Monday, June 06, 2011

How shall we pronounce the first וּבָאוּ in parashat Naso?

Summary: Is it mile'eil or mi'le-ra? I weigh in, considering the meaning of Minchas Shai.

Post: In parashat Naso, in the fifth perek of Bamidbar, there are three instances of the word וּבָאוּ:
במדבר פרק ה
  • פסוק כ"ב: וּבָאוּ הַמַּיִם הַמְאָרְרִים הָאֵלֶּה, בְּמֵעַיִךְ, לַצְבּוֹת בֶּטֶן, וְלַנְפִּל יָרֵךְ; וְאָמְרָה הָאִשָּׁה, אָמֵן אָמֵן. 
  • פסוק כ"ד: וְהִשְׁקָה, אֶת-הָאִשָּׁה, אֶת-מֵי הַמָּרִים, הַמְאָרְרִים; וּבָאוּ בָהּ הַמַּיִם הַמְאָרְרִים, לְמָרִים. 
  • פסוק כ"ז: וְהִשְׁקָהּ אֶת-הַמַּיִם, וְהָיְתָה אִם-נִטְמְאָה וַתִּמְעֹל מַעַל בְּאִישָׁהּ--וּבָאוּ בָהּ הַמַּיִם הַמְאָרְרִים לְמָרִים, וְצָבְתָה בִטְנָהּ וְנָפְלָה יְרֵכָהּ; וְהָיְתָה הָאִשָּׁה לְאָלָה, בְּקֶרֶב עַמָּהּ. 
How are these pronounced? According to Mechon-Mamre, based on Teimani sources:


The second two have trup positioned on the stressed syllable, because that is where the mercha and kadma are placed. And so, we know that they are mi-le'eil, with the stress on the second to last syllable. But what about the first one? A telishah gedolah is placed at the start of the word, to distinguish it from the telisha ketana -- this before the innovation of the fancy tails facing different directions. So, which is the stressed syllable? Sometimes on such words in some manuscripts, the telisha is repeated (just as the pashta is) on the stressed syllable. So where should the stress be?

This is important to know, for a baal koreih -- in fact, someone who lained last week asked me about this, on the basis of a Minchas Shai and various other evidence he encountered, which we will get to, beEzras Hashem.

At Temanim.org, in their Chumash with Tafsir, Onkelos and Rashi, they have the same -- a telisha gedolah at the start but not repeated.

So too the Leningrad Codex:

Or, looking at the actual images:



Similarly, Codex Hilleli only has the one telisha:

And so too the Lisbon Codex:

So too the following Chumash: (You need the djvu plugin to see these inside.) 


1482
תנ"ך. תורה. רמ"ב. בולוניה

[בולונייא : דפוס יוסף בן אברהם קרוויטה ; אברהם בן חיים מן הצבועים, רמ"ב].


The first one I thought had a telisha, but actually did not, is in this chumash:

1491
תנ"ך. תורה. רנ"א. ליסבון
(אשבונה : דפוס אליעזר [טולידאנו], אב רנ"א).


However, this first one is a telisha gedola while the second is a simple circle, for a masoretic note! That masoretic note, on the side, is as follows:
That is, אין בבית תלשא. The second note is on the word המאררים, and it says האלפ בקמץ. And so, there is a masoretic note telling us that there is no repetition of the telisha.

How confusing, since as I noted, the original sign of telisha gedola and of ketana is just the circle, which is why it is sometimes called tarsa, 'shield'! Indeed, looking at Dr. William Wickes, he makes the following point:


Yet here, by making a masoretic note to note its absence, he confounded me! And perhaps in earlier, hand-written texts, it confounded others. I'll save this for the later discussion section, because for now I just want to consider the evidence.

Bomberg's first Mikraos Gedolos

1518
תנ"ך. רע"ח. ונציה
ויניציאה : דניאל בומבירגי, רע"ח.

has just the one telisha gedola. Follow the link to see the direct page. Since these present no new interesting information, I am not putting up pictures.

