Siman 147

The Rambam wrote in chapter 8, law 3: How does a man divorce upon a condition? Not that he should say, "Write a get for my wife on this condition," or "Write and give it to her on a condition," and it goes without saying that he should not write within the get, "On this condition so-and-so divorced so-and-so." Rather, how does he do it? He tells the scribe to write and the witnesses to sign, and they write a valid get without any condition in the world, and afterward he gives her the get and says to her, "Behold, this is your get," or "Behold, you are divorced with this on the condition of such and such," or he says to them or to the messenger, "Give her this get on the condition of such and such." Thus far his words. And so wrote the Shulchan Aruch in section 1; see there.

He further wrote: If he wrote the condition in the get after he finished writing the essence of the get, this is a valid get, whether he wrote it before the signing of the witnesses or whether he wrote it after the signatures of the witnesses. But if he wrote it before the essence of the get, even if he wrote "on condition of such and such," this is a doubtful divorce, for he has retained a right in the body of the get. And likewise, if he stipulated orally before the writing of the essence, this is a doubtful divorce. Thus far his words. And it was transcribed in the Shulchan Aruch, section 2; see there.

And afterwards the Rambam writes many laws regarding conditions and writes the laws of conditions that invalidate the get, such as if he made a reservation in the get by saying "except for so-and-so" or "on the condition that you do not marry so-and-so," and these laws were explained in siman 137. And afterwards in law 16 he wrote, and this is his wording: Anyone who divorces on a condition that invalidates the get, such as if he stipulated to her that she shall not eat meat and shall not drink wine all the days of her life, or that she shall be permitted to any man except for so-and-so, or if he stipulated other conditions upon her before the writing of the toref—if the condition was written in the get and he went back and erased it and gave it to her, she is doubtfully divorced; and if the condition was oral, he takes the get from her and gives it back to her without any condition at all or with a valid condition. Thus far his words. And it was transcribed in the Shulchan Aruch, section 3, and other opinions were not cited at all. And now we will explain, with the help of Heaven, the opinion of the Rambam, of blessed memory:

The Rambam, of blessed memory, holds that before the toref, any condition invalidates the get, whether he wrote the condition in the get or stated it orally, as he explained the reason: because he retains a right in the essence of the get. He means to say that the Torah said, "and he shall write for her a scroll of severance," that at the time of writing there must be a complete severance between him and her; and since he imposed some condition before the toref, it was written upon a condition, and if it is not fulfilled there will be no severance, so he was not completely severed from her at the time of writing. Furthermore, there is a concern of bereirah in this, for when the condition is fulfilled we would need to say it was clarified retroactively that it was written for a complete severance, and regarding Torah matters we establish that there is no bereirah (the Maggid Mishneh). This reasoning is found in the Yerushalmi (Chapter 3, Halacha 1), and even though there is a dispute there regarding this, nevertheless it is possible to interpret there that the conclusion is also thus, see there (that Rabbi Chananya and Rabbi Yochanan hold thus according to Rabbi, and Rabbi Yudan rejects this, and Rabbi Chinana upholds their words; and even though this is according to Rabbi and not the Rabbanan, nevertheless regarding the concern of before the toref even the Rabbanan agree, and like Rava in our Shas, page 85b, according to his system which will be explained before us. And because this reasoning is not clearly established, therefore the Rambam wrote it is a doubtful divorce):

And according to his words, that which is found in the Shas (84b) where Rava said to those who write a get, "Silence the husband (so that he does not mention any condition) until you have written the toref of the get," see there, it is not due to a decree as will be explained according to those who hold not like the Rambam. Rather, from the essential law it is so, for before the toref, any type of condition invalidates the get even orally. And this is what Rava said there: "The dispute is regarding after the toref, etc., but before the toref, all agree it is invalid," see there, for it is so due to the essential law. And because it is possible to interpret that it is not from the essential law but rather for the reason of a decree, therefore he wrote that she is doubtfully divorced. And the invalidity of that which is before the toref applies even when the condition was fulfilled, and even orally; and all the more so when they wrote it, even if he went back and erased it, it is invalid.

