4b - Gilgul Shevu'a

  • Rav Yair Kahn

Translated by David Silverberg




1. ד: "מתיב מר זוטרא ... דלאו הילך הוא", שבועות לח: "טענו כלים ... לישבע עליהן", קדושין כז: "וזוקקים ... קמ"ל".

2. ריטב"א ה. "א"ל ג' פרות ... ואין צורך", גמ' ו. "דאמר רב ששת ... שאינה ברשותי", תוספות ד"ה 'שבועה', תוס' הרא"ש ד"ה 'שלא', תוס' ב"מ צז. ד"ה 'ביום' מהמילים "ועוי"ל ...".

3. שבועות מט. "לכל מגלגלין ... לפתוח לו", תוספות ד"ה 'לכל, ריטב"א ד"ה 'ואמרי' מאי בינייהו', חידושי הר"ן ב"מ ד. ד"ה 'תיתי מגלגול שבועה', ריטב"א קדושין כז: ד"ה 'אמר עולא'.

4. ירושלמי שבועות פרק ז' סוף הלכה א "מהו לגלגל ...".

5. חידושי הר"ן צח. ד"ה 'משכחת לה', 'ומ"מ קשה'.

6. חידושים המיוחסים לריטב"א לד. "אמר רב הונא ... כל זה לשון רבי' יצ"ו", צח. ד"ה 'והשאר'.


1. On what point do Rav Huna and Rav Chisda disagree in Shevu'ot 49a?

2. Why is the oath taken by a laborer different than other oaths with respect to gilgul shevu'a?

3. Is gilgul shevu'a by Torah law? Is it subject to the law that "since he cannot take the oath, he must pay"?

4. Which additional claims can the defendant not be required to incorporate into his oath by way of gilgul?




I. The Source of Gilgul Shevu'a

            According to the Gemara's conclusion (4b), the mishna in Shavuot comes to teach us the concept of "gilgul shevu'a". The Gemara in Masekhet Kiddushin (27b) establishes the Torah origin of the gilgul shevu'a provision, and derives it from the oath taken by the sota:

"Ula said: From where do we know the Biblical origin of gilgul shevu'a?  For it says, 'the [sota] woman responds: Amen, amen.  And it says in the mishna: With regard to what does she say, 'Amen, amen'?  Amen on the curse, and amen on the oath; amen not with this man, amen not from another man, amen I have not sinned as a betrothed woman, married woman, waiting for yibum, after yibum… We find [the concept of gilgul shevu'a in the context of] sota, which involves issur; from where do we know [that we apply the law of gilgul shevu'a] in the context of monetary issues?  The house of Rabbi Yishmael taught [that this evolves from] a kal va-chomer – if for a sota, regarding whom proceedings cannot be initiated based on the testimony of one witness, we apply gilgul, then certainly in monetary issues, where one can initiate proceedings based on the testimony of a single witness, we apply gilgul."

However, this law of gilgul shevu'a seems very difficult to comprehend.  After all, the Torah is very "stingy" when it comes to oaths, and Torah law requires a shevu'a in only three cases: a claim substantiated by only one witness ("eid echad"); when the litigant confesses to part of a claim ("modeh be-miktzat"); and when a watchman cannot produce the item under his charge ("shevu'at shomrim").  But suddenly, when the defendant must indeed take an oath, the plaintiff can force him to swear over all types of other claims he brings against him, and all these oaths have the status of a "shevu'a de-orayta."  We must therefore explain how exactly the concept of "gilgul shevu'a" works.

            A previous Gemara (4a) introduces gilgul shevu'a within the context of Rabbi Chiya's halakha (that testimony of two witnesses on part of a claim, obligates an oath on the remainder): "Rav Papa said: It evolves from the gilgul shevu'a in a situation of a single witness."  But the Gemara rejects this inference: "Gilgul shevu'a in a case of a single witness is different, because one shevu'a brings with it another shevu'a; will you apply this to [two] witnesses, who obligate money [rather than a shevu'a]?"  As we saw in an earlier shiur (#9), the Rishonim argue as to whether this refutation stands according to the Gemara's conclusion, or if in the end we indeed derive Rabbi Chiya's halakha from a tzad hashaveh that includes the halakha of gilgul shevu'a in a case of eid echad.  In order to understand this debate, let us consider the Gemara's comment in Masekhet Shavuot (49a).

