Thursday, June 25, 2009

Did Moshe not take a single donkey? Or a donkey of a single one of them?

In Moshe's complaint in parshat Korach, we have:
טו וַיִּחַר לְמֹשֶׁה, מְאֹד, וַיֹּאמֶר אֶל-ה', אַל-תֵּפֶן אֶל-מִנְחָתָם; לֹא חֲמוֹר אֶחָד מֵהֶם, נָשָׂאתִי, וְלֹא הֲרֵעֹתִי, אֶת-אַחַד מֵהֶם.15 And Moses was very wroth, and said unto the LORD: 'Respect not Thou their offering; I have not taken one ass from them, neither have I hurt one of them.'
But how should we translate chamor echad meihem? The middle word, echad, could associate either with the previous or the next word.

The trup is no help, since chamor has a kadma and echad has a mahpach. Both of these are mesharetim, conjunctive trup, and so there is no pause there to help us parse it.

Later in the pasuk, we have achad mehem, where it means "one of them," so perhaps we should take the earlier run of these two words the same. On the other hand, the vowels at the end of the pasuk are of the pattern patach patach, so perhaps they are different.

Rashi takes it as "donkey of one of them." Thus,
I have not taken a donkey from a single one of them: I did not take a donkey from any one of them. Even when I went from Midian to Egypt, and I placed my wife and sons on a donkey to ride, and I should have taken that donkey from their property, I took only from my own property (Tanchuma Korach 7, Num. Rabbah 10). Onkelos renders it as שְׁחָרִית, ‘expropriated.’ In Aramaic, the king’s service is called שַׁחְוַור.
לא חמור אחד מהם נשאתי: לא חמורו של אחד מהם נטלתי. אפילו כשהלכתי ממדין למצרים והרכבתי את אשתי ואת בני על החמור, והיה לי ליטול אותו החמור משלהם, לא נטלתי אלא משלי. ותרגום אנקלוס, שחרית, לשון ארמי, כך נקראת אנגריא של מלך שחוור:

And so do all the Targumim render it.

In contrast, Ramban takes it as "a single donkey of theirs." Thus:
וטעם לא חמור אחד מהם נשאתי
אמר מה שררה אני משתרר עליהם, כי מעולם לא לקחתי מהם אפילו חמור אחד לעשות צרכי כדרך המלכים או השרים, כי זה משפט המלוכה דכתיב (ש"א ח טז): ואת חמוריכם יקח ועשה למלאכתו. וזה טעם של אונקלוס"שחרית", שכך נקרא אנגריא של מלך שחוור (ב"ב מז א). הזכיר הדבר הקטן שבדיני מלכות.
Rashbam makes the point that the nikkud should help disambiguate, and that it is echad rather than achad means that it is one donkey, not a donkey of one of them. Thus:
לא חמור אחד מהם נשאתי -
אפילו חמור אחד לא נשאתי מהם כשאר משתוררים על העם. ואם כן למה מפשיעים אותי בשררה?!
ואם היה נקוד לא חמור אַחַד מהם (בפת"ח) פתרונו היה חמור של אחד מהם כמו: אַחַד העם, וכמו: את אַחַד מהם.
And here is where I first saw it, in Chizkuni:
chamor echad: According to its vowel points, its explanation is in accordance with its {plain} implication. However, according to the explanation of Rashi it should have been vocalized like achad haam {at the end of the pasuk}.
It seems that the idea is that where אחד precedes the word it modifies, it should be achad, while if it follows the word it modifies, it should be echad.

