KETUBOT 19B: Witnesses Who Come To Disqualify The Witnesses Who Are Signed On A Deed
The Baraita asserts that after a document is authenticated, other witnesses are not believed to say that the witnesses signed on the document were disqualified from giving testimony. The Gemara explains that the witnesses are given limited credence, as the situation is considered one of two witnesses against two witnesses. In this shiur, we shall consider two issues discussed by the Rishonim (ad loc.): 1) When did the witnesses signed on the document disagree with the witnesses who come to disqualify them, and even if we say that they disagreed, do they have the credibility to do so? 2) Why don't we believe the second set of witnesses based on a miggo?
Our Rabbis taught: [If] two] witnesses were signed on a deed and they died, and two [witnesses] came from the marketplace and said: "We know that it is their handwriting, but they were under duress, they were minors, they were disqualified [from giving] evidence," they are believed. But if there are witnesses that it is their handwriting, or where their handwriting was authenticated from another place – from a document that had been contested and had been confirmed in court – they are not believed.
But do we collect with it as with a valid document? But why? They are two against two!… (Ketubot 19b)
I. Defining the Situation as Two against Two
Our passage establishes that in the absence of the argument of "peh she-asur," witnesses are not believed to disqualify the witnesses who are signed on a deed, and therefore we are dealing with a situation of "two against two" – conflicting testimony of two sets of witnesses. This assertion is very difficult to understand, for two different reasons.
First of all, the witnesses signed on the deed do not deny the testimony of the set of witnesses coming to disqualify them. All they did was sign the deed, without relating to their own age, qualification, or being under duress. But even if we can overcome this difficulty, and say that a deed is regarded as having pretensions of being a valid deed, and as testifying to its own fitness, another difficulty still remains. It would seem that even if the witnesses who are signed on the deed would come and declare that they had been adults, fit to testify and not under duress, they would not be believed to make such claims against the set of witnesses coming to disqualify them. Surely the Gemara explicitly states (Bava Kama 73a, Sanhedrin 27a) that witnesses can disqualify other witnesses by accusing them of thievery, and that there is nothing anomalous about the idea that we believe the second set of witnesses more than the first, because the second set do not testify about the matter at hand, but about the witnesses themselves, and the first set of witnesses are not believed to testify about themselves that they are qualified witnesses.
It would appear that the Rosh (siman 13) was so troubled by these questions that he chose to explain our passage in a most novel way:
They are two against two, because the witnesses who are signed on the deed, even though these testify against them that they were disqualified, it is as if there are two others who testify that they are qualified. For there is a presumption that the lender would not sign illegitimate witnesses on his deed, and this chazaka is regarded like witnesses… For even if [the second set of witnesses] say that [the witnesses who are signed on the deed] are disqualified from giving testimony even now, they are not believed to disqualify the deed. For by strict law, even where their handwriting comes out from another place, we should have believed the second set [of witnesses], for they do not come to say that their signatures are forgeries, but rather they impair the deed based on a different disqualification. And the reason that they are not believed is that "we are witnesses" that any deed that is written and signed was signed without any disqualification. Therefore, even if they say that they are disqualified from giving testimony now, they are not believed to disqualify the deed, since their handwriting comes out from another place.
The Rosh, explains that indeed we are not dealing here with two witnesses who testify that the deed is valid. What validates the deed, i.e., what defines the situation as one of two against two, is not two witnesses, but an external presumption that has the force of "we are witnesses" ("anan sahadi") – "For there is a presumption that the lender would not sign disqualified witnesses on his deed… for 'we are witnesses' that any deed that was written and signed was signed without any disqualification."
The Rosh, presumably, was aware of the difficulties in his explanation. It is a forced interpretation of the wording of the Gemara, which simply asks: "They are two against two!"; there is no source teaching that the chazaka of which he speaks has the force of witnesses; and to the best of my knowledge – there is no precedent for treating a clash between witnesses and anan sahad as "two against two." But despite all of these difficulties, the Rosh is not prepared to accept that possibility that we are dealing here with a real clash of two sets of witnesses. It is not surprising that the Rosh remained relatively unique in his approach.
Rashi explains: "They are two against two – the witnesses signed on the deed are two, and these two who testify about them that they were disqualified at that time." He explicitly states that the witnesses testifying that the deed is valid are the witnesses who are signed on the document, and thus we come back to the objections that were raised above.
