Shiur # 32: Trei U-Trei

                  by Rav Yosef Tzvi Rimon               translated by Menachem Weinberg

Our Sugya

In the course of the discussion, the gemara (Kiddushin  66a) discusses King Yannai's status as a kohen.  The gemara  raises the possibility that this was a case of 'trei u-trei,'  'two against two,' i.e., two witnesses said that his mother  was captured, and two said that she was not.

Rashi says that if his mother had appeared we would have  accepted her as eligible for marriage to a kohen based on her  chezkat kashrut (Her original status was one of eligibility,  and there is no clear-cut proof that this status changed)),  and her son would consequently be ruled eligible as a kohen.   However, because the mother did not come and our decision  focuses on Yannai himself, it is impossible to transfer a  chazaka from the mother and apply it to her son.  Therefore we  cannot establish Yannai as a valid kohen based on a chazaka.

Tosafot differ from Rashi and hold that we do transfer  the chazaka of the mother to the son, and he should  essentially be eligible.  However, rabbinic law prevents the  application of a chazaka in cases of trei u-trei.

We do not intend to relate directly to this machloket in  the forthcoming shiur, but rather we will relate to the  central issue of trei u-trei.  After we learn the basics of  trei u-trei, we will be able to understand where Tosafot and  Rashi differ.

Introduction

In the course of this shiur, we will discuss the power of  eidut, testimony.  As is well known, in other legal systems,  testimony is not given the same import as in halakha where 'on  the basis of two witnesses will a matter be established.'  We  will try to clarify whether in the case of trei u-trei perhaps  we do allow a decision based on circumstantial evidence or  other forms of verification.

The gemara in Bava Batra (31a-b) brings a machloket  between Rav Huna and Rav Chisda regarding a case of trei u- trei.  According to Rav Chisda all four witnesses are  considered false witnesses and can no longer testify.   According to Rav Huna, in future testimonies, 'this one comes  independently and testifies, and this one comes independently  and testifies.'

It seems that their machloket is simply whether trei u- trei is an invalidation of the testimony (psul ba-eidut), or  an invalidation of the witnesses (psul ba-eidim).  According  to Rav Chisda the witnesses themselves become invalid and  therefore can no longer testify.  According to Rav Huna, only  this specific testimony was contradicted and thereby  nullified.  Therefore the witnesses are valid for other cases.   In practice, we rule in accordance with Rav Huna, but as we  shall see, his approach requires further elaboration and  explanation.

The Question from Ketubot

In Ketubot (22a, according to Rashi) the gemara brings a  case of trei u-trei regarding the reliability of an individual  to testify.  If two witnesses say that he is reliable and two  testify that he is not, the gemara concludes, that this  individual is invalid to testify as a witness.  The obvious  question is why - Rav Huna clearly stated that trei u-trei  does not invalidate the witnesses from future testimony, but  rather (as the Rashbam in Bava Batra comments) leaves them in  the status of kashrut as per their chazaka.

The Bakh (siman 34) answers that in the standard case of  trei u-trei, there is no direct testimony against either of  the groups that says that they are invalid.  The eidim merely  contradict each other regarding an independent issue.  From  this we infer that one set of witnesses has lied and is  consequently invalid.  Since we don't know which set has lied,  we can apply chezkat kashrut.  However, in Ketubot, the groups  are arguing directly about the individual's validity.  In such  a case the stalemate is treated (by rabbinic law) as an  objective safek, which cannot be decided by a chazaka.

The Chakham Tzvi (see Ketzot 34:6) answers differently.   Normally we would accept the conclusion of the gemara in  Ketubot; that we do not apply chazaka where we have a safek  based on contradicting eidim.  However, in the case of trei u- trei the doubt is weaker than normal, since the witnesses have  two things in their favor (tartei le-tivuta): 1. The witnesses have a chazaka, 2. It is certain that one of  the groups is valid.  In other words, it is clear that there  are valid witnesses here, but we just don't know which they  are.  The chazaka builds on the assumption that each group is  the valid one.

Understanding the Bakh's answer - Approach #1 to trei u-trei

R. Akiva Eiger (Responsa 136-7) raises the following  questions: What if someone demands the return of a loan based  on a shtar (contract) and subsequently two witnesses say that  the defendant paid the loan while two other witnesses testify  that he did not.  Should we say that since this is a case of  trei u-trei, the plaintiff should prevail as usual due to his  proof offered by the shtar  Furthermore, what if two witnesses  come and say that the defendant paid up before the time, and  two say that he did not.  Should we say again that the  plaintiff wins because of the chazaka that a person normally  does not pay before the loan is due. (Bava Batra 5b).

