Does One Witness Preclude the Obligation of an Oath
Our topic is completely engendered from a side point raised on daf 2b. In the case where two "customers" are holding the merchandise, each claiming to have bought it, the gemara asks: "Let us see who paid the seller!" Rashi explains that special trustworthiness (henceforth NE'EMANUT) is given to the storekeeper, before or after he has sold the merchandise and he is, thus, believed when he identifies the customer who in fact bought from him. Consequently, we should award the merchandise based on the storekeeper's testimony as opposed to dividing it equally between the two litigants. The gemara answers that BOTH had paid and the safek is to whom the storekeeper agreed to sell. Rashi explains that the seller's ne'emanut is limited to the case where only ONE customer pays.
Tosafot disagree with Rashi (for a fuller discussion of this argument see Kiddushin 73b Rashi and Tosafot op. cit.) and maintain that the seller's ne'emanut is restricted to the case where the merchandise is still in his possession. Since in our case, the "buyers" are holding the merchandise, Tosafot interpret that the seller should be considered as a regular witness who should be able to testify on behalf of one of the litigants who would then be exempt from the oath of the mishna (the second litigant would have to swear a biblical oath to contradict the witness i.e., the seller). The gemara answers that we are referring to a case where the seller has become confused and does not remember to whom he sold it. Had he remembered correctly, his evidence would have the power to exempt a litigant from the oath of the mishna. [Ri in Tosafot adds another interpretation which is not germane to our topic.]
The Rosh infers from the statement of Tosafot (Rabbeinu Tam) that in every case, one witness may exempt the litigant from an oath. The Pilpula Charifta correctly notes that Rabbeinu Tam did not say so specifically. Indeed, a careful reading of Rabbeinu Tam in Sefer Ha-Yashar #571 shows that Rabbeinu Tam refers only to the oath of our mishna which is not to be imposed when there is one witness; however, the Rosh infers that this exemption would apply similarly to any oath.
Inasmuch as Rashi here does not raise or discuss this issue, one may conjecture if Rashi would agree with Rabbeinu Tam. The Siftei Chakhamim (Shemot 22:12) maintains that Rashi disagrees with Rabbeinu Tam (at least as explained by the Rosh): The Torah says that if a shomer (who normally would take a biblical oath) has "witness" that the object was torn apart, he is not held responsible. Rashi explains that the Torah relates to TWO witnesses. The Siftei Chakhamim comments that this is in keeping with the opinion that one witness does not preclude a biblical oath. However, this is not evident from Rashi in Bava Metzia or Kiddushin. It could be explained that in the strange event that two people paid, we assume that the seller is simply not aware of the true events and he is not believed (see Tosafot Ri Ha-zaken and other Rishonim in Kiddushin). Nevertheless, in regular circumstances, a single witness would exempt one from taking an oath.
The Rosh quotes the Maharam mi-Rotenburg who cites proofs that one witness may exempt one from taking an oath: The Torah says that one witness is not valid in certain cases (avon and chatat) and the gemara (Shevu'ot 40a) deduces that one witness is indeed significant in cases of oaths. There is no distinction made between obligating an oath and exempting one. He also argues that if one witness may cause the one in possession (muchzak) to swear, a fortiori he should exempt the muchzak from an oath. Rav Y.Z. Gustman zt"l notes (Kuntresei Shiurim p. 15) that without these proofs, apparently everyone would argue that although one witness can create an obligation of an oath, he could not exempt one from any oath. One witness is not "believed" in the case of an oath. We merely create the obligation. [Note the difference - in certain laws (such as kashrut), the law states that one witness is "ne'eman" whereas it does not say that one witness is "ne'eman" in cases of oath; it merely says that one witness creates the obligation of an oath.] Therefore, it is understandable that perhaps Rashi and the Ri (who gives an alternate explanation to Rabbeinu Tam) maintain that one witness does not preclude any oath. In fact, the Beit Yosef (C.M. 46:36) explains that the Rama holds that one witness does not preclude an oath imposed by one witness.