A resolution of this difficulty follows from a distinction proposed by the Tosafot:
That which is taught "they were disqualified from giving testimony" – they do not say that they are still disqualified, for if so why are they not believed. Rather they had been disqualified from giving testimony, but now they agree that they are qualified.
The Rid in his Tosafot (printed below on p. 20b) expands upon this idea:
Certainly when they say to them "You are thieves" or "You are relatives," they are believed, because they completely remove them from the category of testimony. But when they say "You were minors" or "You were relatives," since they admit that now they can give testimony, and testify only about that time [in the past], they are two against two. For now [the witnesses signed on the deed] are qualified to give testimony, and they say: "We were not minors," "we were not relatives." And therefore we do not believe the second [set of witnesses].
How are we to understand this rationale? At first glance, we seem to be dealing here with a distinction between disqualification that leaves the first witnesses disqualified from testifying now on the matter, and disqualification that leaves them qualified to testify at this time, and therefore they can stand up to the second witnesses and argue with them as fit witnesses, thereby creating a situation of two against two.
This interpretation, however, is incorrect. Surely we are dealing with witnesses signed on a document who are already dead, and will give no further testimony. The power of their testimony stems from their signatures at the time in question, and nothing more. What then is the difference between localized disqualification relating to the time of signature and sweeping disqualification even now?
It seems that according to the Tosafot and the Rid, the difference between testimony that the witnesses are disqualified even now and testimony that they had been disqualified at the time in question is a difference relating to the essence of the testimony. In order for the testimony of the second set of witnesses to overcome that of the first set, it must be defined as testimony that in its essence relates to the witnesses and not to the matter in question. When Reuven and Shimon establish that Levi and Yehuda are disqualified from giving testimony, this indeed is testimony that is focused on a disqualification of the witnesses themselves. And since this has broad ramifications beyond the specific issue that must be decided now by the court, it is a topic for separate discussion. But when all they do is testify that Levi and Yehuda were disqualified from giving testimony at the time of the signature, the focus of the matter is not the status of Levi and Yehuda as witnesses, but rather the question whether or not the deed before us is valid. In this context Reuven and Shimon's testimony has the same status as that of Levi and Yehuda.
This principle seems to underlie the positions of other Rishonim as well. Let us examine the words of the Ramban:
What does it mean here - "disqualified from giving testimony"? Relatives. For if you say gamblers, what is the question: "They are two against two"? Can't they disqualify them with the accusation of thievery? For certainly if they say: "We know that they are thieves" – they are believed. And they cannot say about themselves: "We are not thieves." Rather it must be that they say [that they are] relatives. And they come to disqualify the deed. And the others testify that it is a valid deed, and there is a presumption that people do not sign documents unless they are unrelated. And therefore we say: Stand the two against the two, and stand the money in the presumptive possession of its owner.
The Ramban does not distinguish between testimony that the witnesses were disqualified when they signed the deed and testimony that they are still disqualified now, but rather between disqualification based on theft, regarding which the second set of witnesses overcome the first set, and disqualification based on familial relationships, regarding which the situation is defined as a case of "two and two." The Ramban explicitly explains the basis of the distinction: When the second set of witnesses testify that the first set are relatives, we say that "they come to disqualify the deed, and the others testify that it is a valid document." That is to say, we have no independent interest in knowing that Reuven is related to the plaintiff. Everybody has relatives, and it is not an independent topic of discussion. The question discussed here is whether or not the document before us is valid, and on this question the first set of witnesses have the same status as the second set. This stands in contrast to the disqualification based on theft, for if we accept it, we define Reuven as disqualified from giving testimony, and this is a separate topic of discussion, regarding which Reuven is not believed.
A question may be raised against the Ramban: Granted regarding the claim that the original witnesses were disqualified from giving evidence, for we can distinguish between thieves and relatives. But what can we say about the other case in the Baraita: "They were minors"? Is it not clear that the question whether Reuven is an adult or a minor constitutes a separate topic of discussion? How can Reuven testify about himself that he is an adult? At this point we may be forced to add within the framework of the Ramban a distinction similar to that of the Tosafot, and say that the first set of witnesses are only believed when the second set testify that they had been minors at the time they signed the document, but they are not necessarily minors now.
