JOFA Fourth International Conference on Feminism and Orthodoxy

New York, November 10th, 2002



Agunah: Problems of History and Authority


Bernard S. Jackson


Centre for Jewish Studies

University of Manchester


Abstract: Solution of the agunah problem requires a clear understanding of the relations between history and authority in the halakhah. The argument of Rabbi Michael Broyde, that conditional marriages represent the most hopeful way forward, itself implies the need for consensus in addressing the technical problems involved. But what is the status of consensus in the halakhah? Historical theories are available to explain its emergence, but do they cast light on its present authority? Equally, we need to address the authority of halakhic statements which may themselves have been based on historical error, on such vital matters as the talmudic text relating to the moredet, the measures actually used by the Geonim to free her, and the authority by which they did so (see also Jackson, "Agunah and the Problem of Authority",, with the relevant Hebrew texts from which this is argued). All this points to a research agenda which I hope to implement in the Agunah Research Unit of the University of Manchester (, resources permitting.




First, may I thank you for your very kind invitation to participate in this conference and address you on the issue of agunah, and for your hospitality, especially that of the Greenbergs over Shabbat.

Much significant progress in alleviating the plight of agunot has been made in recent years, not least as a result of the efforts of Orthodox Jewish women in a number of jurisdictions. I take this opportunity to pay tribute to the memory of Judge Myrella Cohen, zikhronah livrakhah, who passed away just a couple of weeks ago, and who was a leader in the - recently successful - campaign to amend English divorce law by giving the judiciary a power comparable to that in the first New York get law. Myrella had been involved in this campaign for more than a decade, going back to the initiative of the late Chief Rabbi Jakobovits. She would not, however, have dissented from the judgment that such measures - even leaving aside their technical limitations and the argument about the distatefulness of reliance on secular law - are in the strictest sense "parochial": not all agunot live in New York state or in England; that legislative route has to be pursued separately in every jurisdiction where Jews live, and even once legislated - as the English history shows - it may not prove immune from change for quite external, secular reasons.

In the time available to me, I would like to outline to you a much more ambitious project, that of prevention, within the halakhah itself, of the agunah problem from arising at all. But let me stress at the outset: I do not come with any magic solution which could be put to the halakhic authorities and adopted overnight. Those who point to the immense work of generations of halakhic scholars, without achieving any such magic solution, are correct. Rather, I wish to advocate the need for a new round of substantial research, directed largely to addressing the underlying rather than the immediate issues. These underlying issues are largely issues of authority, and I believe that academic research has a major contribution to make, first and foremost in the very definition and clarification of the issues involved.

Some of these underlying issues involve the relationship between history and authority. We sometimes encounter the argument: major historical changes have been made in Jewish law (not least, in the areas of marriage and divorce) in the past; why, then, can we not make the necessary changes to meet the problem of agunah today? Against this, the halakhist will reply: it is not a question of history, but rather of authority. The mere fact that changes were made in the past (even assuming that those who made them had the authority to do so) does not in itself entail the view that we have authority to make such changes today. We have to accept as our starting point the orthodox premise that changes cannot be made today unless the authority exists to make such changes. The issue then become whether such authority exists.

At this point, however, the history of the matter re-enters the debate. The question of authority itself sometimes depends upon historical claims. How, then, are such historical claims to be judged - by the criteria of history, or of authority? Moreover, the authority system of the halakhah has its own history, and itself admits, in principle, of controlled change. If we are not entitled to argue: "just because changes have been effected in the past, the authority must exist to make further changes today", it must follow that we cannot argue either: "just because changes have been not been effected in the past, the authority cannot exist to make them today". I shall return to this in some detail later. But first, I wish to review an important recent analysis of the problem, that of Rabbi Michael Broyde, in the book he published last year: Marriage, Divorce and the Abandoned Wife in Jewish Law (Ktav, 2001). It too seeks to address the underlying problems, though Rabbi Broyde analyses them somewhat differently.


The Present State of the Debate: Rabbi Broyde's analysis

Rabbi Broyde distinguishes five models of exit from marriage, arising from or reflecting different conceptions of the nature of marriage, which have been normative in different communities at different times. Of these, one may distinguish particularly between (i) that of the Geonim and Maimonides which, with some differences between them, gave both parties to the marriage a unilateral right to terminate it (without proof of fault) and (ii) that of Rabbenu Gershom and Rabbenu Tam, which equalized the position between the spouses in a different way, namely by allowing neither party to the marriage a unilateral right to terminate it (without proof of fault). Rabbi Broyde maintains that it is the latter conception which is currently "normative" in most Orthodox (and certainly US) communities. Nevertheless, he appears to take the view that any individual couple may "opt in" to whichever model they like, even if that is not the model normative in their community, by the use of an appropriate clause in the ketubah. The ketubot found in the Cairo Genizah, incidentally, provide an historical precedent for such clauses.

