Thursday, December 08, 2005

parshat Vayeitzei: Who Named Levi?

Who named Levi? The pasuk states {Bereishit 29:34}
לד וַתַּהַר עוֹד, וַתֵּלֶד בֵּן, וַתֹּאמֶר עַתָּה הַפַּעַם יִלָּוֶה אִישִׁי אֵלַי, כִּי-יָלַדְתִּי לוֹ שְׁלֹשָׁה בָנִים; עַל-כֵּן קָרָא-שְׁמוֹ, לֵוִי. 34 And she conceived again, and bore a son; and said: 'Now this time will my husband be joined unto me, because I have borne him three sons.' Therefore was his name called Levi.
JPS thus translates קָרָא as "was called," the passive.

There is another possibility, given by Judaica Press, based on Rashi, in turn based on a midrash in Devarim Rabba. They translate:
And she conceived again and bore a son, and she said, "Now this time my husband will be attached to me, for I have borne him three sons; therefore, He named him Levi.
Thus, קָרָא is translated as "he called," with God as the one performing the act of naming.
He named him Levi: [I wondered why] it is written concerning this one, “and he named,” while concerning all [the others], it is written, “and she named.” There is an aggadic midrash in Deuteronomy Rabbah, [which relates] how the Holy One, blessed be He, sent Gabriel and he brought him (Levi) before Him, and He gave him this name, and He gave him the twenty-four priestly gifts; and because he accompanied him (לִוָּהוּ) with gifts, he named him Levi.
The simple, non-midrashic reading of this, taking the verb the same way syntactically, is "he said," with a lowercase "h," with Yaakov doing the naming. And on a peshat level, it is not clear that Rashi disagrees. He notes the distinction of "she called/he called," and then mentions a midrash which accounts for the sudden shift, while marking it clearly as a midrash aggada. (Or alternatively, he notes the distinction "she called his name"/"his name was called," which carries no real semantic difference, and notes a midrash aggada which plays upon this.)

The trup would seem to favor taking קָרָא as "was called," the passive, for there is a makef (dash) connecting the words קָרָא-שְׁמוֹ, such that they receive only one stress. This is more consistent with the name being the subject ("thefore his name was called") than the object. I do not think we would expect a makef if "his name" were the object. Indeed, in the close parallel, in which Rachel names a son, we read {Bereishit 30:6}:

ה וַתַּהַר בִּלְהָה, וַתֵּלֶד לְיַעֲקֹב בֵּן. 5 And Bilhah conceived, and bore Jacob a son.
ו וַתֹּאמֶר רָחֵל, דָּנַנִּי אֱלֹקִים, וְגַם שָׁמַע בְּקֹלִי, וַיִּתֶּן-לִי בֵּן; עַל-כֵּן קָרְאָה שְׁמוֹ, דָּן. 6 And Rachel said: 'God hath judged me, and hath also heard my voice, and hath given me a son.' Therefore called she his name Dan.
we have no makef between קָרְאָה and שְׁמו. ֹ

It is a little strange for Yaakov to be naming the son, for he is not mentioned in the verse. (For that matter, neither is God.) It is Leah who gives the etymology for the name, and so we should either expect her to actually assign the name, or for Levi to be named, using a passive verb, in which the actor is not specified. True, Leah could have said this and in response Yaakov named him Levi, but that is a bit more difficult.

And so, we have three different possibilities:
1) He was named (presumably by Leah)
2) God named him
3) Yaakov named him

There is actually a fourth possibility I will cover shortly.

Speiser, in Anchor Bible Genesis, emends the text to "she called him," noting "Sam., LXX; MT "was called/he called." That is, the Samaritan Torah has "she called him," as does the Septuagint, and he follows their more logical reading. The Masoretic text, on the other hand, hasקָרָא, which can mean either (1) or (3).

