Consistency in Yesh Em LaMasoret
In Kiddushin 18, we have a brayta with a dispute between third-generation Rabbi Eliezer ben Hyrcanus and fourth-generation Rabbi Akiva.
וּבִפְלוּגְתָּא דְּהָנֵי תַּנָּאֵי דְּתַנְיָא: ״בְּבִגְדוֹ בָּהּ״ –
The Gemara comments: And this issue is taught in the dispute between these tanna’im, as it is taught in a baraita with regard to the verse: “He shall have no power to sell her to a foreign people, seeing that he has dealt deceitfully with her [bevigdo vah]” (Exodus 21:8).
כֵּיוָן שֶׁפֵּירַשׂ טַלִּיתוֹ עָלֶיהָ, שׁוּב אֵין רַשַּׁאי לְמוֹכְרָהּ, דִּבְרֵי רַבִּי עֲקִיבָא. רַבִּי אֱלִיעֶזֶר אוֹמֵר: ״בְּבִגְדוֹ בָּהּ״ – כֵּיוָן שֶׁבָּגַד בָּהּ, שׁוּב אֵין רַשַּׁאי לְמוֹכְרָהּ.
This verse indicates that once the master has spread his garment over her, thereby designating her as his wife, her father may no longer sell her. This is the statement of Rabbi Akiva, who interprets bevigdo as related to beged, meaning garment. Rabbi Eliezer says: “Bevigdo vah” means that since the father dealt deceitfully [bagad] with her and sold her once, he cannot sell her again.
The Talmudic Narrator explains the basis for their dispute, namely what force we grant to the consonantal written text and to the pronounced, oral text, which contains a tradition of vowels.
בְּמַאי קָמִיפַּלְגִי? רַבִּי אֱלִיעֶזֶר סָבַר: יֵשׁ אֵם לַמָּסוֹרֶת, וְרַבִּי עֲקִיבָא סָבַר: יֵשׁ אֵם לַמִּקְרָא, וְרַבִּי שִׁמְעוֹן סָבַר: יֵשׁ אֵם לַמִּקְרָא וְלַמָּסוֹרֶת.
The Gemara asks: With regard to what principle do they disagree? Rabbi Eliezer maintains that the tradition of the manner in which the verses in the Torah are written is authoritative, and one derives halakhot based on the spelling of the words. One relies on the way a word is written, without the traditional vocalization, and therefore it is read as though it were vocalized as bevagdo, which refers to betrayal, not a garment. And Rabbi Akiva maintains that the vocalization of the Torah is authoritative, meaning that one derives halakhot based on the pronunciation of the words, although it diverges from the spelling, and since one pronounces the term as bevigdo, it is related to the word beged, meaning garment. And Rabbi Shimon maintains that both the vocalization of the Torah and the tradition of the manner in which the verses in the Torah are written are authoritative. Consequently, she cannot be sold as a slave after she has been taken as a wife, nor can she be sold again after she has already been sold once.
OK, that is the background.
Personally, I am not entirely convinced that this is the basis for the Rabbi Eliezer / Rabbi Akiva dispute. After all, on a peshat level, even with a chirik in bivigdo, it means “as he betrayed” by the rules of Hebrew grammar. A typical yesh em lamasoret that comes to mind is basukot basukot basukot, where the word is spelled deficiently, so we ignore the peshat meaning and the dikduk, ignore that there are cholam maleis making it plural, and reading the two deficient instances as if it were the singular construct basukat. Here, for yesh eim lamasoret, we need to pretend that the peshat accords with the revocalization of bevagdo bah, because there is no extraneous yud. And the yesh eim lamikra is the farfetched, non-obvious case. (Though maybe I am imposing my sense of peshat and derash onto this, confusing matters.)
If I had my druthers, I would argue with the Talmudic Narrator. I would suggest that the Sage suggesting it is a reference to a beged is engaging in derash, an additional level of interpretation on top of the peshat, because ain mikra yotzei miydei peshuto, we hardly ever abandon the basic plain meaning of the text in the face of additional midrashic interpretation. This would let us arrive in the same place, if we were using Rabbi Shimon. Alternatively, bigdo as a beged is enough peshat like, since the word appears in that form elsewhere, וַתִּתְפְּשֵׂהוּ בְּבִגְדוֹ לֵאמֹר שִׁכְבָה עִמִּי, to make it the sole meaning of the word.
Next, Rashi. I think he makes a deliberate girsological choice here, about which Sage said what. This may be as a result of sevara, logic, rather than seeing any text with this. Thus, Rashi writes:
בדברי ר"ע גרסינן כיון שפירש טליתו עליה - דעל כרחך מיבעי ליה לאוקמי יש אם למקרא כר"ע כדאמרינן בסנהדרין בפ"ק ויש אם למסורת כר"א כדדייקי' בבכורות דרמינן מהא פלוגתא על מילתא אחריתי דאמרינן התם לר"א יש אם למקרא ורמינן עלה מהא דיש אם למסורת ס"ל בפ' כל פסולי המוקדשין ובמכילתא גרסינן שפירש טליתו עליה בדברי ר"ע ודר"א ליתא התם:
He thus establishes the girsa of beged / clothing within Rabbi Akiva (like our gemara) in order to maintain consistency with another gemara, in which Rabbi Eliezer maintains yesh em lamasoret.