So too his second Mikraos Gedolos,

1525
תנ"ך. רפ"ה. ונציה
שער יהוה החדש : ... החומש עם תרגום ופי' רש"י ון' עזרא והנביאים ראשונים עם פי' רש"י וקמחי ורלב"ג והנביאים אחרונים ישעיה עם פי' רש"י ואבן עזרא. ירמיה ויחזקאל עם פי' רש"י וקמחי, תרי עשר עם פי' רש"י ואבן עזרא, והכתובים תולים עם פי' רש"י ואבן עזרא, משלי עם פי' אבן עזרא [ר' משה קמחי], ורלב"ג, איוב עם פי' אבן עזרא ורלב"ג, דניאל עם פירוש אבן עזרא ורבינו סעדיה גאון, עזרא עם פי' אבן עזרא [ר' משה קמחי], ורש"י, דברי הימים עם פי' מיוחס לרש"י, חמש מגילות עם פירוש רש"י ואבן עזרא ... / נערך בידי יעקב בן חיים מטוניס עם הקדמה ממנו.
ויניצייא : דפוס ד. בומבירגי, (רפ"ה-רפ"ו).


So too Bomberg's Chumash:

1524
תנ"ך. תורה. רפ"ד. ונציה
ויניציאה : דפוס ד' בומבירגי, רפ"ד.


So too this one:

1521
תנ"ך. תורה. רפ"א. שלוניקי
(שאלוניקי : [דפוס יהודה גדליה], רפ"א).


So too this one:

1546
תנ"ך. תורה. ש"ו. קושטא
[קושטאנדינא : א' שונצינו], (ש"ו).


So too this one:

1547
תנ"ך. תורה. ש"ז. קושטא
קושטנדינה : א. שונצין, ש"ז.


So too this one:

1680
תנ"ך. תורה. ת"ם. אמשטרדם
אמשטרדאם : דפוס אורי בן אהרן הלוי, ת'ה'ל'ה' [ת"מ].


So too this one:

1703תנ"ך. תורה. תס"ג. ברלין
חמשה חומשי תורה : ... מקושטים ומלובשים בעשרה לבושי אורה ... והמה ... תרגום אונקלוס ותרגום יונתן ב"ע [בן עזיאל] ותרגום הירושלמי ... רש"י והרשב"ם והראב"ע ... המסורת ... ופירוש הרב בעל הטורים ... ובעל תולדת אהרן ... פי' הרד"ק על ההפטרות ...
ברלין : [חמו"ל], תס"ג-תס"ח.


So too this one:

1742תנ"ך. תק"ב. מנטובה
ספר ארבעה ועשרים : עם תוספת ... / מהרב ... ידידיה שלמה מנורצי ... קראתי שמו מנחת שי ... ; הוגה ... על ידי ... כמהר"ר רפאל חיים באזילה ...
מנטובה : דפוס רפאל חיים מאיטליא, [תק"ב]-[תק"ד].


Here is an interesting one. It is in Mendelsohnn's Nesivos Hashalom:

1783תנ"ך. תורה. תקמ"ג. ברלין
ספר נתיבות השלום : והוא חבור כולל חמשת חמשי התורה עם תקון סופרים / [מאת שלמה בן יואל מדובנא ושלום בן יעקב הכהן] ; ותרגום אשכנזי [מאת משה מנדלסון] ; ובאור [מאת משה מנדלסון, שלמה בן יואל מדובנא, אהרן ירוסלב והירץ הומברג].
ברלין : דפוס G. F. STARCKE, תקמ"ג.


He just has a single telisha, as above. But he writes, in Tikkun Soferim, at the bottom -- written by R' Shlomo Dubno (also, see here):

"There is a single telisha in the vav and there is not another in the beis -- Ohr Torah and Minchas Shai. And it is pronounced on the beis."