R. Akiva Eiger answers (in one approach) that the chazaka  and the shtar have no effect and do not influence the outcome.   A doubt raised by trei u-trei can not be decided by other  indicators.  He explains that witnesses are the ultimate proof  and are considered more significant than other forms of  evidence.  Therefore it is not conceivable that a lesser proof  should outweigh the witnesses.  Even though you might think  that eidim and a lesser proof merge to improve the quality of  the evidence, this is not so.  We know that two witnesses are  like 100 - even if another 100 witnesses or groups come to  testify for one side, it is insignificant.  Clearly, lesser  proofs can not be any stronger than additional groups of  witnesses.

From this we can infer that if two witnesses testify that  someone is a valid witness, and two say the opposite, the fact  that this individual has an additional chezkat kashrut is  irrelevant because a chazaka can not have a greater effect  than eidim.  However, in our case, there is no direct eidut  regarding the kashrut of the witnesses.  Rather we merely  infer that one of them must be lying since they contradict.   Therefore we can apply the chazaka, and assume independently  that each one is valid.

We can bring a proof for this approach from Tosafot in  Bava Batra (31b s.v. Ve-zo Ba'a).  Tosafot ask why we don't  believe the second group of witnesses whenever they contradict  the first.  They should be believed because of a migo (see  shiur #27) - they could have testified that the first group is  invalid (e.g. thieves etc.).  In their first answer Tosafot  say that we don't apply migo where there are eidim.  In other  words migo will not help the second pair because they are up  against other eidim.  Again we see that eidim + a proof (like  migo) is irrelevant, because eidim themselves are the maximum  proof possible.

Understanding the power of witnesses

In all areas of Torah, we follow majority rule.  Why then  is testimony different where 2 witnesses have the power of  100?  Following the majority is based on estimation - the  majority rules what seems logical and reasonable.  But in  testimony there is no room for estimation - we need an  accurate reflection of the facts.  The Torah decreed that two  eidim can present the equivalent of a perfectly clear picture  of what happened, and we can even execute capital punishment  on this basis.  Now we can understand, that even if there are  many witnesses, it only gives us additional parallel pictures.   However, it does not qualitatively improve the accuracy (see  shiur #31).  (The strength of eyewitness testimony also stems  from the great stringency that was practiced in checking and  cross-examining the witnesses.)

Summary

We saw that R. Akiva Eiger understands that testimony  represents the ultimate proof, and therefore chazaka and the  like can not decide in a case of trei u-trei.  We also showed  a proof for this approach from Tosafot who say that in a case  of trei u-trei, migo is inapplicable.

Approach #2 to trei u-trei

Tosafot bring an additional answer: There is no migo when  a claim is made by more than one person.  It is possible that  the second answer argues with the first answer (see R. Akiva  Eiger).  Accordingly, the second answer would accept a valid  migo within the context of trei u-trei. 

It could be that the argument between the two answers  revolves around how we perceive contradicting eidut.   Everybody agrees that eidim are the ultimate proof; however,  the first answer (and Rabbi Akiva Eiger) hold that when we  have trei u-trei, we see both groups as still existing, and  therefore there is no value to additional proofs and chazakot  that attempt to support existent eidut.  The second answer  understands that in a case of trei u-trei we assume that the  witnesses have canceled each other out.  That is, we see both  testimonies as invalid and therefore non-existent because of  the contradiction.  Therefore there migo or a chazaka can be  decisive.  (This is known as "keg-man de-leitnehu" - see shiur  #19).

It is difficult to understand this second approach.  If  the eidim are considered non-existent, how can a migo give  credence to their claim!  It must be understood, therefore,  that even those who say ke-man de-leitnehu do not think that  we have erased the testimony as if they never existed.  The  eidim exist, but the special reliability that the Torah gave  them has disappeared.  Therefore, with a migo it is possible  to retrieve this status for one of the groups.

Chazaka in a case of trei u-trei

It could also be that the first answer in Tosafot is not  a proof for Rabbi Akiva Eiger's approach.  The Tosafot spoke  of the fact that migo can not decide between trei u-trei, but  perhaps that is limited to migo and does not apply to chazaka.

The Torah believed witnesses and therefore migo can not  add support to the reliability of the eidim (In a case of trei  u-trei their reliability falls off.  Migo can not return this  reliability in a place where the reliability from the Torah  has fallen off.)  However, maybe specifically for this reason,  side proofs can be decisive.  If we see the witnesses ke-man  de-leitnehu - as if they are no longer existent - then since  the testimony is denied, other evidence can 'take center  stage' and help arbitrate.  Therefore, if the testimony falls  away, the chazaka or the shtar etc. are useful as if no  witnesses ever came at all.  However, migo does not help us  for it is a support of the testimony which has been  disqualified.