II. Who Imposes a Gilgul Shevu'a

According to a Gemara at the end of Shavuot (49a), Rav Huna and Rav Chisda argue whether beit din will initiate gilgul shevua, if the plaintiff did not call for its implementation.  The Amoraim argue as to who in essence casts the secondary oath upon the defendant: Beit-Din, or the plaintiff. It is likely that Rav Huna and Rav Chisda argue with regard to the fundamental underpinnings of gilgul shevu'a.  If Beit-Din initiates the gilgul, then presumably gilgul shevu'a stems from the fact that once this defendant is already obligated to take an oath, we have reasonable doubt concerning the other claims against him.  It is this objective doubt that obligates the defendant to take the secondary shevu'a, and Beit-Din therefore bears the responsibility to obligate the additional oaths just as they must force upon the defendant the original oath required by the Torah.  By contrast, according to the position that Beit-Din does not initiate the gilgul shevu'a, it would seem that despite their obligation to force the defendant to take the original oath, we have no reason for suspicion regarding other claims that have nothing to do with the case at hand.  Beit-Din therefore does not involve itself in other claims against the defendant.  Accordingly, gilgul shevu'a exists as an option granted to the plaintiff to include all types of other claims into the oath taken by his disputant.

III. What Claims Can Be Imposed Through Gilgul

            Now the Rishonim are in disagreement as to which claims are eligible for gilgul shevu'a.  The Gemara (6a) records the comment of Rav Sheshet, that a shomer must take three oaths: "I swear I was not negligent with it"; "I swear I did not use it for myself"; "I swear it is not in my possession."  Tosefot there explain: "Or, perhaps he swears 'I did not use it for myself' only as a result of a gilgul from the oath that 'I was not negligent with it'."  They proceed to cite the position of: "But Rabbenu Tam explains that he cannot swear [that he did not use it for himself] through gilgul if [the animal] is dead before us, for we apply gilgul only regarding something likely [when we have grounds for suspicion]."  The Tosefot Ha-Rosh explains more fully:

"It does not appear reasonable to obligate him [to swear] through gilgul, even though we normally apply gilgul even regarding a doubt … For this is only for something reasonable and common, such as partners, sharecroppers or guardians who might excuse themselves [when stealing from the other].  But someone else, from whom he never stole anything, cannot cast upon him a different oath out of doubt.  And here, too, he cannot cast upon him an oath that he did not use it for himself out of doubt.  For why should we suspect that he used it for himself, given that the reason of 'he excuses himself' is not relevant here?"

According to Rabbenu Tam, we can cast upon the defendant an additional oath only with regard to a likely claim.  Those who argue with Rabbenu Tam, however, allow the plaintiff to force the defendant to swear on any claim he presents, regardless of their likelihood.  (See also Tosefot in Kiddushin 28a, s.v. ne-emra.)  If we perceive gilgul shevu'a as stemming from our suspicion of the defendant, then we would likely accept Rabbenu Tam's position, limiting gilgul to probable claims.  If, however, gilgul is an available option for the plaintiff, to include all his claims against the defendant in the current oath, then we need not, necessarily, distinguish between different types of claims, and we may subject them all to giglul shevu'a.

            A different debate among the Rishonim concerns the principle "mi-tokh she-eino yakhol li-shava meshalem," that a litigant obligated in an oath who is not in a position to swear on the given issue must pay instead.  Does this rule apply to oaths obligated through gilgul shevu'a, as well?  Tosefot later in Bava Metzia (97a s.v. ba-yom) maintain that even to such oaths we apply the rule of "mi-tokh she-eino yakhol li-shava meshalem."  The Ran, however, brings the ruling of the Ra'avad, that one who cannot take an oath obligated through gilgul need not render payment in lieu of the shevu'a.  It would seem that if the obligation of the original oath leads to an additional shevu'a de-orayta because of suspicion, then if the litigant cannot take this oath, he must pay instead, just like any other shevu'a mandated by the Torah.  If, however, we have no suspicion, rather the plaintiff enjoys the right to introduce into the oath other claims, even when unsubstantiated, then presumably we would not apply the principle of "mi-tokh she-eino yakhol li-shava meshalem."