And after the essential part, certainly ab initio they should be careful not to write any condition in the get, and not by law but because not everyone is expert in the difference between after the essential part and before the essential part, and between a condition that nullifies the get, such as "except for so-and-so" or "that you shall not drink all the days of her life," and an ordinary condition (Haggahot Maimoniyot). But ex post facto we have no concern regarding all conditions that do not nullify the get, for in writing and orally they are equal. And even if they wrote the condition before the signatures, even though it appears as if the witnesses signed upon a condition, nevertheless we do not fear this, for we establish that "on condition" effects severance and this is not like a forgery from within. And regarding after the essential part, it is valid in one of two ways: either when the condition was fulfilled, or it was not fulfilled and he canceled it and went back and erased it. And orally when it was not fulfilled, several detailed laws were explained in siman 143 regarding this. And a condition that nullifies the get, if he wrote it in the get even after the essential part, even if he went back and erased it, she is a case of doubtful divorce; and all the more so when he did not erase it, the get is void by Torah law (Beit Shmuel, subsection 3). But orally after the essential part, he takes the get from her and returns and gives it to her without a condition or with a valid condition, for the get was not disqualified. But when he wrote it within it, even after the essential part, this is a doubt by Torah law, for perhaps by this writing the entire get was nullified. And there is one from whose words it is implied that according to the Rambam, regarding conditions that nullify the get where he wrote it after the essential part and erased it before the giving, she is not a case of doubtful divorce, and only if it was not erased at the time of giving does it not help if he went back and erased it (the Gra, subsection 3 in the middle of the discussion, and this is his wording: since he did not erase it at the time of giving, etc.). But in the Maggid, Halacha 16, it is explicit as I have written, and this is his wording there: Behold, if he wrote a condition in the get, if it is from the conditions that nullify the get, etc., in whatever place it was written, even though he went back and erased it and did not mention it at all at the time of giving, the get is disqualified and has no remedy. Thus far his words (and so it appears from the Beit Shmuel, subsection 4). And all this is according to the system of the Rambam, and so appears the opinion of the Rif and the Rosh, of blessed memory.

The Tur brought from a responsum of the Rif who was even more stringent, that even if it was written after the toref, and even if it was written on its back, it invalidates the get. This is the wording of the Tur after bringing the words of the Rambam: And more stringent than this is Rav Alfas, who wrote in a responsum regarding one who gives a get on condition etc. but to write the condition within the get or on its back, he is not permitted, for if he wrote it within it, the husband has a right in it and it is not "karetut." Thus far his words. And these are puzzling matters, for even if we explain that he did not mean literally that it is not "karetut" but rather a rabbinic invalidation, from where do we have this? For the most stringent authority in the Gemara there is Rava, and he says that regarding after the toref, Rabbi and the Sages disagree, for Rabbi decreed on account of before the toref and the Sages did not decree. And it is known that the halacha follows the Sages. And granted that ab initio one should not write even after the toref because Rabbi invalidates it, and regarding gittin we seek to satisfy all opinions—and this is the reason of the Rambam that ab initio one should not write a condition in a get as he wrote—nevertheless, to invalidate it ex post facto we have not heard. And the Rif himself ruled so in the chapter "HaMeguresh," see there, and it requires further study. Therefore, in practice it is difficult to be lenient since it issued from the mouth of the Rif in a responsum and the Tur brought it.

And the approach of our teachers, the Tosafists, is as follows: that a disqualification from the Torah exists only in a condition that nullifies the get and was written before the toref; but if it was written after the toref, even if he said this beforehand, it is not disqualified from the Torah and is not considered as having been written not for the sake of severance, for the intent of the scribe is not for the sake of this condition at the time of the writing of the toref since it was not yet written in the get (Tosafot 84b, s.v. "aval"). And the reason for the matter is that the scribe and the witnesses write only based on the intent of what the husband will desire in the end (Rashba ibid.), and this is from the Torah.

However, by rabbinic decree, any type of condition is invalid before the toref when written in the get, as a decree on account of a condition that nullifies the get, such as "except for so-and-so." And a condition that nullifies the get is invalid even after the toref when written, as we decree on account of before the toref. This applies to writing, but orally, only a condition that nullifies the get invalidates. Similarly, any condition before the toref that nullifies the get, even if stated orally and even if he subsequently revoked it, is invalid as a decree on account of writing before the toref. And that which we said, that any condition before the toref invalidates when written in the get, this is when it was not fulfilled and he erased it, even though he revoked the condition, for we decree on account of a condition that nullifies the get, for which erasing is of no use since it was not written for the purpose of severance (Tur in the name of the Ri). But when it is fulfilled, we have no concern with it, even if it was written before the toref. And any condition after the toref is valid when written, even if it was not fulfilled, provided he subsequently erased it and revoked the condition. And some say that any type of condition is invalid before the toref even orally, for we decree on account of a condition that nullifies the get (Rashba). And even though this too is only a decree lest he write it as I have written, nevertheless, the essence of this invalidity is from the Torah when he does not revoke it; and even at the time of giving, when he says "except for so-and-so" or "that you shall not drink all the days of my life," the get is void from the Torah. Therefore, we are stringent by rabbinic decree regarding any condition even orally if he made the condition before the toref. However, many of our teachers—the Ra'avad, and the Maor, and the Ramban, and the Rashba—hold that other conditions orally do not invalidate at all even before the toref, for we do not decree to such an extent (the Gra, ibid.).