This one has a single telisha, like the others:

1803תנ"ך. תקס"ג. פיזה
ארבעה ועשרים : עם פירוש חומת אנ"ך ... והוא קצת פי' על איזה פסוקים תורה נביאים וכתובים ... / אמר ... חיים יוסף דוד אזולאי ...
פיסא : שמואל מולכו, ובגפן שלשה ש'ר'י'ג'י'ם' [תקס"ג].


Here is an interesting one, organized by Wolf Heidenheim. In this instance, there is repetition of the telisha!

This from:

1818תנ"ך. תורה. תקע"ח. רדלהים
חומש מאור עינים : והיו לאורות ס' עין הקורא וס' עין הסופר / מדויק .. ומסדר ... מאתי וואלף ... היידנהיים.
רעדלהיים : ו' היידנהיים, תקע"ח-תקפ"א.


This second telisha is positioned over the beis.

Next, another with just a single telisha:

1824תנ"ך. תורה. תקפ"ד. סלויטה
חמשה חומשי תורה : עם תרגום אונקלוס ותרגום יונתן ... ותרגום ירושלמי עם פירושיהם ... ועם פי' רש"י ופי' אור החיים / הכינו ... מוהר"ר חיים ן' עטר ...
סלאוויטא : דפוס ש"א שפירא, תקפ"ד-תקפ"ה.

Another, with a single telisha:

1825תנ"ך. תורה. תקפ"ו. אוסטרה
חמשה חומשי תורה : עם תרגום אונקלוס ... פי' רש"י ופי' פנים יפות על דרך פרד"ס / נובעים ממעין ... הגאון ... מור' פנחס בן צבי הירש הורוויץ ... ; עם פירוש ... שם אפרים מהרב הגאון מוה' אפרים זלמן מרגליות ...
אוסטרהא : אברהם קלאהר פיין ודוד בן יצחק מרדכי, תקפ"ה-תקפ"ו.


This exhausts all the ones on JNUL.

In this printing of Shadal's commentary on a chumash (though the Chumash text with trup is NOT from him), we have:

So too, at Bible.ort.org, which I think is associated with the Living Torah chumash:

This modern Mikraos Gedolos only has the one telisha:

Or Torah, R' Menachem de Lunzano writes:
ובאו  אין  בבית תלשא

Minchas Shai writes:
ובאו ־ תלשא אחת בוא״ו ואין אחרת בבית

Masoret HaKeriah (R' Yehuda Leib Shuslowitz) writes:
ובאו ,  הנגינה  בבי״ת

(I don't see Rabbi Meir Abulafia, the Rama, discussing it.)

Looking at Targum Onkelos, we see:

ה,כב וּבָאוּ הַמַּיִם הַמְאָרְרִים הָאֵלֶּה, בְּמֵעַיִךְ, לַצְבּוֹת בֶּטֶן, וְלַנְפִּל יָרֵךְ; וְאָמְרָה הָאִשָּׁה, אָמֵן אָמֵן.וְיֵיעֲלוּן מַיָּא מְלָטְטַיָּא הָאִלֵּין, בִּמְעַכִי, לְאַפָּחָא מְעִין, וּלְאַמְסָאָה יִירַךְ; וְתֵימַר אִתְּתָא, אָמֵן אָמֵן.


This translation is identical to the other ones in Onkelos for uva`u in this section, where if it was stressed differently, we might have expected a different translation.

There is only one explicit piece of contradictory evidence. The same Baal Koreh who asked me the question and pointed out the Minchas Shai also showed me the following: In the chumash with the peirush of Rav Shamshon Refael Hirsch, they have two telishas, and the second is on the aleph of  וּבָאוּ ! That would mean that the stress is on the last syllable! He also conjectured a plausible translation, if it were different. Something along the lines of that 'the water would have entered', rather than 'the water shall enter'. For the difference between mile'el and milera is a difference of tense.

I agree that we can adequately explain any sort of tense if need be, such that we cannot use it as absolute evidence to reject a reading. It is relevant that Onkelos does not vary in his rendition, above. But what we really want to do is look at the traditional masoretic texts.