It should be added that the Ramban's argument that the expression "disqualified from giving testimony" relates to disqualification based on familial relationship, and not to disqualification based on theft, is directed against Rashi, who in one place proposes that we are dealing with gamblers (18b, s.v. pesulei). In the continuation, the Ramban tries to explain Rashi's position:
And some say according to Rashi that here, since, they died, they are not believed to disqualify them based on theft, because they are testifying about the maneh in the deed. But when they are alive they can disqualify them, for they are not believed to say, "We are not thieves," for they have an interest in their testimony. And therefore the Baraita states: "and they died."
Here we have a distinction between the situation in which the first set of witnesses are dead and the situation in which they are still alive. And once again, the Ramban explains the foundation of the distinction. When the first set of witnesses are dead, there is certainly no reason to enter into a separate discussion regarding their personal status as valid witnesses, and our entire interest is the loan now under discussion in the court: "Because they are testifying about the maneh in the deed," and here the first set can stand up to the second set as equals. But when the first set of witnesses are still alive, the testimony of the second set is seen as focusing on the first witnesses' personal fitness, and therefore it wins out against the first set.
II. A Miggo in the Case of "Two Against Two"
Another question that troubled the Rishonim on our passage, and reappears in the Rishonim in many other passages, accepts that the situation is defined as a case of two against two, but asks why don't we decide the doubt in favor of the second set of witnesses, who have a miggo that had they wanted to lie, they could have disqualified the first set by accusing them of thievery. We cannot exhaust this issue here, but we will try to outline several approaches to it.
The Tosafot (ad loc., s.v. ve-im) write:
You might ask: In any event, why are they [the second set] not believed? Let us believe them based on a miggo that had they wanted [to lie] they could have said that they are still disqualified? And similarly in every case of two sets of witnesses who contradict each other – why don't we believe the second set based on a miggo that had they wanted [to lie] they could have disqualified the first with the charge of thievery? The Ri says that with two witnesses there is no miggo, for the one does not know what is in the heart of the other.
According to this answer, the idea of miggo does not apply here, because we cannot build on the fact that they could have put forward a different claim, when the one doesn't know what is in the heart of the other. Why, then, when the signatures of the witnesses of the shtar are not authenticated are the second witnesses believed to say: "We know that it is their handwriting, but they were under duress"? The Tosafot Yeshanim (ad loc.) extend the answer of the Tosafot, and write: "When their handwriting does not come out from another place, they are believed, based on the miggo that had they wanted [to lie], they could have remained silent." It seems that the Tosafot Yeshanim are not referring to the special power of the miggo that they could have remained silent, which we saw at the beginning of the chapter. But rather they mean that the technical difficulty noted by the Tosafot only applies to an alternative claim based on speech, but not to an alternative claim based on silence, regarding which special coordination between the two is unnecessary. On the previous page, the Tosafot themselves proposed a different explanation of the first clause of the Baraita: "Since the Sages required authentication, here their saying, 'This is our handwriting,' is not considered authentication, because they themselves immediately say, 'we were minors' or 'we were under duress'" (18b, s.v. harei).
The Rishonim were not satisfied with the Tosafot's answer, and so they proposed many other reasons for not believing the second set of witnesses based on a miggo. Let us examine the words of the Rashba in Bava Batra, who alongside the Tosafot's answer brings three additional answers:
The witnesses are not believed based on a miggo more than they are believed based on their testimony in such a case. They only said that witnesses are believed based on a miggo in order to overcome the problem that they have an interest in their testimony, or the like. But in this case, not.
And furthermore, why should we believe these witnesses based on the laws of testimony more than the others? Surely the Torah does not believe certain witnesses more than other witnesses based on a miggo.
And furthermore, there is no miggo here, for it is a miggo where there are witnesses, and we do not say [such a miggo]. (31a, s.v. amar lei)
The Rashba's first answer asserts that the force of testimony is based exclusively on the personal trustworthiness of the witnesses, and not on some external-circumstantial proof. The reason that witnesses are believed is not that if two people say something, it is reasonable to assume that it is true, but rather that the Torah regards their testimony as the ultimate proof based on the personal trustworthiness of man. Therefore, their testimony is not made any stronger by some circumstantial proof such as miggo, which establishes that it is reasonable to assume that they speak the truth, for had they wanted to lie they would have had a better way to do so.
