Shiur #04: Masekhet Ketubot 3a - "Afke'inhu Rabanan le-Kiddushai'hu"

  • Rav Yair Kahn

Based on a Shiur Kelali by HaRav Aharon Lichtenstein shlit"a


The halakha of "afke'inhu" - the revocation of kiddushin - appears in our sugya with regard to a condition placed upon a get that was fulfilled through an oness (circumstances beyond one's control).  According to the first version cited by the Gemara, the get does not take effect according to Torah law, but Chazal revoked the kiddushin out of concern for the "tzenu'ot" and "perutzot" (cautions women who would avoid getting remarried fearing that an oness prevented the husband's return; promiscuous women who would get remarried disregarding the possibility that an oness prevented the husband's return).  We find similar examples elsewhere in the Talmud.  In Masekhet Gittin (33a), the Gemara addresses a situation of a husband who sends a get via a messenger and annuls the get after the messenger left, but before he handed it to the wife.  The Gemara rules that the get does not take effect by Torah law, but Chazal revoked the kiddushin, thus allowing the woman who received the get from the agent thinking that it was binding, to remarry.  In both these sugyot, the Gemara explains the process of "hafka'at kiddushin" (the revocation of kiddushin) as based on the principle, "kol de-mekadesh a-data de-rabanan mekadesh" – whenever one betroths a woman, he does so only in accordance with the rabbis' consent.  Before we deal with the meaning of this phrase, let us first address a comment in the Yerushalmi, which explains this halakha differently.


I.          The Yerushalmi's Position: The Power of Bet-Din


            The Yerushalmi explains the effectiveness of the annulled get on the basis of the general principle that the Sages are empowered to "uproot" that which is written in the Torah: "What is Rebbe's reasoning?  According to the Torah he can annul [the get], and they [Chazal] said that he cannot - their words uproot the words of the Torah" (Gittin 4:2).  The difference between the two approaches is clear.  The Bavli's explanation, of "kol de-mekadesh," is grounded in the person's desire to betroth in accordance with the will of Chazal, whereas the Yerushalmi approach is based upon the power of Chazal to nullify the marriage regardless of the husband's wishes.


            As stated, the Bavli's explanation does not build upon Chazal's power to override Torah law, but we may still consider the possibility that the Bavli accepts this approach partially.  The Gemara in Yevamot (110a) deals with a case of revoking one's kiddushin because he betrothed in an improper manner.  Tosefot there (s.v. "lefikhakh") cite the following question of the Ri: "Is the reason why they revoked it because 'kol de-mekadesh a-data de-rabanan mekadesh,' or because the Sages have the power to uproot something from the Torah - [a reason that we could infer] from the fact that it does not mention here, 'kol de-mekadesh a-data de-rabanan mekadesh' as it does in all cases?"  The Ri's question arises from the omission of the clause, "kol de-mekadesh a-data de-rabanan mekadesh" from the sugyot in Yevamot and in Bava Batra (48b) in their presentation of "afke'inhu."  The question thus becomes whether the halakha of "afke'inhu" in these sugyot corresponds to the "afke'inhu" in other sugyot.


            The Ri seeks to distinguish between the different cases.  The discussions in Yevamot and Bava Batra deal with cases of improper kiddushin.  Clearly, a person who betroths improperly has no interest in performing his kiddushin in accordance with Chazal's wishes.  We have no choice, then, but to use the principle of the Chakhamim's power to override the Torah.  By contrast, in our sugya, as well as that in Gittin, we deal with situations of a problematic get where the kiddushin was perfectly acceptable, and we may therefore employ the concept of "kol de-mekadesh."


            At first glance, if we are working with the ability to override Torah law, no distinction exists between kiddushei kesef (betrothal with money or objects of value) and kiddushei bi'a (betrothal through marital relations) with respect to the Chakhamim's power to override the Torah's validation of the marriage.  It is therefore difficult to understand the possibility the Gemara in Yevamot raises of drawing such a distinction.  Tosefot in Ketubot (s.v. "teinach") write: "Even according to the view… [in Yevamot 89b; see Tosefot there] that the Chakhamim are empowered to uproot something from the Torah, this applies only in situations where no transgression is involved, such as there… [in Yevamot]… But where there is a transgression involved, such as when she was betrothed through marital relations, what can we say?"  According to this approach, in a situation of kiddushei bi'a a restriction could perhaps be imposed on the Sages' power given the transgression that will result should they revoke the kiddushin.


            However, even the Ri concedes that in our sugya the revocation of kiddushin evolves not from the Chakhamim's power to override the Torah, but rather from the principle of "kol de-mekadesh," and there are several different ways to understand this principle.