Rabbi Broyde in fact distinguishes five such models: in addition to the above two, there are those of the Bible and the Talmud. But he also identifies a fifth model, which he calls "marital abode as the norm" but which sounds to me rather like the modern secular "irretrievable breakdown" ground, where the woman seeking a divorce separates from the husband and the Rabbinic court concludes that there is no prospect of their getting together again; in such a case, according to the Or Zaru'a, it is permissible to coerce the giving or receiving of a get. It is not clear from the modern authorities Rabbi Broyde cites whether the husband must agree that there is no such prospect. On the one hand, he quotes (142 n.14) Rav Henkin, who appears to require that both parties desire termination of the marriage (in which case it surely falls within the normal parameters of divorce by mutual consent); on the other hand, Rav Feinstein is quoted as saying:

In the matter of a man and a woman who, for these past years, has (sic) not had peace in the house. Since the beit din sees that it is impossible to make peace between them ... it is compelling that they should be divorced, and it is prohibited from either side to withhold a get, not the man to chain the woman to the marriage or the woman to chain the man to the marriage, and certainly not over financial matters (Iggrot Moshe, Yoreh Deah 4:15)

If the consent of the husband is not here required, as seems to be implied by Rav Feinstein, then really the only difference between this and the model of the Geonim and Maimonides is proof of separation and irretrievable absence of shalom bayit. How different is this from the amatla required by many in cases where a moredet claims ma'is alay? Indeed, we have here a good example of a restriction adopted by the halakhah being regarded not as absolute but rather interpreted in the light of its motives. The motive is the traditional fear of the halakhah that the woman is actually seeking an excuse because she is attracted to someone else - a fear which has an interesting historical origin, in my view, in early rabbinic abhorrence of the behaviour of some women of the Herodian dynasty recounted by Josephus (but that is a different story). Here, proof of separation and lack of shalom bayit rebuts that fear. I shall suggest that similar reasoning may be applicable to another restriction holding up progress in relation to our problem, that relating to the maintenance of conditions within a marriage where there is, for the moment, shalom bayit.

Of course, a lot depends in such cases on what measures of coercion the bet din is prepared to adopt, and how effective they may prove. I have to say that this - permissible coercion - does not appear to me to be the preferred route to a solution. We know that there are cases where a husband will resist even the fiercest coercion: witness the notorious case in Israel where the husband died in prison after 32 years of refusal to give a get. Rabbi Broyde regards this as representing "the basic success of the system, not its failure" (p.51). He argues (p.156 n.24) that just as the presence of some crime is not proof that the criminal justice system does not work, so too the presence of some agunot is not proof that the halakhic system does not work. The analogy fails, however, if one takes the view that the presence of any agunot (which Rabbi Broyde himself would prefer not to see) represents a failure in the very structure of the halakhic system, and that modifications in the halakhic system are capable of removing it. Despite the title of Rabbi Broyde's lecture here tomorrow morning (Unsolvable Agunah problems, Intractable Jewish Law, and the Real Limits of Halakhah: Why Some Agunah Problems Simply Cannot be Solved"), I think that the logical implications of his book are that modifications in the halakhic system are indeed capable of removing it. As I have already noted, he appears to take the view that any individual couple may "opt in" to whichever model of marriage (and its termination) they like, even if that is not the model normative in their community, by the use of an appropriate tnai in the ketubah.

Suppose, then, that the model of marriage which a couple decide to opt into is one which provides a unilateral exit to either party, with or without proof of separation and irretrievable breakdown ofshalom bayit? Suppose further that the clause they adopt in their ketubah goes so far as to say that the marriage is conditional upon either party not refusing either to give or receive a get? Rabbi Broyde appears to accept that even such a tnai is in principle halakhically valid (a tnai, he observes, "works for almost every imaginable contingency", p.95) and that its breach results in the termination of the marriage without the need for a get; indeed, he argues even that this is not retrospective termination, so that no doubt is thereby cast upon the (social) status of the children (who would not, in any event, have been mamzerim even if the termination of the marriage had been by retrospective annulment). However, Rabbi Broyde identifies the following practical problem, as a result of which such a clause is quite outside current halakhic practice: "normative halakhah assumes that people forgive tenaim after the couple commence a sexual relationship, and thus the marriage is valid even if the subsequent conditions are breached." It is, however, possible, he argues, to keep conditions in force despite the sexual relationship: the halakhic assumption regarding their intentions is at best a presumption. How, then, may it be rebutted? Rabbi Broyde observes that the procedural requirements for this are very onerous, and that is why such conditions are not used (p.96):

Certainly, all agree that a tenai can be kept in effect if, for example, the couple repeated the condition to a bet din each time before they engage in a sexual relationship.