While we have seen that Rabbeinu Tam says that one witness does preclude an oath, while the Rama and others disagree and say that the oath is still in place, there are a number of opinions which distinguish between different types of oaths. The Ramban (Milchamot p. 1 of Rif) maintains that Rabbeinu Tam merely said that in our mishna, where the oath is enacted by Chakhamim to deter people from grabbing someone's tallit, one witness' testimony would be sufficient to preclude the oath. However, in the case of a biblical oath, he would agree with the Rama and require a shevu'a.
The Nimukei Yosef (op. cit.) cites an opinion that argues with the principle that one witness would preclude the oath of the mishna but not a biblical oath. Nonetheless, in the case where one witness testified and engendered the obligation, a contradictory witness would cancel the witness' testimony and would preclude the oath. (There could be differences made as to the exact situation when the second witness came, see text of the Nimukei Yosef.)
The Sefer Ha-terumot (sha'ar 21 5:2) maintains that although a witness does preclude all oaths (like Rabbeinu Tam according to Rosh), a "heset" oath must still be taken. This oath is mandated by the Chakhamim even in the case of "kofer ha-kol" (he who denies any obligation and there is no evidence to contradict his claim). In the case of one witness who backs up the litigant's claim there is still no clear and indisputable evidence and, thus, a shevu'at heset should be administered.
The Gedulei Teruma (op. cit.) cites the Raza, presumably the Ba'al Ha-ma'or (however, note that the Ramban in Milchamot op. cit. interprets him differently) as well as the Rashba (responsum #1161) that one witness would not preclude any biblical oath; however, he would preclude any oath of rabbinic origin (presumable even "heset").
The last opinion is found in Teshuvot Maimoniyot to Sefer Mishpatim #61 (cited in Rama C.M. 87:6). He maintains that one witness would preclude an oath even in the case where one must swear in order to receive money (nishba ve-notel). The Bach (responsum 64) rejects this opinion and maintains that Chakhamim would not allow a person to collect money on the basis of one witness. The Pitchei Teshuva cites responsum #6 of Chut Ha-shani which differentiates between different cases of nishba ve-notel. In the case where the oath is only a "chumra" and the litigant should collect the money without an oath and Chakhamim demanded the oath, one witness would preclude the oath. However, in the case where the litigant should not collect the money and Chakhamim allowed him to take an oath and collect, one witness would not be sufficient to preclude the oath.
We have thus seen two extreme opinions with 4 variations.
1) Rabbeinu Tam (as interpreted by Rosh) - all oaths are precluded by one witness.
2) Rama (perhaps Rashi and Ri) - no oath is precluded by one witness.
3) Ramban - only the oath of our mishna is precluded.
4) Nimukei Yosef - an oath engendered by one witness as well as our mishna's oath is precluded.
5) Sefer Ha-terumot - every oath except heset is precluded.
6) Shut Maimoniyot - even nishba ve-notel may be precluded. Some disagree with this opinion and others qualify it.
"Migu" - To Exempt an Oath
The gemara (Bava Metzia 3a) cites Rava's question: Why is there a biblical oath in the case of "modeh miktzat" (partial admission)? Tosafot interpret the question as follows - why don't we believe him with a "migu" (alternate claim) that he could have denied everything? The gemara answers that a person would not dare ("me'iz") to deny everything and, thus, no palatable alternate claim exists. Apparently, Tosafot think that a regular migu would preclude an oath; however, a migu de-he'aza (of daring) would not preclude an oath.
The Rosh (Shevu'ot 7:3) cites the Ri Migash that no migu would exempt an oath. He argues that in the case of shomrim or modeh be-miktzat there always is a migu and yet there is an oath. (He must explain our gemara differently than Tosafot - see Rashi ad loc.) The Rosh disputes this point and says that migu in these cases is always a migu de-he'aza. Apparently, the Ri Migash thinks that since you do say a migu de-he'aza in monetary cases (see Bava Batra 36a) there is actually no distinction between these types of migu. Therefore, no migu may exempt an oath.
