The Vexing Concept of Kavu’a
To the Editor:
I appreciated reading Professor Moshe Koppel’s article, “A Simple Explanation of Kavu’a and Parish” (TRADITION, Winter 2018), a response to my own “A Logical Approach to Kavu’a” (in the same issue). We are both trying to present a systematic explanation for the vexing concept of kavu’a, and he demonstrates his thorough engagement with the sources. To continue this conversation, I would like to briefly engage with Prof. Koppel’s critique that my framework does not fit several cases in the Talmud. (I assume the reader’s familiarity with both my original article and his response.)
1. Koppel begins with the case in Ketubot (14b-15a). As he describes it, the Talmud is discussing a fellow who throws a stone “over a wall,” and, as he points out, it is unreasonable to assume that the thrower would know what is “on the other side of the wall.” Here, Koppel is simply misquoting the text. The Talmud writes “zorek even l’gav,” which Jastrow (s.v. “gav”), following Rashi (Ketubot 15a s.v. Amri), properly translates as “one who throws a stone into (a crowd).” There is no “wall.” Without the “wall,” Koppel’s criticism falls away. A fellow who throws a stone directly into a group of people might well be targeting a specific person in the group.
2. There is an apparent tension between my framework for kavu’a and the Talmud’s assumption (Zevahim 73b) that a mixture of permitted and prohibited cows is kavu’a. Where is the potential bias? My article proposed two answers. The first is that the person saw the prohibited animal enter the group, but he can no longer recall which animal it was. Koppel does not engage with this answer. The second answer is that the mixture is not really a case of kavu’a. In support, I cite Tosfot that say it is only a rabbinic prohibition based on the mixture being akin to cases of kavu’a. Koppel argues that Tosfot’s motivations for saying the prohibition is only rabbinic are that the cow should be batel, not based on any difficulty in kavu’a. Even were that true, the Talmudic passage does not contradict my approach because, as Tosfot show, it can be read as not being a case of kavu’a. Moreover, Tosfot was bothered by both bitul and kavu’a: After discussing bitul, Tosfot in Zevahim (73b s.v. Ela) conclude that this “is not kavu’a because the prohibited item is not known in its place.” Finally, Koppel asks why the rabbis would invoke kavu’a where bias is not possible. The Shita me-Kubetzet (ad loc. aleph) answers: “Because of the importance of live animals, we are strict and treat the cows as if the prohibited one were known.” That works cleanly with my approach: The Talmud treats animals as if they were potentially identifiable and thus subject to bias because animals are important. After all, as a general rule, people are far more likely to be able to identify important things than unimportant ones.
3. The Talmud rules that cases of parish follow the majority, but cases of kavu’a are treated as “mehtza al mehtza”—half and half (Yoma 84b; Ketubot 15a; Kidushin 73a; Zevahim 73b). Koppel’s response gives a clear illustration of the principle:
[I]f a piece of meat is found on the street (“parish”), but can be presumed to have originated in one of the ten shops, we regard it as having originated in one of the kosher shops, since they constitute a majority (“kol de-parish me-ruba parish”). But, if someone purchases meat in one of the shops (“kavu’a”), but does not currently know in which shop he had purchased it (either he forgot or never knew), the case is “like half and half” (“ke-mehtsa al mehtsa dami”) (55).
Koppel argues that that the “half and half” of Kavua is treated differently than “a safek ha-shakul (what we today would call a probability of 50%).” I disagree, but, at the very least, this is debatable. At the outset, Koppel assumes that “half and half” does not literally mean a “probability of 50%” but, rather, means a 1:1 mixture. I am not sure that is true, and numerous Ahronim, such as Hatam Sofer (Ketubot 27a) and Hiddushei Rav Shimon Shkop (Ketubot siman 23), explicitly equate the “half and half” treatment of kavu’a with safek ha-shakul. Even accepting Koppel’s assumption, the probability of selecting either element in a 1:1 mixture is 50%, so it is unclear that there should be any difference between an item that comes from a 1:1 mixture and an item with any other uncertainty of 50%. (Rashba, Torat HaBayit, 4:24b-25b, cites a debate on this last point.)
In support of his proposition that kavu’a and safek ha-shakul are treated differently, Koppel asserts that the Talmud (Pesahim 9b) treats the kavu’a case of nine bundles of matza and one of hametz differently than it treats a safek ha-shakul case where a mouse takes definite hametz into one of two houses and we are uncertain as to which. He claims this “flatly contradict[s]” my approach. Rishonim resolve the potential tension between Pesahim’s two rulings with far less dramatic solutions than distinguishing the “half and half” of kavu’a from the 50% of safek ha-shakul. For example, Ba’al ha-Maor (Rif Pesahim 4b) denies Koppel’s premise that the two cases in Pesahim are treated differently: As the Ba’al ha-Maor reads it, in both cases the Talmud requires the relatively simple ritual relinquishment of hametz (bitul) but not the more onerous search for hametz (bedika). Ra’avad (Rif Pesahim 4b) disagrees but explains why the rabbinic requirement to do bedika, over and above bitul, had to apply to cases where potential hametz has definitely been brought into the house (the kavu’a case) but did not have to apply in situations where it is uncertain that anything problematic even entered the house (the two-house-one-mouse case). At the very least, the passages in Pesahim do not “flatly contradict” my approach: as these Rishonim show, there are simple ways to reconcile the cases.