He states this as an example of how the problem might be overcome. Yet in other types of agunah situation, Rabbi Broyde himself celebrates the fact that the problem has been largely resolved through modern technology, particularly relating to communications. Is it not conceivable that personal appearance before a bet din could be replaced, for example, by a secure e-mail message, buttressed perhaps by a statement in the tenai itself that it is presumed to remain in force despite sexual relations, if the latter are preceded by such a message Ö??

For Rabbi Broyde, the problem of the agunah is only marginally one of such "mechanisms"; rather it is a combination of substantive law (when should divorce be permissible? ñ though he appears willing to leave a considerable leeway on this to the parties themselves) and the authority system. To an extent, I agree, but I differ in my view of the nature of the problem relating to the authority system. For Rabbi Broyde, it consists in the lack of practical authority, particularly in the United States, in the rabbinic court system: he notes, for example, the extensive communal and religious sanctions (theharkhakot deRabbenu Tam: see pp. 152f. n.11) advocated by Rabbenu Tam in order to persuade the recalcitrant husband. These amount to excommunication; he notes that such sanctions are indeed applied by the rabbinical courts in the State of Israel - according to one source he cites, even in cases of moredet to'enet maus alay with amatla me'uveret. He laments the fact that any comparable orders issued by a rabbinical court in the US could simply be ignored. Yet all this assumes that permissible coercion is indeed the way, or one of the ways, forward. In my view, we should never reach the stage of having to coerce the husband.

Rabbi Broyde's own formulation of the position relating to tenain in fact reflects what in my view is the true problem of authority: he notes that " all agree that a tenai can be kept in effect ..." if certain procedures are followed. Frequently, we hear that a proposed solution fails because "all" do not agree: in other words, that it lacks a consensus. That suggests that the problem with which we are faced is primarily one not of judicial authority but rather of "legislative" authority, in the sense that a consensus and not merely a majority of contemporary poskim is required. I turn now to the analysis of such problems of authority.


Problems of Authority

Consensus is one of a series of issues which, in my view, have to be addressed before there can be real progress on the agunah issue (and indeed many others in contemporary Jewish law). We need a deep understanding of the history and contemporary authority of consensus in the halakhic system. Consensus was not required for halakhic decision-making according to the Talmud: ahare rabbim lehatot, majority decision-making, was the rule. Some have suggested that consensus entered Jewish law in the Middle Ages, under the influence of the Islamic ijma, one of the four "pillars" of Islamic law. But matters are rarely that simple. We do find discussion amongst the Rishonim, arguing from talmudic texts, of whether unanimity is required for certain types of communal enactment (derived from the authority of the bnei ha'ir in the Talmud), particularly where confiscation of property was concerned. This seems to have spread to other types of communal enactment where the approval of a sage, apparently fulfilling the role of an adam hashuv, in the Talmud was required.

In the 14th cent, Ribash (R. Isaac b. Sheshet Perfet) was asked (Resp. #399) about the validity of a proposed communal enactment which provided:

... that no one may marry any woman except with the knowledge and in the presence of the communal officials, and in the presence of ten persons; and that if anyone should violate the law and marry contrary to these requirements, the marriage is void (nifka'in). At the time a marriage is contracted [in violation of the enactment], the community expropriates the money or other property given to effect the marriage, and the property is considered to be ownerless and of no value. The marriage is annulled (mufka'in), and the woman may marry without any divorce (mibli shum get) and is not even required to obtain a divorce to remove any possible doubt.

Ribash marshalls the relevant talmudic sources, and even argues that the principle of kol hamekadesh extends to the power of a kahal to legislate for marriage. He concludes unequivocally that the community does have the power to adopt the proposed takanah - takanah, note, authorising the use of a tenai which, if broken, would result in annulment of the marriage. That being so, the final paragraph of the teshuvah comes as a surprise:

This is my opinion on this matter in theory. However, as to its practical application I tend to view the matter strictly; and I would not rely on my own opinion, in view of the seriousness of declaring that she needs no divorce to be free [to marry], unless all the halakhic authorities of the region concurred, so that only a "chip of the beam" [cf. Sanh. 7b] should reach me [i e., so that I do not take upon myself the full responsibility, but only part of it].