The problem is one of ambiguity. Most of these texts do not have the second telisha. Where the telisha does repeat in a word, it is to indicate stress, but here, we have no indication of stress, in all the good masoretic texts I have access to. (E.g. Leningrad Codex.)

Similarly, there is ambiguity in the words of the masoretes, namely Or Torah and Minchas Shai. Yes, they say that there is no second telisha in the beis, but just what do they mean by that? I could interpret it in one of a few ways:

  1. Minchas Shai is just commenting on the orthographic tradition. While he readily admits that it is pronounced with stress on the letter ב, it is important to preserve the textual tradition, which developed for whatever reason it developed.
  2. Minchas Shai is also commenting on the tradition of pronunciation. If the second telisha is not present, where it usually informs us of where to stress the word, perhaps we would revert to the standard method of pronunciation, and stress the last syllable.
Here are some further considerations. The other two in context are stressed mile'eil, so perhaps the purpose of commenting is to distinguish this pronunciation from the others. On the other hand, there is no reason to assume that the meaning changed. Another thought: In the very next pasuk, the first word, וכתב, also has a telisha gedola. And it is surely pronounced milera. So why don't Or Torah and Minchas Shai comment on that one? Presumably because there is no need to comment on that one. But here, where the stress is strange, where we would have expected the repeated orthographic symbol, we have to note the divergence. (Strange, though, might be strangeness in being mile'eil unlike most Hebrew words, or strangeness in diverging from the other two ובאוs in context.) Also note that neither explicitly mention that is is mile'eil or milera.

However, I would conclude that there should be penultimate stress (mi-le'eil) on this word, rather than ultimate stress (mi-le-ra). Why?
  1. In fact, as we see from the words of Dr. William Wickes, Masoretes (really, punctuators) often avoid repeating the telisha, and in fact, they were placed before the words, in order to avoid confusion with a masoretic note. This would be the case whether there was ultimate or penultimate stress on the word. And so its absence should not be evidence that it has ultimate stress. So the comment of Ohr Torah and Minchat Shai could well be taken as referring simply to the orthography.
    .
  2. That is how Shlomo Dubno interprets Minchas Shai and Or Torah explicitly in Tikun Soferim, and how Wolf Heidenheim explicitly arranges it in his Chumash. And I can point to two other Chumashim which have this.
    .
  3. Yes, there is this chumash with the commentary of R' Shamshon Refael Hirsch, but who is responsible for the trup in this? It is a pretty late printing, and all the evidence I could find in looking at early codices and chumashim had no second telisha or (later on), a telisha on the beis to indicate the stress, something which need not contradict what Minchas Shai says. If so, my guess is that the printer put this in based on sevara, and then on his own weight of authority. And we have others who I think would outweigh him in authority / knowledge / sevara, among those who explicitly write otherwise. (I freely admit ignorance as to who arranged the trup in this Chumash.) Therefore, I would toss out this evidence as an error. Indeed, as an error I would have made myself, on the basis of sevara.

There are presumably other sources which discuss this. I am going to put this up to the Google Leining group. Perhaps more recent masoretic scholars have discussed this.

But at the moment, my conclusion is that it should be uva'u, rather that uva'u.

8 comments:

yaak said...

I've always pronounced it Mil'eil.

You should do a post for parshat Shelah Lecha on whether the word סלח-נא is mil'eil or milra because I've seen it both ways.

Josh M. said...

The mesorah seems fairly clear-cut, but I don't understand why the word would be mil'el. I understand the latter two are nesog achor, but what's the technical reason for the accent on the first one?

joshwaxman said...

before proceeding, you mean clear cut that it is mile'eil?

if so, are you then saying that under the rules of dikduk, we should have expected the former to be milera?

thanks,
josh

joshwaxman said...