IV. The Uniqueness of Sakhir

            Earlier, we encountered the Gemara in Shavuot which asserts that we do not apply gilgul shevu'a in a situation of a sakhir, a worker whose employer claims to have already paid his fee.  The mishna had ruled that in such a situation, the worker may take an oath and then receive his payment.  The Gemara establishes that this oath taken by the employee is not subject to gilgul shevu'a.  Why do we not apply gilgul shevu'a in such a case?  Tosefot bring the Ritzba's position that this halakha applies not only in the case of sakhir, but rather in all cases of "nishba'in ve-notelin" – situations of shevu'ot taken before collecting payment.  Tosefot write, "Since the oath is for their own benefit, we do not cast upon them [other oaths]."  The explanation would seem to be that since this type of shevu'a works to the benefit of the plaintiff, the defendant does not have the standard status of a "mashbia" – one who imposes an oath, and therefore he does not have the power to add other claims.  Similarly, the Yerushalmi in Shavuot (chapter 7, end of halakha 1) raises the possibility of casting a shevu'at nishba ve-notel upon someone required to take a Torah-mandated shevu'a, through gilgul, but quickly dismisses this possibility.  It says, "There – he swears and does not pay, whereas here, he swears and collects."  According to the Yerushalmi, then, we cannot demand a shevu'at nishba ve-notel through gilgul shevu'a from a shevu'a taken to exempt oneself.  However, the Yerushalmi implies that this is because of the incompatibility between the two oaths.  This requirement of parity between the original oath and those added through gilgul becomes easily understood if we assume that it is the litigant who includes additional claims to the basic shevu'a.  If, however, gilgul shevu'a stems from our suspicion, which would then require a separate, independent oath on other claims, the parity requirement is far less convincing.  Tosefot later cite the Ri's explanation as to why we do not apply gilgul in the case of the sakhir: "It mentions specifically sakhir, because his oath was instituted only to appease [the employer], and they therefore did not obligate him to such an extent that he is subjected to gilgul."  This explanation works well with the approach viewing gilgul as evolving from the reasonable doubt concerning this litigant.  It therefore does not apply in a case of sakhir, where the oath is required only to appease the employer, and it does not reflect any suspicion regarding the position of the sakhir.

            It is reasonable to assume that the possibility of deriving Rabbi Chiya's halakha (an oath resulting from testimony on part of a claim) from gilgul shevu'a in a case of eid echad, depends on this very point.  If gilgul shevu'a stems from the plaintiff's right to include additional claims in the oath, then we cannot build upon gilgul shevu'a in establishing Rabbi Chiya's halakha.  After all, the testimony of two witnesses yields an obligation of money, not an oath; therefore, there is no basic oath into which we can introduce additional claims.  Indeed, we may explain the Gemara's refutation along these lines: "Gilgul shevu'a in a case of a single witness is different, because one shevu'a brings with it another shevu'a'; will you apply this to [two] witnesses, who obligate money [rather than a shevu'a]?"  If, however, gilgul shevu'a results from an objective doubt regarding the defendant, we might not need specifically an original oath in order to activate gilgul to require additional oaths.  To the contrary, what triggers the gilgul shevu'a is not the original oath, but rather the single witness who testified against the defendant and thus gave rise to doubts concerning other outstanding claims against him.  And if one witness can call the defendant's trustworthiness into question, then certainly two witnesses can do the same.

            The Ran, in his chiddushim, comments on this Gemara:

"From here Tosefot deduce that we can apply gilgul shevu'a even if the litigant does not demand it, and we initiate it for him.  For here we compare gilgul shevu'a to witnesses, and regarding witnesses we say that he swears on the remaining sum – meaning, Beit-Din has him swear, even if his fellow litigant does not demand it, like the shevu'a of modeh be-miktzat."

The Ran thus proves from the equation drawn between two witnesses and gilgul shevu'a through an eid echad, that Beit-Din initiates gilgul shevu'a.  We would explain this on the basis of the fact that this equation stems from the understanding that gilgul shevu'a results from objective suspicion, and it therefore stands to reason that Beit-Din initiates the additional oaths, as we explained.

            The Ran then adds that on this point Rav Huna and Rav Chisda are in agreement; according to both views, Beit-Din initiates the process of gilgul shevu'a.  He proves his position from the case of sota, where it is the kohen who adds more oaths onto the basic shevu'a taken by the woman, despite the fact that the kohen parallels Beit-Din, not the litigant.  The Ran writes:

"We may draw further proof to this point from the fact that in the first chapter of Kiddushin (27b), the basic rule of gilgul shevu'a is derived from the sota … This despite the fact that her husband demands of her only [an oath] regarding 'this man' who warned her and she nevertheless secluded herself [with another man].  And although it appears that Rav Huna and Rav Chisda argue on this point… whether or not they initiate gilgul, this refers only to a shevu'a de-rabbanan, with which we are dealing there.  Regarding, however, a shevu'a de-orayta, everyone agrees that they [Beit-Din] initiate gilgul."