Here is the LXX:

29:34 και συνελαβεν ετι και ετεκεν υιον και ειπεν εν τω νυν καιρω προς εμου εσται ο ανηρ μου ετεκον γαρ αυτω τρεις υιους δια τουτο εκαλεσεν το ονομα αυτου λευι

I could not find the Samaritan Torah online, (though would appreciate someone sending me a link), but I assume that they have the text קָרְאָה rather than קָרָא.

{Update: Indeed it does. Thanks for the link! Here is the page in question.}

I take exception to Speiser's comment, because it implies that MT could not mean "she called." In fact, it can (as I will soon explain). By making this an emendation and claiming that the MT means something else, we get to apply emendation, which looks nice and scholarly, but several other things occur:

1) He partially undermines his own argument. If all texts are in agreement, then the conclusion is more solid.
2) Those who give primacy to the MT will take the text otherwise. This will lead them to argue against his emendation, and also to think that the correct meaning of the must be "was called" or "he called."
3) The MT gets maligned - a data point is created in which the Masoretic text was wrong.
4) In fact, the thought process which involves justifying the MT could explain the relationship of the Samaritan text and the LXX to the MT, which then yields a deeper understanding of the issue.
5) Parenthetically, Speiser is possibly wrong, and the correct reading may indeed indeed be "was called." A closer analysis might have revealed this.

Indeed, the most likely meaning of the term appears to be "she called." Consider all the other sons (and one daughter):
  1. Reuven: (29:32): וַתִּקְרָא שְׁמוֹ רְאוּבֵן
  2. Shimon: (29:33): וַתִּקְרָא שְׁמוֹ, שִׁמְעוֹן
  3. Levi: (29:34): עַל-כֵּן קָרָא-שְׁמוֹ, לֵוִי
  4. Yehuda: (29:35): עַל-כֵּן קָרְאָה שְׁמוֹ, יְהוּדָה
  5. Dan (30:6): עַל-כֵּן קָרְאָה שְׁמוֹ, דָּן
  6. Naftali: (30:8): וַתִּקְרָא שְׁמוֹ, נַפְתָּלִי
  7. Gad: (30:11): וַתִּקְרָא אֶת-שְׁמוֹ, גָּד
  8. Asher: (30: 13): וַתִּקְרָא אֶת-שְׁמוֹ, אָשֵׁר
  9. Yissachar: (30:18): וַתִּקְרָא שְׁמוֹ, יִשָּׂשכָר
  10. Zevulun (30:20): וַתִּקְרָא אֶת-שְׁמוֹ, זְבֻלוּן
  11. Dinah (30:21): וַתִּקְרָא אֶת-שְׁמָהּ, דִּינָה
  12. Yosef (30: 24): וַתִּקְרָא אֶת-שְׁמוֹ יוֹסֵף
  13. Ben-Oni/Binyamin (35:18): וַתִּקְרָא שְׁמוֹ, בֶּן-אוֹנִי; וְאָבִיו, קָרָא-לוֹ בִנְיָמִין
Each of these is called by his or her mother. There are one or two exceptions. Binyamin is called Ben Oni by his mother, while his father is the one who names him Binyamin. And the other possible exception is Levi, if we read קָרָא as "he called him." Perhaps we can match קָרָא by Binyamin to קָרָא by Levi.

However, Binyamin is in a separate section, and is an exceptional case - and even there, we have וַתִּקְרָא. Let us consider the other cases.

In each, the mother names the child. In most instances, this is stated via: וַתִּקְרָא אֶת-שְׁמוֹ. However, there are three in the middle which deviate. Speiser engages in a smoothing operation. First, smooth all three exceptions to be identical. This could be accomplished by
transforming the three exceptions to קָרָא, or else by transforming the three exceptions to קָרְאָה. The second step is to smooth these three exceptions with the surrounding verses. In each of those other verses, the mother names the child (וַתִּקְרָא), and thus it is likely that these all are really קָרְאָה. The reason for a simple perfect (=past) tense rather than the vav conversive form וַתִּקְרָא is the word עַל-כֵּן, which interrupts the flow such that we cannot have וַ here. Another reason to favor the transformation to קָרְאָה - two of the verses have קָרְאָה and only one has קָרָא, and we should follow majority rule, for this entails fewer emendation and fewer claims of the text having become mangled.

