Tokh Kedei Dibbur in Eidim Zomemim

  • Rav Moshe Taragin

The gemara in Ketubot addresses the issur of eid zommem (a special type of false witness) in relation to other issurim.  To highlight the singularity of eid zommem, the gemara searches for a point of distinction between it and other issurim.  For example, the gemara in Ketubot (33a) notes that an eid zommem is punished even though he wasn't warned prior to committing the sin.  Generally, one only receives an onesh for a sin committed subsequent to hatra'a - ample warning.  Otherwise "ein onshin ela im kein mazhirin" - we cannot administer punishment unless a warning was issued. 


            Eid zommem, however, is an exception.  The gemara suggests that the lack of warning for an eid zommem is not a fundamental issue but a technical one.  There is no practical way of warning eidim since issuing a warning to each witness prior to testimony might deter them from giving testimony.  Moreover, the potential cost of warning all eidim not to issue false testimony rules out issuing prior warnings.  Furthermore, once their testimony has been offered a warning is meaningless.  Hence, we can conclude that eid zommem is qualitatively - on a technical level - different from other aveirot.


            Subsequently, the gemara questions the above stated logic.  Why not, the gemara asks, alert the eidim immediately AFTER they finish their testimony.  Halakha recognizes the immediate few seconds after an action as 'relevant' to that action.  This is known as the halakha of 'tokh kedei dibbur'.  "Kedei dibbur" refers to the amount of time it would take to greet a rebbi with the greeting "Shalom alekha rebbi u-mori."  According to the gemara these few seconds still allow a reversal of the original action.  For example, if someone executes a kinyan and within a few seconds withdraws his intent, the kinyan is voided.  Similarly, if someone articulates a 'neder' and within a few seconds invalidates it, the neder is nullified.  By similar logic, the gemara suggests that we should warn the eidim within a few seconds of their testimony.


            Though the gemara asserts this possibility and at no point rejects or even examines this option, from a purely logical standpoint the option remains highly problematic.  The gemara in Sanhedrin (8b, 72b) describes the purpose of hatra'a as "le-havchin bein shogeg le-meizid" - to distinguish between an intentional sinner and an unintentional one.  Essentially, anyone who has not been warned can always claim ignorance.  Since we cannot prove otherwise we exonerate all those who have not received warning.  Once warned, however, the action was obviously performed intentionally and warrants a penalty.  Given this view of the role of hatra'a it is difficult to accept the possibility of a warning which FOLLOWS an aveira - even if it follows immediately.  If, at the moment the aveira was committed the perpetrator was not warned, he may be assumed to be an unintentional culprit and cannot be punished.  What difference does it make if he received a warning immediately after the aveira?  Even a warning issued within kedei dibbur of an action should have no meaning!!  How can the gemara in Ketubot sanction such warnings?


            In order to resolve this issue, we must change one of two perceptions.  We have just isolated a problem in applying the rules of kedei dibbur to hatra'a.  To reconcile this problem we must alter either our conception of kedei dibbur or our view of hatra'a and its function - possibly both.  This article will examine how these concepts must be adapted to enable the position in Ketubot which validates a warning tokh kedei dibbur.  It will not represent the only method of understanding these concepts.  It will merely isolate certain perceptions which must be maintained to arrive at the gemara's conclusion.


            As previously stated, the common approach to the halakha of tokh kedei dibbur is that it allows a person a 'window' of a few moments to reconsider an event or statement and possibly retract it.  This indeed is how Rashi in Ketubot (33a) describes this halakha.  Given this view of kedei dibbur our question remains: regardless of his ABILITY to retract the eidut, the warning did not coincide with action, but succeeded it!! 


            The gemara in Nedarim (87), however, suggests an alternate version of kedei dibbur.  The gemara discusses the mitzva of tearing garments upon hearing of the loss of a close relative.  According to the gemara the keri'a must be performed with accurate knowledge.  If, by chance, someone tore his clothing thinking that his brother had expired only to discover that it was really his parent, the keri'a, based on incorrect knowledge is rendered invalid and must be repeated.  However, if he discovers the true identity of the departed within a kedei dibbur of his keri'a, the act is valid.  Evidently, the gemara considers kedei dibbur more than just a window of opportunity to retract statements or recall actions.  No actions are being retracted in this instance!!


