In discussing kiddush hachodesh (sanctification of the new moon), the gemara (R.H. 25b) equates testimony with the BEGINNING of din which must be done in the daytime, as derived (San. 34b) from the verse, "on the DAY that he causes his sons to inherit." (Devarim 21:16) Furthermore, even if judges themselves observed the new moon at night, they may not declare the new month on the basis of their observation alone, but must testify the following day verbally. Tosfot (s.v. k'gon) explains that this is because their observation is equivalent to testimony, and is not valid at night. [Similarly, the gemara (B.B. 113b) teaches that even if three people attended the final-will declaration of a person at night, they may not execute it on the basis of their observation alone without testifying about it in the daytime.]

The Talmud Yerushalmi, however, quoted by the Rosh (B.B. 8:3), writes that b'dieved (post facto), if judges mistakenly judged at night, their ruling is valid. Similarly, the Rashbam (B.B. 114a s.v. afilu) writes that if the court collected testimony at night, b'dieved they can judge the following day based on this testimony. The Shulchan Aruch (C.M. 28:24), however, rules that if testimony was collected at night it is not valid, even under the extenuating circumstances that the witnesses had to leave immediately.

The RaM"A (C.M. 5:2) quotes the opinion of the Yerushalmi that if judges wrongly judged at night their judgement is valid. Yet, regarding the collection of testimony, the RaM"A does not quote the opinion of the Rashbam. His silence implies that he accepts the decision of the Shulchan Aruch that testimony which was collected at night may not be relied on! Why does the RaM"A quote the lenient opinion regarding judgment at night, but not regarding testimony at night?

The SM"A (5:8 and 28:63) resolves this apparent contradiction in the following manner. In the first case, where the case was judged at night and the ruling was already delivered, it is truly post facto, and so the judgment is valid. But if testimony alone was collected at night, and beit din would now like to rule based on it, it is not totally post facto. We can further explain this distinction, that if the ruling has been issued, it achieves the status an act of beit din, whereas if the testimony alone has been collected, we still consider their statements as mere talk. (It would seem according to this that if testimony was collected and a ruling was issued based on it at night, it would be valid according to the RaM"A.) This distinction is difficult, however, since the Rashbam explicitly considered valid a case in which the testimony alone was collected at night.

To explain the SM"A, we must first analyze more carefully the gemara's contention that testimony is like the beginning of din. In Shavuot (30b), the gemara derives that witnesses must stand from the verse, "The two men [and those] who have the grievance shall stand before Hashem." (Devarim 19:17) Rav Huna this derives from the phrase, "[Those] who have the grievance," that the litigants must also stand at the rendering of the decision, "since witnesses are similar to the END of din." This gemara seems to contradict the gemara in Rosh Hashana which compares testimony to the beginning of din.

We must explain, therefore, that the gemara in Shavuot is not actually equating testimony with the end of din. Rather, it is teaching, by analogy, that whomever beit din is dealing with must stand: just as we find that the witnesses must stand when they testify, similarly, the litigants at the time of the ruling must also stand. It is also possible to answer that collection of testimony has both stringencies -- it must be in the day, like the beginning of din, on account of the judges, and the witnesses must stand, like the end of din, on account of the litigants. (See Pilpula Charifta, B.B 8:3 [10])

[It is interesting to note that regarding the issue of daytime the Torah is more stringent about the beginning of din, whereas regarding standing, the Torah was more stringent about the end of din. We can suggest a simple rationale, that a judge's main effort is in the deliberation, during which he clarifies the law, and therefore the Torah required that it be during the day. (The SM"A 5:7 suggests that if the lights are on it is possible to begin din even at night.) However, for the litigants, the main point of the din Torah is the decision -- as Moshe tells Yitro, "I judge between a man and his fellow, and I make known the decrees of G-d" (Shemot 18:16) -- and therefore the Torah required that they stand at the end of din.]

The status of testimony as beginning or end of din would seem, in fact, to be a dispute amongst the Tosfot. According to one answer in Tosfot (San. 19a s.v. Yanai, as well as Tosfot Shav. 30a s.v. she'im), only the witnesses themselves have to stand at the time of testimony, but not the litigants. This would fit in well with the answer above. In the second answer, however, Tosfot suggests that the collection of testimony is like the end of din even vis-a-vis the LITIGANTS, and that they too must stand then. According to this opinion, the contradiction between the two gemarot returns. We must then differentiate that in regards to STANDING, testimony is similar to the end of din, whereas regarding DAYTIME it is similar to the beginning of din.

Based on this, we can explain the SM"A's distinction. The Rashbam holds like the second opinion in Tosfot that collection of testimony is also considered an act of beit din, so that if it was collected at night it is valid, post facto. The RaM"A, however, rules like the first answer in Tosfot, that the collection of testimony is similar to the beginning of din in all respects. [In fact, we do not find anywhere in the Shulchan Aruch or RaM"A that the litigants must stand at the time of testimony.] Therefore, only a ruling issued at night is valid, but acceptance of testimony is not an act of beit din, but merely talk.

The Rambam, as well, seems to follow the first answer of Tosfot, that testimony is like the end of din. He writes (Hil. San. 13:7) that witnesses who testify that someone was already judged with a death penalty for murder in a certain court can be put to death, "provided that they testify in a court of twenty three [judges]." The Ohr Samayach infers from this that in general, the collection of testimony can be done by three even in capital cases, like Rav Saadya Gaon, against the Ramban (Devarim 17:6) who holds that all capital cases require collection of testimony before twenty-three judges. Only in this unique case that the testimony represents the entire rendering of the decision (since there is no room for discussion here other the validity of the testimony), we require acceptance before twenty-three. The fact that he does not require collection of testimony before twenty-three judges would imply that he maintains that testimony is not like the end of din. This is further evident from the fact that he does not require the litigants to stand at the time of testimony.

[In regards to kiddush hachodesh, the gemara teaches (R.H. 25a), "YOU are to designate -- even mistaken, even misled, even intentionally," and therefore Beit Din is permitted to confuse and scare the witnesses into testifying or contradicting their testimony. The Ohr Samayach (Hil. Kiddush Hachodesh 2:8) claims that, as such, kiddush hachodesh is unlike other judgments where the witnesses clarify the facts, and the judges render the verdict on the basis of the testimony. Here, the declaration "Mekudash! Mekudash!" includes within it a degree of acceptance of testimony and clarifying of facts. Based on this he explains the gemara (R.H. 25b), "You might have thought that cross-examining the witnesses is like the beginning of din, and "Mekudash! Mekudash!" is like the end of din, so that they could sanctify at night ... therefore, it teaches otherwise." Here, since even the declaration of mekudash has an aspect of beginning of din, it too has to be in the daytime.]