I do not know if all this went through his mind. A more simple way at arriving at this conclusion: there are 12 people named here, and with only one exception, the mother does the naming. In the single exception, there are two variant textual traditions, represented by Sam. and LXX, which have קָרְאָה. The most logical path would be to follow the two texts and assume that the one (the MT) is wrong - and to follow the two instances of קָרְאָה in the MT, and assume that the one occurrence of קָרָא is wrong. Thus, we should emend to קָרְאָה.

However, we need not emend קָרָא to make it mean "she called." Let us examine this phonologically. Neither the aleph of קָרָא nor the heh of קָרְאָה is pronounced. Both are matres lectiones, to help with the reading of the word (or in the case of קָרָא, for etymological reasons as well).

Perhaps the word was originally, correctly קָרְאָה. What then happened is commonly called "elision of aleph." That is, the sheva followed by aleph are not pronounced. We see a very common occurence of this phenomenon in the word וַיֹּאמֶר, where the aleph, while present, is not pronounced. On the occasion where aleph is pronounced, it will be at the beginning of a syllable or at the end followed by a shva nach (resting shwa). In וַיֹּאמֶר, there is no shva.

Another example which people frequently get wrong is in Tehillim 34:10, though it also accurs in bentching:
י יְראוּ אֶת-ה קְדֹשָׁיו: כִּי-אֵין מַחְסוֹר, לִירֵאָיו. 10 O fear the LORD, ye His holy ones; for there is no want to them that fear Him.

There is no shva under the resh or aleph, telling us that only one is pronounced. In this case, it is the aleph. Note also the shva under the yud instead of the chirik we would expect. The word is pronounced ye-ru, rather than yir-`u {update: or rather, yi-re-`u}. In other words, יְרוּ.

That the word is pronounced without the aleph - יְרוּ - does not change its syntactic or semantic meaning. This is a purely phonological phenomenon.

Similarly, if we have the same elision of aleph for the word קָרְאָה, we would pronounce it קָרָה, though we might as well keep the aleph and have קָרָא so that people can identify the root. After all, neither the aleph nor heh are actually pronounced.

And once again, that the word is pronounced differently is a phonological phenomenon, but has no bearing on syntax nor semantics. The word would still mean "she called." And Speiser would not have to appeal to an emendation.

Alternatively, the text can be written in one way (ketiv) and pronounced otherwise (krei). There are some fairly odd spellings. Perhaps the text was written קרא but should be vocalized קָרְאָ. Not always do we have matres lectiones at the end of words. There are times that they are regularly absent - e.g., the kha ending in לְשׁוֹנְךָ and the ta ending in אָכָלְתָּ. Sometimes u endings of words disappear and we call it ketiv and krei. Consider also the case of motzi shem ra, in Devarim 22, with the multiple occurrences of הַנַּעֲרָ. The MT does not mean "boy." Rather, this is a variant spelling.

So too, קָרְאָ might be a variant spelling. Or, as mentioned above, this could be a phonological phenomenon of elision of aleph such that it was even meant to be pronounced קָרָא.

What now of the other two textual traditions? The LXX is a translation, and just reflects meaning. If the word means "she called" even in MT, then it could just as well reflect a translation of the MT. In terms of the Samaritan Torah, if this is a variant spelling of קָרְאָ in MT, the Samaritan text could simply be smoothing this out by adding the heh at the end. Or perhaps it is also a translation. As has been demonstrated (link goes to essay by R' Gil Student), the Samaritan text itself emends the text to smooth things out. And as Rabbi Chajm Heller writes (same link), they have a tendency to modift the text as a means of interpretation. Furthermore (same essay):
Additionally, we find slight changes of wording in the Samaritan Torah to fit the text to the Samaritan dialect of Hebrew. For example, the Samaritan scribes consistently removed verbs in the infinitive. Also, certain Aramaic words found their way into the text. Both of these types of changes reflect the translation of the biblical text into the Samaritan dialect, a phenomenon very consistent with the previously mentioned commentaries that were also inserted into the text.
So this is really a translation in Samaritan dialect, a Hebrew Targum if you will, such that if it reflects the meaning, this does not necessarily reflect an original wording.