II.         A Condition Set by the Groom - As If He Made the Kiddushin Contingent on the Chakhamim's Will


            The Ritva writes in our sugya: "The rabbis revoked the kiddushin from him - meaning, he is similar to one who said, 'Behold, you are betrothed if my father agrees,' in which case if he [the father] does not agree, she is not betrothed."  This explanation is drawn from the sugya in Kiddushin 63 regarding the case, of a betrothal on condition "that my father agrees."  How is this condition to be interpreted?


            We may consider two possibilities:


1. A standard tenai (condition) - like all other tena'im.  According to this understanding, we would require all the "mishpetei ha-tena'im," the regulations governing tena'im (according to the view among the Rishonim that a stipulation articulated in the form of "al menat" requires the "mishpetei ha-tena'im.")  According to the Rambam and Ra'avad, who maintain that one may retract his condition (as long as there is still a possibility to fulfill it) and would then not require a second betrothal, here, too, the individual could retract his tenai and wish for his kiddushin to take effect even if his father does not consent.  Similarly, it stands to reason, according to this approach, that this tenai would take effect only if the individual explicitly articulates it.


2. Subjection to the father's will.  The Gemara in Masekhet Nedarim teaches that one who takes an oath "al da'at rabim," or "in accordance with the will of the masses," has no possibility of "hatara" - having his oath annulled.  The explanation is that we do not have here a condition on the vow; rather, the individual subjects his will to that of the masses.  It is their will, therefore, which is required in order to have the oath annulled.  Quite possibly, an oath "on condition that my fathers agrees" operates in a similar fashion.  The individual subjects his will to his father's, and we therefore include the father's intention within "da'at mekadesh" - the intention upon which the kiddushin hinges.  Therefore, just as the intention of the groom is required as a basic ingredient of the betrothal, and is not an external condition, so too the father's consent becomes integral to the kiddushin. According to this approach, we do not require the "mishpetei ha-tena'im." Furthermore, the individual cannot later retract his condition, since we now include the father's intent as part of the "da'at mekadesh."  Similarly, even if the individual did not expressly articulate his stipulation, it may nonetheless take effect, once we assess his intent and discover that he wishes to subject his will to his father's.


            In our sugya, the individual does not articulate this condition subjecting his kiddushin to the will of the Chakhamim.  It would seem, therefore, that we cannot interpret it as an ordinary tenai.  But according to the second understanding, the individual's will may be subject to the will of the Chakhamim even without his having explicitly stated as such. 


Even in the absence of a verbal stipulation, we may nevertheless consider the possibility of a standard tenai, that the kiddushin should take effect only in accordance with the will of the Sages.  There are two opinions as to how this would work:


1. Tosefot establish that at the kiddushin ceremony the groom adds the phrase, "ke-dat Moshe ve-Yisrael" - "in accordance with the religion of Moshe and Israel" - in order to express the subjection of his will to that of the Chakhamim.  According to this view, we could understand this as a standard tenai.  Even according to Tosefot, however, it is unclear what would happen if the groom did not add this clause when performing the kiddushin.  It stands to reason that even if he omitted it, his will is nevertheless subject to the will of the Chakhamim.


2. Tosefot in several places claim that when we can make a clear presumption of a person's intent - "umdana de-mukhach" - we do not require the formalities of the "mishpetei ha-tena'im."  The umdana itself creates an outright tenai; hence, if this implicit condition is not met, the kiddushin become "kiddushei ta'ut" - a mistaken betrothal, performed with erroneous preconceptions.  Accordingly, we can assume that we have here an outright tenai, regardless of whether or not the groom expressed it verbally.


            The problem with this approach is that it does not allow for any distinction between kiddushei kesef and kiddushei bi'a, thus making it difficult to understand how the Gemara entertained such a distinction.  Indeed, the Ritva explains that no fundamental difference exists between these two methods of kiddushin, and the distinction lies only in the wishes, rather than the power, of the Chakhamim.  Namely, the Gemara considered the possibility that the Chakhamim would not want to revoke kiddushei bi'a, since they have no interest in transforming one's marital relations into a "be'ilat zenut" (sexual union outside the context of marriage).  The Gemara then concludes that in truth, this prospect does not trouble the Chakhamim, and "afke'inhu" can therefore apply in cases of kiddushei bi'a, as well.  Even initially, however, the Gemara believed that it was within the Chakhamim's power to revoke kiddushei bi'a if they so wished.


III.        The Individual Intends For His Get to Be Effective


            The Rishonim discuss how marriage can be retroactively revoked.  At first glance, such a possibility can cause many serious problems.  The Gemara often discusses the situation of one who marries a relative permitted to him - such as his niece - whom he would wish to protect from capital punishment if she betrayed him ("shema yechapeh al bat achoto").  If there exists the possibility of retroactive hafka'at kiddushin, in such a case the husband can simply revoke the kiddushin and thus spare his niece the death penalty.  Similarly, we could very easily eliminate the status of mamzerut (illegitimacy) by simply having the husband retroactively revoke his kiddushin after his wife betrays him!  The Rishonim from the Spanish school explain that in truth, the marriage is not annulled retroactively.  In their view, since the husband knows that the Chakhamim can revoke the kiddushin, he decides when giving the get to accept their rulings.  Therefore, even though the condition was fulfilled through an oness, the divorce takes effect nonetheless.


