It was stated (Bava Kamma 72): Abaye said: A zomeim witness is disqualified (for any other testimony) retroactively (from the time that he testified). Rava said: He is only disqualified from the time that he is found to be a zomeim.
The Gemora explains: Abaye said that he is disqualified retroactively, for it is at that time that he became an evildoer, for the Torah states: Do not place your hand with an evildoer to be a corrupt witness. Rava said that he is only disqualified from the time that he is found to be a zomeim, for his disqualification is itself a novelty (so why should we stretch it). This is because the two sets of witnesses are two against two, so why should we listen to the second set more than the first? Accordingly, we can only apply the disqualification novelty from the time that they become zomemin.
Tosfos asks: Why is it such a novelty that we believe the second set of witnesses over the first? They should be believed, for they have a migu that they could have disqualified the first set by testifying that the initial witnesses are thieves (“believe us when we say that they were with us, for if we would have wanted to lie, we could have said that they are thieves!”)!?
Tosfos answers that we do not apply the principle of “migu” by two witnesses, for each one of them is not aware as to what the other one is thinking.
Furthermore, it is similar to a “migu against witnesses,” where the migu is not effective, and since in this case, the first set of witnesses are testifying that they did in fact witness the event, they are clashing with the migu of the second set. In such types of cases, a migu is not effective.
The Tzlach answers that this is a migu for half a claim and that is why it is not effective. If they would have only testified that the initial witnesses were thieves, they would become disqualified, but they would not be liable to pay at all. However, if they would testify that the first set was not there at the time that they said the event took place, they would be disqualified and they would be liable to pay. Accordingly, the migu principle would not apply in this case.
Read more!
The Gemora explains: Abaye said that he is disqualified retroactively, for it is at that time that he became an evildoer, for the Torah states: Do not place your hand with an evildoer to be a corrupt witness. Rava said that he is only disqualified from the time that he is found to be a zomeim, for his disqualification is itself a novelty (so why should we stretch it). This is because the two sets of witnesses are two against two, so why should we listen to the second set more than the first? Accordingly, we can only apply the disqualification novelty from the time that they become zomemin.
Tosfos asks: Why is it such a novelty that we believe the second set of witnesses over the first? They should be believed, for they have a migu that they could have disqualified the first set by testifying that the initial witnesses are thieves (“believe us when we say that they were with us, for if we would have wanted to lie, we could have said that they are thieves!”)!?
Tosfos answers that we do not apply the principle of “migu” by two witnesses, for each one of them is not aware as to what the other one is thinking.
Furthermore, it is similar to a “migu against witnesses,” where the migu is not effective, and since in this case, the first set of witnesses are testifying that they did in fact witness the event, they are clashing with the migu of the second set. In such types of cases, a migu is not effective.
The Tzlach answers that this is a migu for half a claim and that is why it is not effective. If they would have only testified that the initial witnesses were thieves, they would become disqualified, but they would not be liable to pay at all. However, if they would testify that the first set was not there at the time that they said the event took place, they would be disqualified and they would be liable to pay. Accordingly, the migu principle would not apply in this case.