Some more data:
Shemot 7:28:
http://kodesh.snunit.k12.il/i/t/t0207.htm#28
כח וְשָׁרַ֣ץ הַיְאֹר֮ צְפַרְדְּעִים֒ וְעָלוּ֙ וּבָ֣אוּ בְּבֵיתֶ֔ךָ וּבַֽחֲדַ֥ר מִשְׁכָּֽבְךָ֖ וְעַל־מִטָּתֶ֑ךָ וּבְבֵ֤ית עֲבָדֶ֨יךָ֙ וּבְעַמֶּ֔ךָ וּבְתַנּוּרֶ֖יךָ וּבְמִשְׁאֲרוֹתֶֽיךָ׃

this is not a nasog achor, i think since the stress of the following word 'beveitecha' is not on the first syllable, and yet the stress is on the bet.

also, devarim 28:2:
http://mechon-mamre.org/c/ct/c0528.htm
ב וּבָ֧אוּ עָלֶ֛יךָ כָּל־הַבְּרָכ֥וֹת הָאֵ֖לֶּה וְהִשִּׂיגֻ֑ךָ כִּ֣י תִשְׁמַ֔ע בְּק֖וֹל יְהוָ֥ה אֱלֹהֶֽיךָ׃

again, with stress on the bet.

and 28:15:
טו וְהָיָ֗ה אִם־לֹ֤א תִשְׁמַע֙ בְּקוֹל֙ יְהוָ֣ה אֱלֹהֶ֔יךָ לִשְׁמֹ֤ר לַֽעֲשׂוֹת֙ אֶת־כָּל־מִצְו͏ֹתָ֣יו וְחֻקֹּתָ֔יו אֲשֶׁ֛ר אָֽנֹכִ֥י מְצַוְּךָ֖ הַיּ֑וֹם וּבָ֧אוּ עָלֶ֛יךָ כָּל־הַקְּלָל֥וֹת הָאֵ֖לֶּה וְהִשִּׂיגֽוּךָ׃

and 28:45:
מה וּבָ֨אוּ עָלֶ֜יךָ כָּל־הַקְּלָל֣וֹת הָאֵ֗לֶּה וּרְדָפ֨וּךָ֙ וְהִשִּׂיג֔וּךָ עַ֖ד הִשָּֽׁמְדָ֑ךְ כִּי־לֹ֣א שָׁמַ֗עְתָּ בְּקוֹל֙ יְהוָ֣ה אֱלֹהֶ֔יךָ לִשְׁמֹ֛ר מִצְו͏ֹתָ֥יו וְחֻקֹּתָ֖יו אֲשֶׁ֥ר צִוָּֽךְ׃

more in a bit. i don't have an explanation ready just yet, but this mile'eil does seem to be the pattern in general, even without nasog achor.

kol tuv,
josh

Milhouse said...

R Michoel Slavin can be heard here (at about 11:20) reading ובאו mil'eil, and וכתב mil'ra.

joshwaxman said...

yaak, milhouse:
thanks.

Josh M.:
I'll try to explain, but I am not an expert on dikduk. the reason it is mile'eil is that it is a hollow root, בוא. if the root were באה, then it would be milera. this is apparently the way of distinguishing the two in Biblical Hebrew. see this earlier parshablog post on Minchas Shai and SHAtu vs. shaTU.

thanks for putting the question out like that. i think this actually resolves a lot of the difficulties in this; i'll try to explain in a follow-up post.

kt,
josh

Anonymous said...

I tried to post a reply on the Leining blog but it didn't take.

This word is always mileil throughout Tanach.

Nachei-ayin ("hollow") verbs have their accent on the first root letter (with the exception being s/f 3rd-person present tense (BaAH vs. BAah) to differentiate it from past tense.

Vav Hahipuch does not operate on nachei-ayin verbs in the pl 3rd-person past tense, so this remains mileil even though there is a vav hahipuch here and it is in future tense.

Josh M. said...

That resolves my question, thanks.

LinkWithin

Blog Widget by LinkWithin