Ribash is not willing to bear the responsibility for this decision alone; he requires the concurrence of "all the halakhic authorities of the region". In short, whatever unanimity rule may have applied to the members of the kahal in enacting the takanah is now applied also to the adam hashuv from whom a haskamah is sought. But what is the status of such a demand for consensus? Ribash himself puts it not in normative, but rather in sociological or even psychological terms: I do not wish to take sole responsibility for this. Professor Menachem Elon observes (Jewish Law, II.856) that this reflects a desire "to divide the responsibility for the decision among as many authorities as possible"; perhaps we should say, rather, that it reflects a desire to divide the responsibility for authorisation of the decision among as many authorities as possible. But we still have to examine the process by which such a sociological or even psychological requirement has become - if indeed it has become - a normative requirement, without which no comparable takanot could be enacted today.

Two other issues involving the relationship between history and authority also call for further analysis. One relates to the text of the Talmud itself. If the Talmud ranks as the highest authority, we clearly need to pay careful attention to the text of any relevant talmudic passage. One talmudic text vital to our question, that of the moredet in Ketubot 63b, has a problematic text. According to the traditional text:

... if she says, however, "He is repulsive to me (ma'is alay)," [Amemar said] she is not forced (lo kayfinin lah). Mar Zutra said: She is forced (kayfinin lah).

According to this, the traditional text, the issue between Amemar and Mar Zutra is whether the wife is to be compelled back (into marital compliance). Mar Zutra takes the view that she is; Amemar takes the view that she is not. However, recent work towards a critical edition of the Talmud text has revealed a significant variant. MS Leningrad Phirkovitch reads:

... if she says, however, "He is repulsive to me," [Amemar said] he is forced (kayfinin ley). Mar Zutra said: She is forced (kayfinin lah).

Here, Amemar takes the view that it is the husband who is coerced, which can hardly mean anything other than that he is coerced to give her a get. The issue raised by the variant text of Amemar's opinion is of great importance for the later development of the halakhah. It may cast light on whether the final position of the Talmud, that of Rabbanan Sabora'i, not only entitled the moredet to a getafter a twelve-month waiting period, but also admitted of coercion. The Geonim accepted and developed the institution of compulsion against the husband of a moredet, but their view was ultimately rejected by Rabbenu Tam. For Rabbenu Tam, the Geonim had no authority to go beyond the Talmud, and the Talmud referred to coercion, in the case of the moredet, only in respect of the wife, not in respect of the husband. But Rabbenu Tam did appears not to have had access to this variant MS tradition. Suppose that scholarship ultimately concludes that the variant represents the original text, so that the Talmud does (in the opinion of Amemar, at least) contemplate coercion of the husband? Supposing also that such a text was available to the Geonim, and formed the basis of their takanta demetivta? Would such an historical conclusion be taken into account by halakhic authority?

A recent study of this problem by Rabbi Moshe Bleich (Tradition 27/2 (1993), 22-55) cites the view of Rabbi S.Y. Zevin, the editor of the modern volume of variae lectiones, that:

a variant talmudic text is significant only when it can be demonstrated that an early-day authority based his ruling upon that version of the text.

But should that apply even when manuscripts become available which were not available at all to the earlier authorities? Is the situation not comparable to the principle of hilkheta kebatra'i, where account is taken of the fact that the new argument could not have been known to the earlier authorities? Rabbi Bleich, however, concludes:

for halakhic purposes, it is the consensus of contemporary authorities that inordinate weight not be given to newly published material. Even earlier authorities who gave a relatively high degree of credence to newly discovered manuscripts did so within a limited context. Accordingly, formulation of novel halakhic positions and adjudication of halakhic disputes on the basis of such sources can be undertaken only with extreme caution.

Again, we see the crucial role played by consensus, not only in the substance of the law but also in underlying issues relating to the authority system itself.