            Can a person stipulate from the outset that his get will take effect even if the tenai is fulfilled through an oness?  This issue would depend on how we understand the application of oness to tena'im placed on gittin (see previous shiur, #3).  According to the approach that we presume that the individual never had oness situations in mind when placing the condition, he could, it would seem, accept upon himself the validity of the get even should the tenai be fulfilled as a result of an oness.  How such a concession would work we can understand based on an analysis of the concept of "mekach ta'ut" - a transaction conducted on error.


            There are two forms of mekach ta'ut.  In some cases, the transaction automatically becomes null and void, and the individual involved need not even wish to have it annulled.  This occurs when a person makes a purchase thinking he receives wheat, and it turns out that he received barley; the transaction is annulled ipso facto.  In other instances, the individual can claim that his transaction was conducted erroneously, and then it is annulled.  But if he does not come forth with a claim, then the transaction remains intact.  Such is the halakha concerning the purchase of overtly defective merchandise (such as an ox purchased for plowing that is discovered to have a limp) or if after the purchase the individual acquires some information that had he known beforehand, he would never have gone through with the transaction.


            Into which of these two categories of mekach ta'ut should we place the Gemara's situation, of oness be-gittin?  The husband reached the river and found that it had overflowed its banks, and he shouted, "Look, I'm coming!  Look, I'm coming!"  Do we view the condition as having been fulfilled (since he didn't return by the stipulated date) only he has the claim of oness, or perhaps the tenai has not been fulfilled at all, since initially he intended for the get to take effect only if he intentionally does not return (as if he had expressly said, "… if I intentionally do not return")?


            According to the first approach, we could claim that from the outset, the individual foregoes on his ability to make a claim of oness, since he knows that the Chakhamim want the get to take effect.  According to the second possibility, we are forced to assume that the person agrees to forego on the condition that the tenai be fulfilled willingly, and the get therefore takes effect even if circumstances beyond his control prevented him from returning on time.


            In any event, according to this approach the Chakhamim merely express their wish that the get take effect; they do not exercise any authority.  The individual tailors his da'at in accordance with the Chakhamim's wishes, and by his own doing, the get is valid.


            This approach, too, allows for no fundamental distinction between kiddushei kesef and kiddushei bi'a, and even the Ritva's solution will be useless.


IV.       "Hefker Bet-Din Hefker"


            This approach explains that the Chakhamim revoke the kiddushin through their power of "hefker Bet-Din hefker," allowing the rabbis to confiscate personal property.  When dealing with a marriage initiated by kiddushei bi'a, only then do they employ their special power granted by the principle of "kol de-mekadesh."  Here we can readily understand the distinction drawn between kiddushei bi'a and kiddushei kesef, and the Gemara's progression becomes clear.  There is room to debate, however, whether or not we may apply "hefker Bet-Din" to this context.  There is a question whether the money transferred in kiddushin has to belong to the groom.  If actual ownership is not required, delaring the money ownerless would be insufficient in invalidating the kiddushin. Presumably, "hefker Bet-Din" would work here by having the Chakhamim declare that the kiddushin money belongs to the bride.  But the Rishonim debate the question of whether "hefker Bet-Din" empowers the Chakhamim to transfer property, or only to render it hefker (see Rashba, Gittin 36).  Additionally, this approach would require us to explain the Gemara's conclusion, applying "afke'inhu" even to marriages initiated through kiddushei bi'a. 


The Status of the Marital Relations


            With respect to the revocation of kiddushin effectuated through relations, the Gemara asks how the Chakhamim can retroactively transform the initial union into a be'ilat zenut.  The question, of course, arises, why did the Gemara ask only with regard to the initial bi'a?  Seemingly, "afke'inhu" retroactively renders all marital relations over the course of the marriage into be'ilot zenut!  We have two ways to resolve this question:


1.         True, the revocation affects all the marital relations.  But the Gemara had difficulty understanding only how the Chakhamim can directly transform a bi'a into be'ilat zenut; as far as the subsequent relations are concerned, they are delegitimized only as an indirect result of the Rabbinic action and do not require justification.