According to the Rashba's second answer, the doubt created by the conflicting testimony of two sets of witnesses cannot be resolved by way of a miggo. For surely two witnesses are like a hundred witnesses, and even if we add a hundred witnesses to the one set, this will not strengthen it against the second set of witnesses. And there is no reason to assume that adding a miggo should help more than the addition of another hundred witnesses.
The Rashba's third answer is based on the second set of witnesses standing up against the miggo and canceling it, for a miggo cannot stand up to witnesses.
Let us conclude with yet another answer, cited by the Tosafot Yeshanim (ad loc.):
That the second [set of witnesses] is believed is anomalous, for why accept the evidence of the one pair rather than that of the other. Therefore, they are not believed based on the miggo that they could have disqualified them by accusing them of thievery, unless they say so explicitly.
In the previous shiur, we saw the words of Tosafot (19a, s.v. modeh) in the name of Rashi, that when there is a claim regarding which a person does not have built-in credibility, only that when it is put forward it is accepted – we cannot believe a person who does not put it forward, based on a miggo that he could have put it forward, because this is not a claim that he has "in his pouch." The Tosafot said this with respect to the miggo of forgery, and we mentioned that a similar principle appears in other Rishonim in various contexts. In the context of our question as well – why not decide the doubt created by "two against two" based on a miggo – we can see something similar to this in the first answer of the Rashba that we saw above: The power of witnesses itself is an anomaly, and the Torah recognized this anomaly with respect to things that they say, but not to things that they could have said. It seems that a similar principle also emerges from the words of the Tosafot Yeshanim: That we should believe the second set of witnesses who claim that the first set of witnesses are thieves is an anomaly, for there is no special reason to believe the second set any more than the first, and therefore miggo does not help.
The words of the Tosafot Yeshanim are, however, exceedingly difficult, for the Gemara whose wording they cite – "for why accept the evidence of the one pair rather than that of the other" – explicitly contradicts what they say:
If a witness has been proved a zomem, Abaye says: He becomes disqualified retrospectively [from the time when he gave his evidence in court]. Rava says: He is disqualified only for the future [from the time when he is proved zomem]. Abaye makes the disqualification retrospective on the ground that the witness has been shown to have been wicked at the time when he gave evidence, and the Torah says: "Do not accept the wicked as a witness." Rava, on the other hand, holds that the disqualification begins only from the moment when his deceit is proved, because the whole procedure of proving witnesses zomemim is anomalous. For this is a case of two witnesses against two; why then accept the evidence of the one pair rather than that of the other? At least let it take effect only from the time when the anomalous procedure is employed.
Some say that Rava really agrees with Abaye that the disqualification is retrospective, but rejects here this principle on practical grounds, because its adoption might adversely affect purchasers.
What practical difference is there between the two versions? Where two witnesses have proved one of a pair zomem, and other two witnesses have proved the other one of the pair zomem; or again, where the disqualification of the witnesses is based upon an accusation of thievery brought by a subsequent pair. According to the version that Rava based his view on the fact of the procedure being anomalous, it would not apply here, whereas according to the version which makes his reason the fear of adversely affecting purchasers, it would hold good even here. (Bava Kama 72b-73a)
The Gemara states that even according to Rava, who argues that the whole procedure of proving witnesses zomemim is anomalous, for "why accept the evidence of the one pair rather than that of the other," this relates only to the procedure of proving witnesses zomemim. But when the second set disqualifies the first set based on an accusation of thievery, there is nothing anomalous about believing the second set of witnesses, for they are certainly dealing with the witness itself and not with the case at hand, and therefore they act on a plain regarding which the first set cannot at all give testimony. How then can the Tosafot Yeshanim cite this Gemara and assert that the power to disqualify witnesses with the accusation of thievery is anomalous?
We must say (and so I once heard from Rav Gigi, shlita) that when the second set of witnesses claim that the first set of witnesses are thieves, they are no longer in a situation defined as two against two and turn into a set of witnesses dealing with other people. As long as they do not make this claim, they remain in the state of two against two, and we cannot give them greater power based on the fact that the definition of the situation could have been different. This position, however, remains very difficult, and requires further examination.