My final example of an area where issues of the relationship of history and authority arise concerns the measures actually taken by the Geonim in the takanta demetivta. Rav Sherira Gaon wrote that in the case of a moredet, "we compel him (kofin oto) to grant her a divorce forthwith". Kofin normally refers to physical coercion: thus, the husband is coerced (beaten) into writing (or authorising the writing), and delivery of the get. On this formulation there is no suggestion that the court itself takes over any of the required formalities. Yet there are hints of the use of a greater judicial power in some Geonic and later sources. According to the Halakhot Gedolot (ascribed to Rav Shimon Kiara, 9th cent.): "... we grant (veheyvinin lah) her a bill of divorce immediately". Similarly, Rav Shmuel ben Ali, Head of a Babylonian school in the second half of the twelfth century, writes that "they grant her (notnin lah) an immediate divorce" The use of the plural in these sources (veheyvinin, notnin), suggesting that the get is here effected by an act of the court rather than the husband, becomes more explicit still in an anonymous 13th-cent. responsum, which uses the expression: "they wrote (vekotvey lah) her an immediate bill of divorce". Such a view would seem to be implicit also in a number of clauses of ketubot from the Genizah. Indeed, the Rosh (R. Asher b Yehiel (Asheri), Germany, 1250-1328), despite following Rabbenu Tam on the general issue of the takanta demetivta, appears to have interpreted the Geonic practice not as coercion but rather as annulment (hafka'at kiddushin):

... For they relied on this dictum: "Everyone who marries, marries in accordance with the will of the Rabbis" [bKet 3a], and they agreed to annul the marriage (lehafki'a kakiddushin) when a woman rebels against her husband (Resp. 43:8, p.40b)

There, is, however, no necessary incompatibility in these various positions: they could be taken as steps which have to be taken in sequence - leading ultimately, but only as a last resort, to annulment. Indeed, there is support for such a progression in other responsa.

The Ashkenazi rejection of the takanta demetivta ultimately prevailed, despite the fact that Rabbenu Tam may well have represented a minority position in his day. His arguments, however, were based upon a series of historical claims or assumptions, not all of which may turn out to have been sound. We need to ask, inter alia:

(a) What was the original text of Amemar's ruling on the wife proclaiming ma'is alay in the Talmud?

(b) What did the Geonim mean (and practice) by compulsion? Were they willing, in the final resort, to override the husband's resistance, whether by having the court authorise the writing and delivery of the get, or by hafka'at kiddushin?

(c) By what authority did the Geonim proceed: was it by interpretation of the Talmud (or a different talmudic textual tradition), takanah, custom, tsorekh hasha'ah?

Moreover, we also have reason to ask:

(d) Did the Rishonim have accurate information as to what the Geonim did and on what authority they based themselves?

(e) Do we have accurate information on the reasoning of Rabbenu Tam?

On all these (and several other such) questions, we may ask whether the authority of the tradition is affected by what may turn out to have been historical errors concerning its prior development. Thus Rav Shlomo Riskin argues (Women and Jewish Divorce, Ktav, 1989, p.86):

If it was the Geonim who initially provided for a coerced divorce, then if the Geonic decrees are ever rejected, their provision for a coerced divorce must be rejected as well. If, however, it was the Rabbanan Sabborai - i.e., the Talmud itself - who provided for a coerced divorce, then even if we were to reject the Geonic decrees granting the wife monetary compensation, we would nevertheless be forced to uphold the provision for a coerced divorce. Such is the position of Alfasi.

Or do we take the view that, like an erroneous textual tradition, error may be validated by subsequent acceptance? Not necessarily. In discussing hilkheta kebatra'i, Professor Elon (Jewish Law, I.271) quotes Rema:

In all cases where the views of the earlier authorities are recorded and are well known and the later authorities disagree with them - as sometimes was the case with the later authorities who disagreed with the geonim - we follow the view of the later, as from the time of Abbaye and Rava the law is accepted according to the later authority. However, if a responsum by a gaon is found that had not been previously published, and there are other [later] decisions that disagree with it, we need not follow the view of the later authorities (aharonim), as it is possible that they did not know the view of the gaon, and if they had known it they would have decided the other way.



In a lecture I delivered in London 18 months ago, and whose unannotated text is already available on the internet ( I argued for a particular combination of "mechanisms" and of bases for the authority to adopt them. Shortly, I shall publish a much expanded and annotated version of that argument. I put it forward, however, not as a fully researched proposal, but rather as the basis of a research agenda. The difficulties are not to be underestimated. The late Chief Rabbi Yitshak Herzog himself proposed, at the foundation of the State of Israel, atakanah in conjunction with explicit terms in the ketubah in order to provide equal succession rights for women. He failed to persuade the halakhic authorities of the day. This suggests that more fundamental research into the underlying obstacles to progress remains needed. This is the reason why I am seeking to establish an Agunah Research Unit at Manchester ( I hope that it will receive the support, moral and material, of the community.