2.         Perhaps the relations conducted during marriage are not retroactively considered be'ilot zenut.  The Gemara in Masekhet Yevamot (107) discusses a debate between Bet Shammai and Bet Hillel concerning the halakha of "mi'un."  A young orphan girl may be married off by her mother or older brother, but has the option, even after the betrothal, of refusing the marriage, in which case the kiddushin are retroactively annulled.  Bet Hillel and Bet Shammai argue as to whether the girl has the option of mi'un even after nisu'in.  According to Bet Shammai, there can be no such thing as a conditional nisu'in, since a person is not willing to transform all his marital relations into be'ilot zenut.  Bet Hillel, however, allows for the possibility of mi'un after nisu'in, or a conditional marriage.  How would they respond to Bet Shammai's argument?  They could respond that the husband has no such concern, as he feels confident that his condition (in this case, the wife's satisfaction with the marriage) will be fulfilled.  Alternatively, they could perhaps claim that the interim relations do not become be'ilat zenut once the marriage is revoked.  This can be explained in light of the Rambam's comments in Hilkhot Ishut (1), where he defines a "be'ilat zenut" as an act of bi'a performed "for the purpose of zenut [non-marital intercourse], without kiddushin."  The Rambam appears to require two conditions for a bi'a to be considered a be'ilat zenut - "for the purpose of zenut" and "without kiddushin."  Accordingly, an unmarried man who has relations with an unmarried woman without the intention of zenut does not violate this prohibition.  This would include cases where a person sleeps with an unmarried woman thinking she is his wife.  As this was not done "for the purposes of zenut," but rather in a mistakenly presumed marital context, the prohibition of zenut has not been violated.  According to Bet Hillel, perhaps, if an individual has relations with a woman within a given framework - such as a conditional marriage - we cannot consider these relations be'ilot zenut, and he does not violate any prohibition.  The same may very well apply in our sugya: since the couple was married, their relations do not retroactively become be'ilot zenut in the event that the Chakhamim revoke the marriage.


Revoking the Status of "Mamzerut" Through "Afke'inhu"


            The mishna in Yevamot lists two ways of undoing the marriage of a young girl - through mi'un, and through her receipt of a get.  In the first case, the kiddushin is retroactively annulled; hence, both the husband and wife can marry the relatives of the other, and she may marry a kohen (as she is considered as never having wed, she does not have the status of a "gerusha" - divorcee).  If she receives a get, however, the marriage is terminated, not revoked.  Therefore, the relatives remain forbidden, and she may not subsequently marry a kohen.  Which of these models does hafka'at kiddushin resemble - the retroactive annulment of mi'un, or the termination of a get?  According to the approach that the Chakhamim are merely the cause for the effectuation of the get, then clearly we must consider the woman divorced, and she may not marry a kohen.  But what happens according to the other views?  It is possible that in the cases addressed in Bava Batra and Yevamot, the Chakhamim retroactively annul the marriage, whereas in the situations of problematic Gittin, the Chakhamim worked with the get and granted it rabbinic validation.


            Some Rishonim allowed for the elimination of the status of mamzerut through the mechanism of "afke'inhu."  The Maharsham suggested staging a situation of oness be-gittin in order to retroactively annul a marriage and thereby revoke the status of mamzerut.  Discussions regarding such a possibility appear in the "Seridei Eish" and "Minchat Shelomo."


            If we view "afke'inhu" as based on the power of Chazal to override the Torah, as the Yerushalmi claims, then clearly the possibility of implementing it today is very unlikely.  Therefore, we will encounter greater difficulty implementing "afke'inhu" in cases of improper kiddushin than we would in situations of a problematic get, where, according to the Bavli, "afke'inhu" operates through the principle of "kol de-mekadesh."  In any event, any application of "afke'inhu" must clearly have the backing of the leading scholars of the generation, and should not be applied without their support.

Translated by David Silverberg



1.         3b: "de-Tanya… lo akar nafshei"

2.         Tosefot, s.v. "ve-lidrosh"

3.         Sanhedrin 74a: "Amar Rabbi Yochanan mi-shum Rabbi Shimon ben Yehotzadak… hakha nami hana'at atzman shani"; Pesachim 25a: "Amar Rabbi Yaakov amar Rabbi Yochanan ba-kol mitrap'in… de-ha-hu gavra sumak tefei"

4.         Ba'al ha-Ma'or - Sanhedrin 17b in the Rif, s.v. "Abayei"; Ramban's response in Milchamot Hashem; Chiddushei ha-Ran - Sanhedrin 61b, s.v. "itmar": "de-ha af-al-gav de-be-avodat elilim amrinan de-yeihareig ve-al ya'avor… im hitro bo"




1.         In which situations does the halakha of "yeihareig ve-al ya'avor" (that one must surrender his life rather than transgress) apply?  Would relations between a gentile man and married Jewish woman fall under this category?

2.         According to the straightforward reading of the sugya in Ketubot, must an arusa (betrothed woman) surrender her life to avoid relations with a gentile?

3.         What are the various solutions raised by the Rishonim for this contradiction?

4.         On what basis does the Ramban distinguish between different situations of "yeihareig ve-al ya'avor"?