(Translated by David Strauss)
 See Tosafot, Gittin 18b (bottom), s.v. amri.
 The Ri Migash (brought in the Shita Mekubetzet, s.v. u-magbinan, and in Tosafot Rid, in the section printed on p. 20a) proposed that we can distinguish in the context of our passage between the testimony of "they were under duress," which indeed creates a situation of two against two, because the second set of witnesses essentially disagree with the substance of the testimony of the first set, and the testimony of "they were minors," or "they were disqualified from giving testimony," which does not create a situation of two against two. However, as noted by the Rid, even though this distinction is attractive, it is refuted by the passage itself, for the Baraita asserts that in all three cases the witnesses who come to disqualify the witnesses who are signed on the deed are not believed, and the Gemara qualifies this and says that they are believed, only that there are other witnesses who are also believed, and therefore we have a situation of two against two. According to the Ri Migash, we should have accepted the words of the witnesses absolutely. This matter requires further study.
 The Gemara above, p. 16b, states: "If whoever is married [as] a virgin is talked about, when witnesses come, what of it? These witnesses are liars!" According to some Rishonim, we learn from here that some levels of anan sahadi are stronger than witnesses and can define them as false witnesses (see Rashba, Bava Metzia 5a, s.v. hahu). But here we are not dealing with such a level, for we do not define the witnesses as false witnesses, but rather we see the assessment of reality (umdena) as equal to witnesses and creating a situation of two against two.
 See, however, Tosafot, above, p. 18b, s.v. harei, who, at least in their question, adopt the view of the Rosh. See also Tosafot Shantz and Tosafot ha-Rosh (ad loc.).
 See also the wording of the Rambam (Hilkhot Eidut 7:7): "The two [witnesses] who are signed on the deed testify against the two [witnesses] who testified about them that they are disqualified [from giving testimony]."
 The Ramban implies that this is true even if they testify that they are still relatives. It should be noted, however, that a disciple of the Ramban, the Re'a in his novellae ad loc., combines the Ramban's distinction with that of the Tosafot, and writes that only when the second set of witnesses testify that the first set had been relatives at the time of the signature, but are no longer related, do we say that it is a situation of two against two.
 The Ramban adds a line that requires further study: "And there is a presumption that people do not sign documents unless they are unrelated."
 The Tosafot in Bava Kama (72b, s.v. ein) relate to our passage, and distinguish between relatives and thieves as does the Ramban, but they do not explain their distinction. In the wake of the difficulty that we raised against the Ramban, there is room to consider a different explanation for this distinction. We may have to qualify our assumption that a person is not believed to testify about himself. A person is certainly not believed when he comes to court and gives testimony, the subject of which is himself. But there is room to say that every testimony includes a certain statement on the part of the witness about himself. First of all, when Reuven testifies that Shimon lent money to Levi, he is saying two different things: The main testimony – Shimon lent money to Levi; and the secondary testimony – I saw this with my own eyes. The second component is essential for validating the testimony, for if not, the testimony could be invalid: hearsay testimony, testimony of knowledge without having seen anything (see Bet Yishai, no. 109). And it may be added that when a witness gives testimony, he also asserts that he has the basic identity of a qualified witness – he is telling the truth and not under duress, he is not disqualified from giving testimony (he is neither a minor nor a relative). This stands in contrast to the disqualification of a thief, which is based on concern about the witness's trustworthiness, and not his personal identity as a witness, and therefore he is not believed.
 The Rashba in Bava Batra (31a, s.v. amar lei) suggests a variation: "Here it is different, for even if one of them says, 'This is not my handwriting,' the deed is invalid."
 These answers also appear in other Rishonim on our passage and on other talmudic passages (see, for example, Tosafot, Bava Kama 72b, s.v. ein). We chose to cite from the Rashba, because he clearly distinguishes between the three answers.
 See Rambam, Hilkhot Yesodei ha-Torah 7:7; 8:2; Hilkhot Sanhedrin 24:1.
 Ran on the Rif, Kiddushin 27b in Alfasi, s.v. li-nesu'in himneh (the Kovetz Shiurim, II, 3, 2-5, already noted the connection between the Ran and the aforementioned Tosafot); see also Ramban, Bava Batra 34a, s.v. ve-yesh she-pirshu.