Perhaps we can support this idea from Rabbenu Yona.  He  relates to a complex case in Bava Batra (32a) where a rumor  had circulated about a kohen claiming that he was the son of a  divorcee (thereby disqualifying him from kehuna).   Subsequently, two sets of eidim came with conflicting  testimony.  The gemara concludes that the kohen is assumed to  be valid by virtue of his chazaka.  The Rishonim ask: Why  shouldn't the rumor provide support for those eidim that  testify to his disqualification?  Rabbenu Yona answers that  the rumor is irrelevant because it is based on people's  general opinion and is therefore no better than two witnesses.   But then he adds (32a s.v. Ve-ata Ed): "But chazaka surely  supports testimony even though two witnesses that help them do  not, because chazaka is not the same type as testimony."  It  seems from this that side proofs can decide a case of trei u- trei, while all forms of testimony are invalid.

However it is possible to understand Rabbenu Yona  differently.  R. Yona did not say that the chazaka works  because it does not relate to the testimony, as we explained  in the Tosafot.  Rather, he said that "chazaka is not a type  of eidut."  It appears from his wording that no factual  verification can be made whatsoever, but here the chazaka is  not verification.  Since we see the eidim in a case of trei u- trei as nullified, we leave the case in its previous state -  i.e. the chazaka.  (Even though chezkat beirur is usually  better than chezkat hanhaga, in the situation of trei u-trei  this switches.  Since chezkat beirur deals with facts, it is  ignored where there are eidim.  But chezkat hanhaga does not  deal with the facts.) 

We explained the opinion which allows for other forms of  verification in the case of trei u-trei based on ke-man de- leitnehu.  Based on this we can also understand Tosafot (Bava  Batra 32a s.v. Anan).  The gemara says that when trei u-trei  argue whether or not the husband died, the woman can not get  married, but if she did marry she does not require a divorce.   Tosafot ask why - she should have to get divorced because she  has a chazaka of being an eshet ish (a married woman)?  They  answer that the chazaka that a woman is careful when she gets  married is stronger than the chezkat eshet ish.  However, we  must ask - how can we factor in this chazaka since we have  explicit witnesses - trei u-trei!  It must be that we view the  eidim as ke-man de-leitnehu, non-existent and therefore a side  proof can be decisive.

According to this approach, in cases of trei u-trei we  cannot bolster the testimony of the contradicted eidim.   Nevertheless, it is possible to verify the facts utilizing  other indicators.

It could also be that the rationale of this opinion stems  from a variant understanding of trei ke-mea (two eidim are  like a hundred): According to our first understanding the  power of two witnesses is absolute, and, therefore, even if  they bring more eidim, there will be no enhancement of the  evidence.  Perhaps, the power of two eidim is not considered  to be absolute. Nevertheless, two eidim are equivalent to 100,  since the testimony of 100 adds nothing qualitatively.   Therefore, when the eidim contradict, other forms of  verification can actually help to tip the scales. (See Kovetz  Shiurim Baba Kama 49)

Summary

We saw three possible approaches in deciding a case of  trei u-trei.  1. Eidim are considered as absolute proof.  Even when two sets  of eidim contradict we consider the eidim as ke-man de-itnehu,  the witnesses are still considered to be present.  Therefore,  no verification is irrelevant. 2. Eidim are considered as absolute proof.  However,  contradictory eidim are ke-man de-leitnehu, the witnesses are  canceled out and other evidence is decisive. 3. Eidim is not considered as absolute proof.  As long as  there are no contradicting witnesses, this testimony is the  best proof and no other evidence can have sway.  Only when the  testimony is contradicted and we can not use it is there  meaning to additional proofs

We introduced this shiur by quoting the argument between  Rashi and Tosafot regarding the use of Yannai's mother's  chazaka in a case of trei u-trei.  According to Tosafot,  according to torah law the chazaka of the mother can be  applied to Yannai since we view the eidim as non-existent -ke- man de-letnahu.  However, on the rabbinic level we consider  the eidim existent, and therefore cannot apply any chazaka.

According to Rashi, we would apply chazaka in the case of  trei u-trei if the chazaka related to Yannai directly;  however, we cannot apply the chazaka of Yannai's mother in the  case of Yannai.  Tosafot question this ruling based on a  gemara in Ketubot (13a) which states explicitly that the  chazaka of the mother affects the child.  Furthermore, if we  can determine (based on her personal chazaka) that Yannai's  mother is qualified for kehuna, it follows automatically that  Yannai qualifies as well.

We can suggest that according to Rashi there are two  types of chazakot.  One type is a chezkat beirur which uses  chazakot in order to verify the facts.  This type of chazaka  is a form of proof.  The second type is chezkat hanhaga which  guides the ruling when the facts cannot be verified.  When the  chazaka functions as verification, the chazaka of the mother  can be applied to the son as well, since the same events  determine both cases (see Ketubot 13a).  However, in our  gemara which is a case of trei u-trei, chezkat beirur cannot  be applied (if we consider the eidim ke-man de-itnehu).   Consequently, we are dealing only with a chezkat hanhaga.   Since the mother has a personal chezkat kashrut, we would not  disqualify her without sufficient proof.  Yannai, on the other  hand, is disqualified mi-safek, lacking a personal chezkat  kashrut.