The explanation of this controversy is generally explained as follows (see Pri Moshe Shevu'ot Mammon and numerous Acharonim): Is the obligation of the oath basically connected to a monetary obligation; however, one may swear to avoid payment or is the obligation basically a requirement to buttress your claim through an oath. If the obligation is connected intrinsically to a financial obligation, then the argument of the Rosh is well taken. Inasmuch as a migu exempts one from payment, it should also exempt one from an oath. [Of course, this raises the obvious question which will be discussed later if a migu de-he'aza exempts one from payment, why does the Rosh agree that it does not exempt one from taking a shevu'a?] The Ri Migash, on the other hand, apparently thinks that the obligation of an oath was placed upon the litigant in order to support his claim. A migu may also support his claim but it is not as strong a proof as an oath. Therefore, a migu cannot exempt an oath.
[Note: A parallel approach can help explain why one witness would or would not exempt the obligation of a shevu'a (see first part of this shiur). If the obligation of a shevu'a exhausts itself in the requirement to buttress a claim, then the issue of one witness can be reduced to the extent his evidence is convincing. If, however, the obligation of a shevu'a is rooted in a monetary obligation, then perhaps the evidence of one witness is irrelevant. However, the kal va-chomer of the Maharam mi-Rotenburg may nevertheless be applicable.]
In order to explain the question that was raised according to the Rosh, namely why does a migu exempt an oath but a migu de-he'aza which is effective in financial matters does not exempt an oath, we will refer to classic understanding of migu. Rav Elchanan Wasserman (Kovetz Shiurim Vol. 2 siman 3) (as well as numerous Acharonim), explains that a normal migu may work in one of two ways:
1. It indicates that the person is telling the truth as otherwise he would have preferred the alternate claim.
2. It could be based on the position of power that the alternate claim could have awarded him.
In the case of a migu de-he'aza there is no clear indication that his claim is true, since he would prefer to make the less daring claim. However, his position of power, based on the alternative option remains intact. Rav Elchanan explains that this position is sufficient to win a financial argument. However, in order to preclude an oath, it is insufficient. Regarding an oath, we must be convinced that the litigant is indeed telling the truth and this is simply not so in the case of migu de-he'aza. Therefore, although a migu de-he'aza exempts monetary obligation, it can not preclude an oath. This, of course, assumes that even the Rosh agrees that the obligation of a shevu'a demands support for the claim, and is not related to a monetary obligation. Therefore, a migu de-he'aza may erase a monetary obligation simply by the power inherent in the position established by the migu. However, only a migu which convinces us of the truthfulness of the claim, can effectively exempt the litigant from the obligation of a shevu'a.
To summarize: Ri Migash thinks that no migu may exempt an oath as the obligation to swear is basically related to his veracity. However, one who considers the obligation of an oath as analogous to a monetary obligation, migu would be as effective as in financial matters. We used the classic explanation of "migu" to clarify the position of the Rosh that although a regular migu may exempt an oath and a migu de-he'aza is effective in financial matters, nevertheless a migu de-he'aza can not preclude an oath.
Sources for the next shiur:
1. Bava Metzia 2b "leima matnitin de-lo ke-ben Nanas … agbihuha," 3a "bein le-rabbanan … mi-ba'al ha-bayit."
2. Shevu'ot 45a "ve-hachenvani al pinkaso keitzad … shelo be-shevu'a."
3. Ritva Shevu'ot 47a s.v. amar ben Nanas, Bava Metzia 2b s.v. de-i ke-ben Nanas.
4. Ba'al Hamaor Shevu'ot [33b be-dapei ha-Rif] "ve-hareini moser lekha … litol belo shevu'a," Ramban ibid. till "shechenvani chashud eino notal."
5. Rambam Hilkhot Milave ve-loveh 16:5.
6. Rambam Hilkhot Shevu'ot 1:4-7.
7. Sefer Ha-mitzvot Aseh 7 and in Ramban's commentary.
1. What is the distinction between a shevu'at shav and shevu'at sheker?
2. In the case of "chenvani al pinkaso", if both the storeowner and workers were to swear, would the problem be one of shevu'at sheker or shevu'at shav?
3. Why does the Ramban reject the position taken by the Ba'al Hamaor?
4. How can the opinion of the Ba'al Hamaor be defended?