And what is called "before the toref," the Tur wrote that prior to "and behold you are permitted to any man" is called before the toref. However, the Rashba wrote in a responsum that this is not called the toref (Beit Yosef), for even without this there are several phrases in the get that are significant like "and behold you are permitted to any man," and in the ancient forms of gittin this phrasing was not written at all, and Rabbeinu Tam instituted to write this (26a). And we have already written this in siman 125.

And behold, for the halacha we certainly act stringently according to the words of the Rif in a responsum, that one should not mention any condition in a get, and even orally it is forbidden to mention any condition prior to the writing of the get, according to the words of the Rambam and the Shulchan Aruch; and if he mentioned it, it is a doubtful divorce. And if he is hastening on his way and is compelled to go before the writing of the get and wants to command to write a get and deliver it to an agent on a condition, then he should command that they write it and give it into the hand of the agent and he should not mention any condition, neither before the scribe nor before the witnesses to the signing. However, afterward he should say to the agent in the presence of other witnesses: "That get which I commanded to give into your hand, you shall deliver to my wife on such-and-such condition," and the scribe and the witnesses to the signing will not know of this at all (Beit Shmuel subsection 1 in the name of the Maharit ibn Lev). And there are those who hesitate regarding this, for ultimately he mentioned the condition prior to the writing of the get. Therefore they wrote that he should also deliver it to the agent without specification and say to him: "Do as you wish, either deliver it to her without any condition or with a condition," but beforehand he should reveal his intent to the agent that his will is for such-and-hersuch condition; and automatically, when he delivers it to the agent, even though he gives him permission to do as he wishes, presumably the agent will act according to his will (Knesset HaGedolah). And behold, there are other ways in this matter, but in truth one should distance oneself from any condition in a get, and so is the simple custom among us not to give a get on a condition, except in the case of a person facing death when he is a kohen, as I wrote in siman 145, for with a condition the matter is close to coming to a stumbling block, and I have never heard of giving a get on a condition. And so one should conduct himself, except in a place of pressing need and agunah where it is not possible otherwise.

There is one who says that regarding what the Rambam wrote, that with an oral condition one must take it back from her and deliver it to her without the condition or with a valid condition, this is specifically regarding a condition of "except for so-and-so," where the first giving is effective to disqualify her from the priesthood. However, with a condition of "all the days of your life," which is not effective for anything, he does not need to take it from her; rather, he nullifies the condition and says to her, "Behold, this is your get without the condition that I stated," and that is sufficient. For it was thus explained in siman 137 and 138, that whenever the first giving was of no use, he does not need to take it from her (Beit Shmuel, subsection 6). And so too with all conditions that do not nullify the get, if he nullified the condition, he does not need to take the get back from her, but rather returns and says to her, "Behold, this is your get" (ibid.).

There are those who say that if he made some condition before the toref, and according to the Rambam the get is invalid and according to those who disagree with him it is valid, as has been written—if the condition was not made according to its law, even if there is a doubt whether it was made according to its law or not, and this get was delivered into her hand, the get is valid by reason of a double doubt: a doubt whether it was made according to its law, and even if you say it was made according to its law, there is a doubt whether the halacha follows the Rambam or those who disagree with him; and when the condition is fulfilled or he nullified the condition as has been explained, it is valid (Knesset HaGedolah in the name of Rabbi Y. ibn Lev). And this is only according to those who hold that a double doubt permits in a case of a legal presumption, but according to those who hold that a double doubt does not permit in a case of a legal presumption, she is forbidden, for she is under the presumption of a married woman. And it is possible that even those who hold so permit here, for there are three doubts, since according to the Rambam as well it is only a doubt of divorce as has been written (and so wrote the Shach in Yoreh Deah siman 110 in the rules of double doubt, letter 32, and see there in the Taz section 15). However, it seems to me that this is not relevant to be called three doubts, for the two doubts—the doubt of the Rambam and the opinion of those who disagree—are one doubt: whether before the toref is valid or not.