The Ritva, however (in Kiddushin), refutes this proof and claims that in the sota procedure the kohen is in fact considered, with respect to the woman, the litigant, rather than the Beit-Din.  He writes:

"Sota features something that generally does not exist.  Regarding sota, the kohen himself initiates the gilgul oaths, whereas we generally hold that Beit-Din does not initiate gilgul on his behalf, unless the plaintiff demands it … according to Rav Chisda we initiate, whereas according to Rav Huna we do not initiate.  Halakha follows Rav Huna, but Rav Huna agrees that when it comes to sota, the kohen initiates, only sota is different because it is a case of issur [as opposed to a monetary issue], and all Israel are responsible for one another and required to distance [others] from violation.  They are therefore considered 'litigants.'  Here, however, where we deal with monetary matters, if the litigant does not make a claim, we should not turn ourselves into advocates and initiate on his behalf [the demand] to cast upon him [additional oaths]."

It is worth noting that the Rambam argues with both Ran and Ritva. In Hilkhot Sota (4:17), the Rambam indicates that gilgul shavua in the case of Sota is initiated by the husband.

V. Two Types of Gilgul

            Earlier we mentioned a dispute between Tosefot and the Ra'avad as to whether we apply the rule of "mi-tokh she-eino yakhol li-shava meshalem" to an oath brought about via gilgul.  We explained that this question hinges on the fundamental nature of gilgul shevu'a, whether it is due to suspicion regarding the defendant, or to the plaintiff's right to introduce other claims into the initial oath.  In his chiddushim (to 98b), the Ran proposes a middle position on this issue:

"And if you argue that to the contrary, we should say that since he is obligated in an oath through gilgul but cannot swear – since he claims he does not know – he should pay, for we have already proven earlier that even regarding gilgul we say 'mi-tokh she-eino yakol li-shava meshalem' – this is no argument.  For when do we say this – only regarding a gilgul that comes about through a definitive claim.  But with regard to a gilgul of a claim of uncertainty, although he must take an oath on it, if he says 'I don't know' he is not considered one who is obligated in an oath and who must pay if he cannot swear.  Rather, he swears that he does not know and is exempt, and to this the commentators z"l have agreed."

In other words, according to the Ran, we apply "mi-tokh she-eino yakhol li-shava meshalem" to a gilgul shevu'a only if the secondary claim is a definitive one.  When, however, the plaintiff adds onto the oath an uncertain claim, then although the defendant must take an oath because of the law of gilgul shevu'a, he need not pay if he cannot take the oath.  It seems to me that the Ran viewed the halakha of gilgul shevu'a as including both types.  When the plaintiff advances a clear, definitive claim, then the defendant bears the obligation to take a separate, independent oath due to the objective doubt concerning his reliability.  This oath, like all oaths required by the Torah, is subject to the rule of "mi-tokh she-eino yakhol li-shava meshalem."  If, however, the plaintiff adds onto the oath an uncertain claim, then we have no basis for suspicion, and we therefore do not require a new, independent oath.  Nevertheless, the plaintiff may still add this claim into the original oath; clearly, however, we would not apply "mi-tokh she-eino yakhol li-shava meshalem" on the second claim.

            Similarly, we saw that according to Rabbenu Tam, gilgul shevu'a applies only to likely claims, and we explained his view on the basis of the approach that gilgul stems from our suspicion of the defendant.  Tosefot (97, end of s.v. ba-yom) allow adding even uncertain claims through gilgul if they seem likely, just as we have the sota take additional oaths because she has already come under suspicion.  Claims that do not seem likely, however, can be added onto the shevu'a only if they are made with certainty.  This, too, can be seen in the laws of sota: just as we add onto the basic claim against the sota – which is made with uncertainty – other similarly uncertain claims, so may we add onto definitive monetary claims other similarly definitive claims – even if they do not seem likely.  It emerges from Tosefot that two different tracks exist within the halakha of gilgul shevu'a.  One track stems from our suspicion, and is therefore restricted to likely claims.  The second track results from a combination of additional claims and their inclusion within the original oath, for which we require a definitive claim, to integrate with the original claim.