I did a search for אחד מהם to see how the evidence stacks up. We have:
במדבר פרק טז
  • פסוק טו: וַיִּחַר לְמֹשֶׁה, מְאֹד, וַיֹּאמֶר אֶל-יְהוָה, אַל-תֵּפֶן אֶל-מִנְחָתָם; לֹא חֲמוֹר אֶחָד מֵהֶם, נָשָׂאתִי, וְלֹא הֲרֵעֹתִי, אֶת-אַחַד מֵהֶם.
which is the pasuk under consideration. We also have:
דברים פרק כה
  • פסוק ה: כִּי-יֵשְׁבוּ אַחִים יַחְדָּו, וּמֵת אַחַד מֵהֶם וּבֵן אֵין-לוֹ--לֹא-תִהְיֶה אֵשֶׁת-הַמֵּת הַחוּצָה, לְאִישׁ זָר: יְבָמָהּ יָבֹא עָלֶיהָ, וּלְקָחָהּ לוֹ לְאִשָּׁה וְיִבְּמָהּ.
שמואל ב פרק יג
  • פסוק ל: וַיְהִי, הֵמָּה בַדֶּרֶךְ, וְהַשְּׁמֻעָה בָאָה, אֶל-דָּוִד לֵאמֹר: הִכָּה אַבְשָׁלוֹם אֶת-כָּל-בְּנֵי הַמֶּלֶךְ, וְלֹא-נוֹתַר מֵהֶם אֶחָד.
מלכים א פרק יט
  • פסוק ב: וַתִּשְׁלַח אִיזֶבֶל מַלְאָךְ, אֶל-אֵלִיָּהוּ לֵאמֹר: כֹּה-יַעֲשׂוּן אֱלֹהִים, וְכֹה יוֹסִפוּן, כִּי-כָעֵת מָחָר אָשִׂים אֶת-נַפְשְׁךָ, כְּנֶפֶשׁ אַחַד מֵהֶם.
מלכים א פרק כב
  • פסוק יג: וְהַמַּלְאָךְ אֲשֶׁר-הָלַךְ לִקְרֹא מִיכָיְהוּ, דִּבֶּר אֵלָיו לֵאמֹר, הִנֵּה-נָא דִּבְרֵי הַנְּבִיאִים פֶּה-אֶחָד טוֹב, אֶל-הַמֶּלֶךְ; יְהִי-נָא דבריך (דְבָרְךָ), כִּדְבַר אַחַד מֵהֶם--וְדִבַּרְתָּ טּוֹב.
דברי הימים ב פרק יח
  • פסוק יב: וְהַמַּלְאָךְ אֲשֶׁר-הָלַךְ לִקְרֹא לְמִיכָיְהוּ, דִּבֶּר אֵלָיו לֵאמֹר, הִנֵּה דִּבְרֵי הַנְּבִאִים פֶּה-אֶחָד טוֹב, אֶל-הַמֶּלֶךְ; וִיהִי-נָא דְבָרְךָ כְּאַחַד מֵהֶם, וְדִבַּרְתָּ טּוֹב.
In each of these cases, achad meihem means one of them, and meihem echad means the same, but has echad after the modifying word. However, we have the nikkud of asher yatzar, which is echad meihem, rather than achad meihem. And we also have a famous pasuk in Tehillim:

תהילים פרק קו
  • פסוק יא: וַיְכַסּוּ-מַיִם צָרֵיהֶם; אֶחָד מֵהֶם, לֹא נוֹתָר.
How do we justify Rashbam and Chizkuni's statement, in light of that pasuk in Tehillim? Maybe I am misunderstanding their point...

Let me close with a citation from the LXX translation on this pasuk. It is messed up in other ways, but:
15 And Moses was exceeding indignant, and said to the Lord, Do thou take no heed to their sacrifice: I have not taken away the desire of any one of them, neither have I hurt any one of them.
This is like Ramban and Rashbam.

This is not to say that Rashi cannot decide to interpret a pasuk against the nikkud. Shadal points out that Rashi does this on occassion, and ascribes this not to accident or concern with drash over peshat, but to a position that nikkud is post-Sinaitic and that therefore one is entitled to argue against it to determine the peshat. But maybe we can simply say that Rashi was not convinced that this was the intent of the nikkud, given the example in Tehillim.

Update: As Anonymous (please choose a pseudonym, because otherwise it is very hard to keep track of who is who!) points out in the comment section, a classic distinction made by baalei dikduk between echad and achad is that echad is meyuchad, a particular one, while achad is "any one." Perhaps this is akin to a bound and unbound variable in logic. I believe that one must be careful not to take it too precisely, to make meyuchad into some known entity, because that might cross the line from dikduk to midrash. But tentatively, I can see how this works with many pesukim.

If so, then it is a strong possibility that this is Chizkuni's point, and Rashbam's point. "I did not take a single donkey of theirs" is a case of meyuchad. You cannot point to any particular donkey that happened to belong to them that I took. As opposed to "I did not take a donkey of any of them." So this is what Chizkuni and Rashbam mean by the nikkud.

Meanwhile, Rashi has the context at the end of the pasuk, where we would expect the same bigram to mean the same thing is such close proximity.