            This gemara is apparently attempting to redefine the concept of simultaneity, addressing a very basic and fundamental question:  Within what block of time can events be considered as having occurred simultaneously?  According to the gemara events which occur within a kedei dibbur of each other, even though one actually preceded the other, are considered simultaneous. 


            This should not be too surprising.  All measurements of time are subjective and based upon convention.  We assign the label 'second' to a particular unit of time but that certainly doesn't mean that smaller units do not exist; we might be incapable or unwilling to measure those units but they exist!!  Hence, even events which occur during the same second are not REALLY simultaneous; convention merely allows us to CONSIDER them as such (i efshar letzamtzeim).  By analogy, within the halakhic system the block of time known as "kedei dibbur" is defined as an 'elementary' or 'atomic' unit.  Events which occur within this block are categorized as simultaneous - even if to the naked eye they appear sequential.  In the gemara's instance the discovery of the identity of the deceased within a "kedei dibbur" defines the knowledge as simultaneous to the keri'a.  The Yad Rama in Bava Batra (129b) reiterates this concept - within a kedei dibbur the initial intention is revoked.  By redefining kedei dibbur in this fashion we are able to begin to reconcile the appearance of this concept in Ketubot (33a).  If the warning was issued within a few seconds of the eidut it is considered - in the eyes of halakha - as having coincided with the eidut, and an onesh may be administered.




            Assuming tokh kedei dibbur merely allows retraction we might not justify issuing a warning tokh kedei dibbur of the eidut.  If, however, this unit of time defines simultaneity, a hatra'a issued within this block is viewed as contemporary to the actual aveira.


            A problem, however, remains.  Even if from a formal standpoint halakha considers the subsequent warning as simultaneous to the actual aveira, from an empirical standpoint - in terms of real experience - the warning was not issued prior to the event.  If we are to adopt the gemara's position that hatra'a's central role is to inform the person of his sin and define him as meizid, what concerns us isn't the formal timing but the actual!!  If he wasn't actually forewarned he may still claim ignorance.  Justifying this position in the gemara Ketubot requires more than a modification of the concept of tokh kedei dibbur.  We must also inspect the role of hatra'a and detect another possible function - one which might be more formal in nature.


            Though the gemara in Sanhedrin several times designates hatra'a as that which distinguishes between meizid and shogeg, one gemara apparently adopts a different view.  The gemara (41a) concludes that an onesh cannot be imposed unless a warning was given and the subject "yatir atzmo le-mita" (he forfeited his life or he surrendered himself to the punishment in question).  In practical terms this requires the subject to hear the hatra'a and actually defy the warning by declaring "I still intend to carry out the crime."  This gemara establishes a second or alternate role for hatra'a - to intensify the austerity and severity of the aveira making it so grave as to warrant an onesh.  This goes far beyond distinguishing between ignorance and knowledge.  This issue addresses the nature of the action - was it accompanied by onesh or not?  The Rambam cites this halakha (Hilkhot Sanhedrin 12;2) and concludes that for this reason even a talmid chacham who cannot claim ignorance requires a warning.  A talmid chacham even without being warned cannot excuse himself due to lack of knowledge; his action, though, is not serious enough to warrant onesh unless accompanied by - and in defiance of - hatra'a.


            This second dimension of hatra'a seems more flexible. Indeed one might still insist that in order to be considered in contempt of a warning, the event must be subsequent to that warning and in defiance of it.  Alternatively, one might claim that as long as halakhically the two events are simultaneous the action is considered severe enough to warrant penalty.  His defiance consists in his not recanting the testimony after hearing the hatra'a.  If we isolate this second function of hatra'a (assuming this were the only function of hatra'a) we might afford greater time flexibility and allow the warning to be issued afterwards.  As long as it was issued immediately after, within kedei dibbur, it is formally considered simultaneous and the action is severe enough to justify a penalty.




1.  Understanding the interaction between two halakhot demands a thorough grasp of each particular halakha.  Whenever one faces a gemara which discusses how 'x' applies to 'y' both 'x' and 'y' must be examined.

2.  Distinctions between formal and practical halakhot are crucial.  What tokh kedei dibbur accomplishes on a formal plane might have little relevance to the actual experience.  Is hatra'a part of that experience or is it too a formal requirement?