917 has it.)
Consider: How should one say “al tig’u bimshihai“? In Chronicles, there is a disjunctive pashta on the first word, leaving the dagesh in the following beth, but in the Psalms, tig’u has a conjunctive mercha, making the next word “vimshihai.” Or “Lo tirtsah, lo tinaf.” When read with one set of trop, the tavim are strong, while when read with the other set, the tavim are weak. When you are speaking or lecturing, or perhaps reciting those verses as part of your prayers but not as part of whole paragraphs, which set of trop do you use? I have no answer at the moment. But this doubt must exist when the liturgy adds sela to the end of a sentence. Do we read it as though it were one of the Psalms?
I hope that this will provoke further discussion of the matter, because I have been asking myself those last questions for years now, and have not found any direction.