In terms of trup, I don't think that Anonymous' comment is convincing. (See below in the comment section.) Indeed, I was toying with various examples of disjunctive trup in the middle of it, but shortened my dismissal of the trup above from what was originally a paragraph to a sentence or to, about why it was not helpful. Yes, I agree that had the trup been different, it could help parse, such that e.g. a pashta one word earlier could make it, on echad, would make it:
Not a single donkey | of theirs did I take.

But I am not convinced that the absence of such disjunctive trup proves the opposite. Because we can set it up as:
What did I not take? a single donkey of theirs.
or
What did I not take? a donkey of any of them.

5 comments:

Anonymous said...

The trup IS of great help, but not on the words you discuss. "Meihem" has a disjunctive trup, thus separating it from the word "nasasi". This is surely part of Rashi's reasoning. According to the Rashbam, it should have read "chamor[mahpach] Echad[pashta] meihem[munach] nasasi[zakef]".

I have not seen the Shadal you quote, but I would not put words into Rashi's mouth as far as nikkud being post-Siniatic or not; this is an ancient dispute, and there's no need to drag Rashi into it. Rashi's commentary relies heavily on dikduk/nikkud and is always the more accurate from that standpoint.

In this instance, Rashi is very much in line with the nikkud (and the trup as well, which I explained above). The difference between "echad" and "achad" is as follows: "echad" means "a particular one" (in the language of the ba'alei dikduk, a "meyuchad") and "achad" means "any one". That is why "achad" has a plural ("achadim") and "echad" does not; "echad" implies uniqueness.

Thus Rashi was unwilling to translate as the Rashbam/Chizkuni do. There is nothing special about the donkey; Moshe wouldn't be saying "I didn't take a particular donkey". Therefore "echad" must be modifying "from them", meaning "I didn't take a donkey from any paricular/special one of them", meaning, the community chest which is set up to support those in need. Which is EXACTLY why Rashi continues:
"אפילו כשהלכתי ממדין למצרים והרכבתי את אשתי ואת בני על החמור, והיה לי ליטול אותו החמור משלהם, לא נטלתי אלא משלי."
Rashi takes this from the Medrash, which elaborates that Moshe was the community leader and thus should have been paid from the community chest, as is commonplace; nonethless he refused to take anything. That is the particular thing to which "echad" refers, and that is why Rashi quotes this medrash here (otherwise, doesn't it seem out of place? What does this incident have to do with Moshe coming from Midyan?).

See "Havanas HaMikrah" from Heidenheim who elaborates.

Finally, with respect to the nikkud of Asher Yatzar, the RZ"H in Shaarei Tefillah says that the correct nusach should indeed be "achad", since we are not referencing a particular body part, but "any" body part, as explained above.

joshwaxman said...

the disjunctive trup on meihem is indeed plausible as proof. good point, though i need to think more about it when i can actually think.

i don't agree that אֲחָדִים is most certainly the plural of Achad to the exclusion of Echad. How can you tell this?! The vowel under the aleph is a chataf-patach, which is the equivalent of a sheva, except that it is a gutteral. what would you expect the segol to become? yeled (of an admittedly different pattern) becomes yeladim. how do you know the chataf patach comes from achad and not echad?

i have not seen Heidenheim inside. it sounds like an interesting idea. but it sounds kind of midrashic to me. (echad as meyuchad.) i'll look at the evidence, but what i see so far is not so convincing. but again, i am not thinking straight.

כִּמְעַט שָׁכַב אַחַד הָעָם, אֶת-אִשְׁתֶּךָ
sure, this could mean any random person, but surely it is one person. why is it different from one locust?

וְזֶה הַדָּבָר אֲשֶׁר-תַּעֲשֶׂה לָהֶם, לְקַדֵּשׁ אֹתָם--לְכַהֵן לִי: לְקַח פַּר אֶחָד בֶּן-בָּקָר, וְאֵילִם שְׁנַיִם--תְּמִימִם
one individual cow?

אֵיכָה יִרְדֹּף אֶחָד, אֶלֶף
is a particular person intended?

but we have in Shofetim 17:
וְהָאִישׁ מִיכָה, לוֹ בֵּית אֱלֹהִים; וַיַּעַשׂ אֵפוֹד, וּתְרָפִים, וַיְמַלֵּא אֶת-יַד אַחַד מִבָּנָיו, וַיְהִי-לוֹ לְכֹהֵן.
was it not a single individual of his sons?

