by Rabbi Aryeh Klapper
Rashi to Genesis 35:13: “In the place where He had spoken with him” – I do not know what this teaches us.
“I don’t know what this teaches us”– why not simply be silent? I suggest Rashi is taking a stand for his methodology. One might think this unanswerable exception disproves the rule that every word in Chumash teaches something, undermining a fundamental basis of Rashi’s comments about everything else in Torah. No, Rashi says; I am sure this phrase and every phrase teaches something, even if I can’t figure it out what it is. Perhaps you will figure it out.
Lehavdil, I had a similar experience this week. I had the privilege of discussing how to teach Talmud with wonderful educators at two NY day schools. One sterling young mechanekh and I later glanced together at a sugya he was teaching, and I tried using it to instantiate one of the principles I evangelize for: that students cannot understand a Talmudic passage fully unless they precisely and rigorously understand the logical forms represented by the technical terms in the passage. A few minutes later, I blithely repeated the example to another thoughtful teacher. He pointed out that I had been thinking mechanically; in this case it was not clear that following the form increased rather than decreased understanding, and in my haste to make a point I hadn’t taken the time to think through the specifics of the text. This was great mussar to me, and a challenge as well. Is this really an exception? If yes, does my principle survive? Perhaps the general principle is correct, but I simply misunderstood the particular form.
I decided to honor these beautiful conversations, and try to follow in Rashi’s spirit, by committing to publishing about the specific case without knowing what conclusion I would reach. This happily generated another spirited and thoughtful conversation with Deborah Klapper, who insisted that I try to model a research path that high school teachers could reasonably use to test hypotheses similar to mine, and that high school students could be taught to use independently. Here is the first part of the suyga, as it appears in the Vilna shas on Kiddushin 30a:
How far must a person go in teaching his son Torah?
Said Rav Yehudah said Shmuel:
כגון (=As in the case of) Zevulun son of Dan, who was taught by his father’s father mikra, Mishnah, and Talmud, halakhot and aggadot.
מיתיבי (=An attack question based on a text seen as more authoritative):
If he taught him mikra – he does not teach him Mishnah.
and Rava said: Mikra – this refers to Torah.
Like Zevulun son of Dan, and not like Zevulun son of Dan.
Like Zevulun son of Dan – in that he was taught by his father’s father.
Not like Zevulun ben Dan –
There it was mikra, mishnah and Talmud, halakot and aggadot
whereas here it is mikra alone.
The fundamental structure here seems clear.
1-3: Rav Yehudah, citing Shmuel, uses the case of Zevulun ben Dan to instantiate a principle that answers the opening question. The problem is that Shmuel’s case has at least two possibly significant particulars: the grandfather as teacher, and the comprehensive curriculum. The Talmud initially understands Shmuel’s case as instantiating the principle that a father must teach his son all the things that Zevulun ben Dan was taught by his grandfather.
4-5: The Talmud attacks Shmuel by claiming that he is contradicted by a beraita (a Tannaitic text not found in the Mishnah. Tannaitic texts are generally treated as more authoritative than memrot of Amoraim, such as Shmuel’s statement here).
7-11: The Talmud responds that Shmuel and the beraita agree that Zevulun ben Dan’s grandfather taught him far more than he was required to. Shmuel was using Zevulun ben Dan only to instantiate the principle that grandfathers, and not just fathers, are obligated to teach children.
You perhaps noticed that this outline completely ignores line 6, Rava’s statement. Why does that matter? I was confident that the vav/and of “and Rava said” is formally a subordinating conjunction, by which I mean that it makes Rava’s statement part of the argument from the beraita. If this is correct, we should expect the attack on Shmuel to be valid if and only if we understand the beraita in the way that Rava understood it. But this seems not to be the case. The beraita clearly says, before any interpretation from Rava, that a father need not teach his son both mikra and mishnah, whereas we initially understood Shmuel to require both (plus Talmud, halakhot, and aggadot). Rava’s comments therefore seem irrelevant to the argument based on the beraita. Does this mean I misinterpreted the form, or that forms are less crucial than I had argued?
One way to test a claim that Talmudic literary form A = Talmudic logical form 1 is to look up a number of parallel cases. So I opened the Bar Ilan Responsa Project and asked it to search for the words מיתיבי and ואמר, in that order, and with no more than a 25 word gap between them. This yielded a total of other 15 cases, of which 11 were irrelevant (for example the ואמר was said by a character in a beraita rather than an Amoraic legal authority). Here’s what I found in the 4 parallel cases:
Eiruvin 29a: Rav Nachman states one can make an eruv techumin with a kav of tapuchim. מיתיבי introduces a beraita that states that for the purpose of distributing the poor tithe, 5 afarsakim is considered “giving”, and Gorski bar Dari in the name of Rav Manashe bar Shkovli in the name of Rav said: The same is true regarding eruv. This attacks Rav Nachman, as our initial assumption is that tapuchim and afarsakim are alike for the purposes of eruv, and that it takes more than 5 tapuchim to make a kav. In this case, the attack question works only if one accepts the statement introduced by and X said; otherwise we would be comparing eiruvin and maaser ani with no basis, which would be like comparing apples and apricots. Score one for my hypothesis.
Bava Kamma 16a: The Mishnah has a list of animals including the bardelas. Rav Yehudah identifies the bardelas as the nafreza, and Rav Yosef (or the editor) identifies the nafreza with the afa. The מיתיבי introduces a beraita in which R. Meir adds the tzavua to the Mishnah’s list, and Rav Yosef said: The tzavua is the afa! This attacks our previous identification of the bardelas as the afa; in that case R. Meir would merely be repeating an item already on the list. Here, the attack question works only if we accept Rav Yosef’s statement that tzavua = afa. Score two for my hypothesis.
Meilah 16b: The Talmud reports that Rabbi Yose bar Rabbi Chaninah was praised by Rav Yochanan for reciting a beraita that declares that for both tum’ah and eating, less than an olivesize of sheratzim is sufficient. The מיתיבי introduces a beraita which declares that for the purposes of tum’ah less than an olivesize is sufficient, and Rav Yochanan said: One only receives lashes for (eating) an olivesize. This attacks the earlier report that Rav Yochanan praised the beraita which did not require an olivesize. Score three for my hypothesis.
Pesachim 54a: Someone reports that Rav Yochanan agreed with a statement that one makes the berakhah over flame after Shabbat and after Yom Kippur. The מיתיבי introduces a beraita that declares that one makes this berakhah only after Shabbat, with R. Yehudah commenting that one makes it together with the berakahah over wine rather than on the first flame one sees, and Rav Yochanan said: The Halakhah follows Rabbi Yehudah. This attacks the earlier report about Rav Yochanan’s position. Score four for my hypothesis. Four out of five isn’t bad, but it certainly isn’t absolute proof, and of course one might suggest that my interpretations of the four cases suffer from confirmation bias (albeit a bias that seems to be shared by many rishonim.)
A second test was to check whether my hypothesis was shared by great classical commentators. A quick check of Bar Ilan’s mefarshim-acharonim tab showed me that the Pnei Yehoshua and Hamakneh along with many, many other acharonim make yeoman efforts to explain how Rava’s comment in our sugya is a necessary component of the מיתיבי attack. However, I admit that I do not find any of their answers satisfying. Therefore, at least for now, I thank my colleagues very much for their stimulating conversation, and can only say, as per Rashi on Chumash, that I don’t know what Rava teaches us here, but I remain confident in my methodological hypothesis.
I am very open to discussion as to how high school students would react to this admission from a teacher, or to reaching this point themselves.