Daf 21a: Reformatting a Get after its Composition

The mishna on 19a permits the drafting of a get upon the hand of a servant or the horn of a cow, so long as the entire servant or cow is subsequently transferred.  For his part, Rabbi Yosei Ha-gelili invalidates these media for writing a get since he interprets the word "sefer" (Devarim 24:1, 3) literally: a get must resemble a formal document.  A text sketched onto a live being cannot be considered sefer.  By contrast, the Rabbanan understand the term sefer as referring to sefirat devarim, that the get must narrate the actual divorce.  Hence, they do not impose formal standards upon the choice of medium; they allow a get to be written upon a living organism, as long as the entire organism is subsequently transferred. 

            The ensuing gemara questions the need to transfer the entire animal.  In the case of writing a get upon a servant, it is obvious that we cannot sever the servant's hand to deliver the get; given no other choice, we must transfer the entire servant.  In the case of the animal's horn however, why not simply detach the horn from the animal and transfer it alone, without transferring the entire animal?  The gemara responds that Halakha does not allow any 'surgery' (ketzitza) to occur between the drafting of a get and its ultimate delivery.  As the gemara on 21b expounds "Ve-khatav… ve-natan" in those verses, a get is only valid if it is missing ONLY the writing and the immediate delivery.  If, however, a get is missing the writing, the formatting and the delivery, the get becomes invalid.  By writing a get upon an item (e.g., a horn) which one ultimately severs, one has retroactively written a get upon an item not IMMEDIATELY prepared for its delivery.  This shiur will explore the nature of the invalidation of cutting a get after it has already been composed.

            Most Rishonim claim that this gemara is not describing a problem pertaining to the drafting of a get upon something attached to the ground (see Tosafot, s.v. Yatza, quoting Rabbeinu Shmuel, for a notable exception).  The gemara on 21b allows a get to be written upon a plant in a perforated pot (atzitz nakuv), even though it is considered attached to the ground (since it receives its nutrients from the ground).  One may draft the get upon the pottery or even the plant, as long as the entire pot is transferred.  Even Rava, who invalidates drafting a get upon the leaves of this plant, is only concerned with the possibility that the person will subsequently detach the leaf and then transfer the get, thereby rendering the get invalid since the surgery occurs after the writing.  Rava would admit that if, after writing the get upon the leaf, the owner transferred the entire plant to his wife, the get would be valid.  This gemara seems to indicate that however we might understand the disqualification of tearing or reformatting a get after writing, it has little to do with the fact that the get was written on something attached to the ground.  In fact, the mishna itself equates writing a get upon the horn of an animal to writing it upon the hand of a servant, and it validates each instance - as long as the entire being is delivered.  Generally, Halakha considers servants as land ("avadim hukshu le-karkaot"); yet, one is allowed to write a get upon a servant!  Evidently a get may be written upon 'land' or items equated by Halakha thereto.  The problem with drafting a get upon the horn of an animal and subsequently detaching the horn must be explained in a different manner.  How can this halakha be understood?

            The Ramban deliberates between many opinions regarding this law, ultimately adopting the position of Rabbeinu Hai Gaon.  He claims that the disqualification of reformatting the document of a get only applies if the husband originally intended to detach the get; in the absence of this original intention, a get may be cut.  Said otherwise, according to the Ramban, there is no problem with the act of reformatting a get.  If, however, the document was earmarked for cutting, such a text is inherently invalid.  The get must be written upon a document which is immediately ready for delivery, without any delay.  By planning to reformat the document, one defines the get as "omeid le-ketzitza" and renders it unsuitable for immediate delivery.  This perspective accords well with the notion established in shiur #2 that, as opposed to the case of standard documents, in the instance of get the actual writing is deemed part of the formal divorce process.  Rabbi Eli'ezer requires that the get be written lishmah - with a designation to the particular couple involved, a requirement we do not witness in the composition of other documents.  We have also witnessed certain halakhot 'kicking in' once the get has been written even though it has yet to be delivered (for example, recall Rabbi Shimon's position that the husband loses his rights to peirot immediately after the get has been drafted, and Beit Shammai's claim that as long as the get has been written - EVEN THOUGH IT WAS NOT DELIVERED – the woman cannot be remarried to a kohen).  The Ramban's explanation of our gemara would seem to reflect the unique role which the composition of a get plays within the overall process.  As it is not merely preparatory in nature, but essential, it must be performed upon matter which can immediately be delivered.  If the current document cannot be legitimately presented 'as is,' but must first be reshaped, we cannot view the writing as the first stage of delivery.  According to the Ramban, intending to reformat the get converts the medium into something undeliverable.  The get must be drafted on a deliverable medium, or else the writing cannot initiate the overall process.

            Rabbeinu Tam (cited in Tosafot) argues with Rabbeinu Hai Gaon and disqualifies cutting a get even if the change had not been pre-planned.  Of course, Rabbeinu Tam admits that minor cosmetic cutting would not invalidate the get.  However, any major ketzitza would disqualify the get.  Evidently, the invalidation has little to do with the definition of the document as undeliverable.  If a person drafts a get without intention to reformat, he has clearly prepared a deliverable get; why should his subsequent decision to cut the get invalidate a get which was already properly prepared?  The writing was legitimately executed, and the delivery will ultimately be performed!

            Apparently, Rabbeinu Tam introduces a new requirement: continuity between the drafting and the delivery.  No significant action - premeditated or otherwise - can separate between the writing and the delivery.  Even if each component of the divorce process (writing and delivering) was performed properly, the process will be invalidated if they were not uninterrupted.  How might we explain the added requirement of continuity which Rabbeinu Tam asserts? 

            The past shiur (#15) considered the added element of geirushin: communication.  Given the fact that divorce provides not only legal release, but personal separation, a device to mediate the latter aspect was added to the otherwise formal, legal process.  A husband does not merely deposit a document; he also communicates the termination of their relationship.  We explored two possible instruments to provide this communication:

1) The verbal declaration of "Harei at mutteret;"

2) The delivery of the document - netina.

May we view the writing as part of the communication? Communicating the separation which is latent in any divorce demands writing the text (as mere verbal separation would be inelegant or cheap) and physically presenting that written message to its subject.  Disrupting the continuity of these two acts is tantamount to canceling the communicative nature of this process.  If, after writing, the get is reshaped, the initial intended message incorporated in the original get is never accurately delivered to the woman.

            The Ramban understands this stipulation as pertaining to the nature of the actual writing.  It must be performed upon material which can easily and immediately be transferred.  Only if a subsequent reformatting was originally planned do we define the medium as unsuitable and negate the value of the original drafting.  The Ramban's position supports only one conclusion: that the composition plays a vital (rather than incidental) role in the geirushin process.  It must therefore be performed upon something suitable for the culmination of that process.  What role ketiva (writing) plays is in no way elucidated by the Ramban. 

            By contrast, the Rabbeinu Tam begins to suggest the precise role served by the ketiva.  Any action - even if originally unintended - which disrupts the continuity between the ketiva and the netina will invalidate the process.  Evidently, this continuity is vital.  Whatever function the ketiva serves, it must be continuous to the netina.  This shiur posits that as that role is to convey the divorce, lack of continuity would essentially nullify the intended communication.  Communication, by its nature, must be personal.  As opposed to non-Jews, Jews convey their separation in prepared and formalized text.  However, to retain the personal nature of this communication, it must be directly sent from the pen of the husband (or his agent) to the hands of the woman.