כִּי הִנֵּה הָאֶבֶן, אֲשֶׁר נָתַתִּי לִפְנֵי יְהוֹשֻׁעַ--עַל-אֶבֶן אַחַת, שִׁבְעָה עֵינָיִם; הִנְנִי מְפַתֵּחַ פִּתֻּחָהּ
isn't this a particular stone? why achas?

i'm not convinced yet that it is borne out, but i will think about it. it is an interesting theory, though. but even if so, whatever the particular fine distinction is, it would seem that one can kvetch examples towards one or the other, such that we might be overapplying the distinction.

at the very least, it does not seem that rashbam is thinking along those terms.

i don't like the "what is bothering rashi" questions. rashi cites midrash more than 80% of the time. and midrash plays on connections. rashi is citing a midrash tanchuma which is found on parshat korach. how is citing a midrash local to the parsha out of place? that midrash reads:
לא חמור אחד מהם נשאתי
מה שהיה דרכי ליטול, לא נטלתי מהם.
בנוהג שבעולם, אדם שהוא בהקדש, נוטל שכרו מן ההקדש. ואני בשעה שהייתי יורד ממדין למצרים, היה דרכי ליטול מהן חמור, שבשביל צרכיהם ירדתי, ואף על פי כן לא נטלתי.

וכן שמואל הצדיק אומר: הנני ענו בי נגד ה' ונגד משיחו, את שור מי לקחתי וחמור מי לקחתי (ש"א יב ג), שור שהייתי מקריב עליהם מקורבנותיהם ומבקש רחמים עליהם, וכן למשוח עליהם מלך משלי היה, שנאמר: עגלת בקר תקח בידך ואמרת לזבוח לה' באתי (שם טז ב).
וכן הוא אומר: כי זבח היום לעם בבמה (שם ט יב). ולא נטלתי משלהם.
ובשעה שהייתי חוזר ועושה דיניהם וצרכיהם, והולך וסובב על עיירות ישראל, שנאמר: והלך מדי שנה בשנה וסבב בית אל והגלגל והמצפה ושפט את ישראל (שם ז טז).
דרך העולם, בעלי דינים הולכין אצל הדיין. ואני, הולך וסובב מעיר לעיר וממקום למקום.

the midrash seeks for a specific instance in which Moshe used a donkey, found this one, and makes the connection. and rashi cites it not necessarily because he is bothered by anything grammatical. and you could ask a similar question on the hundreds of midrashim rashi brings down. (and find false answers, because grammatical irregularities or possible reinterpretations are often the fodder for midrash.)

regardless, i do not think that rashi is more particularly focused on nikkud and trup than, say, ibn ezra. and more than that, he has a tendency towards reworking midrash into his perush.

if you want to read Shadal on Rashi (and many other meforshim) on this point, you can see him here:
http://parsha.blogspot.com/2007/12/shemot-vikuach-al-chochmat-hakabbalah.html

where he has several examples of Rashi going against the nikkud and trup.

kt,
josh

joshwaxman said...

also, i don't buy echad meihem as specifically a community chest. indeed, the fact that we are identifying a particular instance of taking a donkey, one could argue that echad as meyuchad should apply to the donkey.

in terms of the trup, certainly if it was like that, it would be clear. but in the absence of that, I am not sure it is definitive.

what did I not take? one donkey of theirs.

what did I not take? a donkey of one of them.

true, a parse of:
what did I not take from them? one donkey.
would be clearer. but the first two parses seem possible.

which was why i was not certain in the first place that the absence of disjunctive trup tells us anything.

kt,
josh

joshwaxman said...

also, taking your achad vs. echad distinction you put out as true... then, this is likely the very argument that Chizkuni and Rashbam are making, rather than about position, like i was guessing above.

that is, a single donkey would be a meyuchad one. not a single injust like "there was not a single locust left," or choosing a particular cow for an olah, or "echad mehem no notar," not a single one of them remained. but if it is a donkey of any of them, then we are not highlighting any individual, but it would mean "not a donkey of *any* of them." and then we should have achad, not echad. this would work out quite well with rashbam and chizkuni, but not so well with rashi.

thanks!
kt,
josh

joshwaxman said...

see my updated post. thanks again. your comments have been extremely helpful in clarifying this. (the only thing I could ask is that you choose a pseudonym for the future, so that i can tell you from the other anonymous commenters.)

kt,
josh

LinkWithin

Blog Widget by LinkWithin