I can point also to Keritot 17b, a named Amora telling us what system Rabbi Eliezer follows:
אמר לו ר' אליעזר סבר יש אם למסורת מצות כתיב
Rava said to Abaye: Rabbi Eliezer holds that the tradition of the manner in which the verses in the Torah are written is authoritative. Consequently, as the singular form “mitzvat” is written, he maintains that a provisional guilt offering must be brought in a case of uncertainty involving a single item, such as the fat of a koy.
Meanwhile, look at (the two comments of) Tosafot, who cites Rabbeinu Tam, who wants to have Rabbi Eliezer as the one with the beged / clothing.
מיהו קשה לר"ת דאין דרך הגמרא להקדים דברי ר"ע לדברי ר"א שהיה רבו לכך נראה לר"ת דגרסי' כיון שפירש טליתו עליה דברי ר"א ור"ע אומר כיון שבגד בה וניחא השתא שדברי ר"א קודמין לדברי ר"ע שהיה תלמידו וה"פ ר"א סבר יש אם למסורת ומסורת הוי לשון בגד דהיינו פריסת טלית מדלא כתיב בבוגדו בוי"ו או בביגדו ביו"ד אע"ג דבשמעו לא כתב בוי"ו היינו משום דליכא למיטעי ור"ע סבר יש אם למקרא היינו לשון בגידה דאי לשון בגד הל"ל בבגדו דמבגד יאמר בגדו כמו מנגד נגדו ומחלב חלבו ואע"ג דקרינן ותתפשהו בבגדו ולא בבגדו התם ליכא למיטעי וי"מ בהאי דכתיב אשר לא יעדה וקרינן לו בוי"ו ר"ע סבר יש אם למקרא דהיינו לו בוי"ו והשתא מגלה על בבגדו שהוא לשון בגד פריסת טלית ר"א סבר יש אם למסורת וכתיב לא באל"ף והשתא מגלה על בבגדו שהוא לשון בגידה:
A rather solid point he makes is that Rabbi Eliezer ben Hyrcanus is the teacher of Rabbi Akiva, so his position should come first in the brayta. (Rabbeinu Tam also works hard to maintain old, difficult girsaot, in the face of rabbinic authorities who would emend them to simplify the read of the gemara.) Therefore, Rabbeinu Tam explains that the masoret with בגד implies a garment, and not betrayal. A betrayal should have a cholam to indicate the presence of a kamatz, bevogdo bah, or bevagdo bah. And Rabbi Akiva would read it with / as if it has a vav, so he gets his betrayal.
Now me: Even though our oral tradition, as we read it in shul, and as we have now printed nekudot showing how the Tiberian Masoretes read it, with a chirik!
Look, it is farfetched, but Rabbeinu Tam was cornered.
Now for my radical suggestion.
I would say that Rabbeinu Tam is absolutely correct in who said what in terms of clothing vs. betrayal, that Rabbi Eliezer said clothing. And, Rashi is absolutely correct in which of mikra vs. masoret maps to which, that mikra maps to clothing and masoret maps to betrayal. And this is precisely what the Talmudic Narrator intended.
But then, we have the problem that Rashi raised, an inconsistency between our gemara and others about which Tanna holds which position about lamikra or lamasoret.
I would answer that:
(1) Again, let’s argue with the Talmudic Narrator, and lamikra vs. masoret isn’t really in play here.
(2) Generally, the Talmudic Narrator is NOT consistent between sugyot about who says what. For instance, we often encounter a derasha chain — the game of musical chairs in which each interpreted Biblical phrase maps to a concrete law, and each concrete law maps to a Biblical phrase. Once we say A interprets J to derive X, and B doesn’t, we ask, what does B use J for? And, says the Talmudic Narrator, B uses J to derive C. Then, how does A derive C? Well, A uses K to derive C. Then what does B use K for? And this can go on forever, in shifting fashion. The Talmudic Narrator frequently ends it by asserting that A doesn’t interpret doubled language, because dibra Torah kilshon benei adam. Or naal vs. hanaal, B does not pay heed to.
But then, as Tosafot, who have a global view, point out, this is inconsistent. Here you said A doesn’t interpret doubled language, but there, you said he does! And they give a harmonized answer, such as A doesn’t interpret doubled language unless there is something real necessary to derive. So really it is optional.
But the real answer, I think, is that the (post-Amoraic) Talmudic Narrator isn’t consistent across masechtot, either because he’s one person and cannot always recall what he suggested everywhere, or much more likely, because there were different redactors, different Talmudic Narrators, and each focused on their own set of sugyot.