Thus, Speiser can be correct in having the word mean "she called" without needing to emend the text.

However, if I were Speiser, I would have emended the text in the opposite direction. The problem is not the one occurrence of קָרָא, but the two occurrences of the word קָרְאָה. I would have changed both to קָרָא, to mean "was called."

Speiser likes to divide the text into segments - P, J, E, what have you. Yet his eye is on Elokim/ YKVK distinctions. There is a much more obvious distinction, however - between the עַל-כֵּן and the וַתִּקְרָא. Indeed, Speiser assigns the first two עַל-כֵּן's (that is Levi and Yehuda) to J, because the name YKVK is used with Yehuda, and the third עַל-כֵּן, the naming of Dan, to E, because Elokim is used.

Nonsense. If any grouping is to be made, it should be of this middle group of עַל-כֵּן stuck in the center of a ground of וַתִּקְרָא. To be fair, Speiser notes that the division in this section of Yaakov's children is tenuous - "The borderline between J and E is sometimes uncertain in this section." But this is not an issue of borderline. This middle section, which stands out, contains both J and E, and comes from the same source, if we are even to posit different sources. (In fact, there are many other issues with an J/E analysis here, which I will not go into here because it will take us too far afield.) (Note also that there are other elements in the narrative that link other groups of sons together.)

If these three are distinct from וַתִּקְרָא, there is no need to smooth to match the context. Those other verses can have her (with who she is changing from verse to verse) naming the children, while these three need not have her explicitly naming the children.

Furthermore, from a syntactic perspective, the words עַל-כֵּן, "therefore," used to introduce the section and take it out of the vav conversive, sets the stage for a distancing of actor. "was named" is dependent on the existence of "therefore," and thus we would expect "was named" to follow "therefore," or else there is no need for the shift.

Finally, forget going after the majority when we move from קָרָא to קָרְאָה. For קָרְאָה is most decidedly not the majority. Sure, it is in the majority locally: two to one. However, globally, when we search (here at snunit) for the words עַל-כֵּן קָרָא, you will get 13 matches, where it is clear that what is meant is "was called." In constrast, if you do a search for עַל-כֵּן קָרְאָה, you will find two matches - namely, the two in parshat Vayeitzei. And these might have been changed by a scribe to match all of the other instances of וַתִּקְרָא.

Thus, the reading קָרְאָה is outnumbered 13 to 2. This makes sense if the words עַל-כֵּן are to introduce the passive. Perhaps Speiser should emend the two instances of קָרְאָה to קָרָא.

If we do not wish to emend, perhaps we can explain that in truth the word קָרָא was intended. At the same time, the Author wished to match the other instances of וַתִּקְרָא, and so in general wrote קָרְאָה, which can perhaps be pronounced the same (or at least similarly). After all, in all these instances, it was the mother than named the child. To convey both, a multivalent reading was created, the Author maintaining the spelling in the first to make sure the reader saw the base meaning in all three cases.


Anonymous said...


Here you can find Samaritan Pentateuch on-line (scanned):

Just enter "any" for both, when it asks username and password.


סדר זרעים, מסכת ברכות

פרק ו,כג: ר' יהודה אומר שלש ברכות צריך לברך בכל יום ברוך שלא עשני גוי ברוך שלא עשני אשה [ברוך] שלא עשני בור

Best regards,

joshwaxman said...


it indeed says קראה.

I'll have to add that link to my bookmarks.

all the best,


Blog Widget by LinkWithin