            We should add that if we are dealing with suspicion concerning the defendant, one might distinguish between additional claims somehow related to the central claim, and those which are not.  One might argue that specifically regarding claims connected to the original case does an objective doubt arise concerning the defendant's status.  With regard to claims unrelated to the original claim, we have no reason to suspect the defendant and thus no basis to require an independent oath because of doubts concerning his trustworthiness.  Later in the Gemara (5a), we find that according to Rami Bar Chama, a shevu'at ha-shomrim is taken only in a case of a guardian charged with three items: one he admits owing, a second that he denies, and a third which he claims to have been lost due to excusable circumstances.  Some Rishonim understood that according to this view, shevu'at ha-shomrim is but a gilgul of shevu'at modeh be-miktzat.  The Chiddushim Ha-meyuchasim Le-Ritva (98a) adopts this approach but distinguishes between this instance of gilgul and gilgul shevu'a in general:

"One can answer that this law of Rami Bar Chama differs from general cases of gilgul.  For in cases of gilgul generally, if he says, 'Swear to me' we make him swear; otherwise, we do not.  But in Rami Bar Chama's case of gilgul, even if he does not say, 'Swear to me,' we make him swear, for this [oath] originates from the Torah.  And this is why the verse had to write shevu'at ha-shomrim, for we would not have deduced it from gilgul generally, and for this reason we say 'mi-tokh she-eino yakhol li-shava meshalem'."

In other words, in standard cases of gilgul shevu'a, only the plaintiff can cast the additional shevu'a on the defendant, and Beit-Din does not initiate the gilgul.  Likewise, in standard cases, if the defendant cannot take an oath forced upon him through gilgul, he need not pay.  In Rami Bar Chama's case of gilgul, however, the defendant must pay if he cannot take the oath. 

Earlier in the masekhet (34a), the Chiddushim Ha-meyuchasim Le-Ritva writes that two out of the three oaths taken by a shomer are a result of gilgul.  Here, too, Beit-Din initiates the gilgul process even though the plaintiff himself did not demand additional oaths.  This would indicate that in these situations of gilgul, where a strong connection exists between the different claims, gilgul operates as an independent shevu'a obligation resulting from the objective doubt that has arisen concerning guardianship.  Therefore, it is Beit-Din who initiates the gilgul process, and if the defendant cannot take an oath, he must pay.  In a standard situation of gilgul, however, where the claims are not related, although the plaintiff may, indeed, include additional claims into the oath, they are not added due to any doubt that has arisen.  Therefore, Beit-Din would not initiate such a gilgul, and if the defendant cannot take the oath, he has no obligation to pay.

            Finally, we should note that the provision of gilgul shevu'a constitutes a very big chiddush and drastic extension of the shevu'a obligation, particularly according to the approach that the plaintiff can bring into the oath all his claims against the defendant.  Even after our whole discussion and all the explanations described, the reasoning underlying this halakha remains mystifying.  Although we cannot conclusively determine the reasoning underlying the Torah's laws, we are nevertheless bidden to make an effort at understanding.  It seems to me that we can better understand the law of gilgul shevu'a in light of what we explained in shiur #4.  There we suggested that the oaths Beit-Din imposes upon a litigant ("shevu'ot ha-dayanim") resembles somewhat the oath of a sota.  When we reach an impasse on the legal human plane, we lift our eyes to the heavens for direction and a way out of the deadlock.  Not always is this road open to us.  Only in rare instances, when a serious doubt arises (namely, the husband's warning followed by the wife's seclusion with the suspected man) and we have not even a single witness, we are prepared to have the sota swear and to pay the heavy price of erasing the Almighty's Name.  But the moment we travel this road, we have the opportunity to clarify everything.  The halakha of gilgul shevu'a is derived from sota.  And the shevu'at ha-dayanim mandated by the Torah is also a very specific phenomenon limited to those situations where a suspicion arises but we have no witness to clarify the case for us on the legal plane.  Therefore, the moment we are prepared in any event to have a litigant swear, at the risk of a swearing in God's Name falsely, heaven forbid, his opponent has the opportunity to resolve all his doubts on the religious plane, through the law of gilgul shevua.


Sources for the next shiur:

  1. Tosafot, Bava Metzia 5a, s.v. she-ke-negdo.
  1. Shevu'ot 44b: "She-ke-negdo chashud… yachloku."
  1. Ibid. 47a: "Hayu sheneihem chashudim" … 47b: "chala al sheneihem."
  1. Ibid. 48b: "Ikla Rav Nachman …KeRebbi Elazar avad".
  1. Tosafot, Bava Metzia 97b s.v. h"n.



1. Why don't we apply the rule of "mitokh she'eino yachol lehishava meshalem" to the case of one disqualified from taking an oath (chashud)?

2. Tosafot suggest two answers to this question. Try to discern the basic difference between them?

3. Who argues on the rule of "mitokh?

4. What is the rationale for the rule of mitokh?