Skip to main content
Iyun in Gittin -
Lesson 17

Daf 19b: Reading a Document to Illiterate Eidim

Mekorot for the following shiur: 

1) Gittin 19b, "Tanya kavatei ... shalchuha"
Ramban s.v. Ha de-ka'aru
Rashba s.v. Ve-she'ein
Tosafot Ha-Rosh s.v. Ve-im

2) Gittin 9b, "De-tanya ... lahem"
Tosafot s.v. Korin
Ramban 29b s.v. Ha de-amar, "Ve-im tishal ... ve-chotmin"

3) Rambam Malveh Ve-loveh 24:5-7; Gerushin 1:23 with Maggid Mishneh

Questions:

1) How do the Chakhmei Sefarad solve the problem of "eid mi-pi eid?"

2) How might Tosafot Ha-Rosh offer a different solution?

3) What is the relationship between Tosafot on 9b and the Ramban on 29b?

4) How might we explain the Rambam's condition that we can only read the document to witnesses who "understand the language of the document?"

 

Having considered the solution to witnesses who are incapable of signing their own names, the gemara considers the situation of illiterate eidim, who cannot read the document, or shtar, they are asked to sign. The gemara suggests that another individual read it for them. However, this practice is problematic, since the eidim have no firsthand knowledge of the text to which they affix their signatures. Their 'reader' might intentionally falsify information and thus allow the eidim to sign incorrectly. For this reason, Rabban Shimon ben Gamliel limits this practice to gittin as a special dispensation to prevent widespread cases of aguna.  When the gemara subsequently suggests a unique allowance for Rav Nachman (or other court officials who will inspire the reader to 'read' accurately), it recognizes an exceptional circumstance, which even Rabban Shimon ben Gamliel would embrace. Although Rabban Shimon ben Gamliel generally rejects this solution, he would make a special allowance if the chances of forgery were diminished.

 

            The Rishonim raise a more fundamental issue about this practice. Aside from the practical question - providing the illiterate witnesses with an accurate sense of what they are attesting to - a more legal or structural question emerges. Witnesses must have firsthand knowledge; they cannot testify about an event that they heard or read about. This form of testimony, referred to as that of eid mi-pi eid, is invalid. Eidim who discover the contents of a document by listening to it being read to them receive its information secondhand; this form of testimony should be invalid. (This question assumes a legal basis of eid mi-pi eid; if the entire concern with secondhand information surrounds its veracity, we might overcome that hurdle if the reading is executed in front of individuals such as Rav Nachman or others who inspire an accurate rendition).

 

            The Ramban raises the eid mi-pi eid concern and provides a response that suggests two different approaches. Initially, he justifies this practice because the contents of the document are merely gilui milta. The Rishonim use this term to refer to secondary information necessary to provide the framework or context of an event and do not treat it as testimony proper. For example, two witnesses must testify that a chalitza process occurred; however, one witness is sufficient to ascertain that Reuven is indeed the brother of the departed husband. Since this knowledge is only the predicate for the central event, one witness is sufficient to establish it. The Ramban appears to make a similar claim concerning our gemara. The eidim have personally witnessed the composition of the document. The specific content of the document does not effect the issue about which they are testifying; rather, they must be apprised of its content to assure that their signature is valid. As this information is only 'accessory information,' it can be supplied in an indirect manner.

 

            This first strain in the Rambam is articulated by the Ramban in another context and supported by Tosafot in Gittin (9b). Tosafot pose the same question as the Ramban and answer, 'Eidim do not have to testify about the event; instead, it suffices that they testify that this story is written in the document.' Tosafot remind us that witnesses do not actually view the event described in the document; all that they witness is the composition of the document describing an event that will take place in the future. The unique quality of a shtar allows it to serve as testimony even though its signatories have never witnessed the event about which it testifies. Receiving information about the event through other sources would have indeed violated their testimony as that of eid mi-pi eid. Since, however, they are only discovering the contents of the document, they are allowed to arrive at this information through a public reading. The text itself appears in front of them; they merely have difficulty understanding the meaning of this text.

 

Some Rishonim agree with Tosafot's claim but limit the scope of this case, insisting that some basic knowledge on the part of the eidim is required. If they are totally illiterate, having the document read to them would indeed constitute eid mi-pi eid. The Rosh claims that the eidim themselves can read, but they have minor difficulties with this particular document. The Rosh's claim is textually based in the episode of Rav Nachman, who was clearly literate but could not read a foreign language. The gemara also cites the example of Rabbi Yehuda, who could read though he had poor eyesight. According to the Rosh, if a person has no capacity whatsoever to read a document, others reading for him WOULD INDEED invalidate his testimony as eid mi-pi eid.

 

            The Kesef Mishna imposes an even stricter limitation. The Rambam cites our gemara but adds that the solution is only viable provided that the eidim recognize the language of the document. If they are so illiterate that they cannot read the language in which the document is written, then the presentation of this document's contents through alternate sources would indeed constitute eid mi-pi eid. If, however, they understand the language but cannot read the actual document, they can be informed of its contents through a public reading.

 

From Tosafot themselves, this issue is not entirely clear. Tosafot present their explanation without the qualifications suggested by the Rosh or the Kesef Mishna respectively. One could claim that, as the text itself lies in front of the witnesses, its interpretation is a secondary matter and can be furnished by external sources. Thus, even if the witnesses have absolutely no understanding of the language and are totally dependent upon the public reading, they can sign.

 

            The Ramban echoes this same issue in a different context. The gemara in Gittin (29) considers a situation in which the husband asks three people to supervise the divorce process and they in turn request eidim to sign upon a get. The Ramban questions the ability of these witnesses to sign the get when in fact they were not directly requested to do so by the husband (but rather by his agents). The question itself can be answered through alternate approaches but the Ramban's response - especially the syntax - is illuminating. Specifically in the case of get, eidim have no clue about the event of the divorce (as they do not witness the actual divorce proceedings). They merely sign and attest to the fact that the husband requested their signatures on a valid get he was preparing. They can affirm the husband's request based on their having heard it from him personally or having heard his request conveyed through his selected representative. Once again the Ramban reminds us that eidei chatima have nothing to offer regarding the actual event of the divorce. Rather, they speak of the composition of the document and, in the unique case of get, about the husband's specific request for their signatures.

 

            Having established the first element that the Ramban utilizes to justify this exercise, we can now consider the second aspect. Toward the end of his discussion, the Ramban introduces a second reason for allowing the witnesses to have the text read for them. As the contents of the document can be subsequently checked (milta de-avida le-igluyei) we allow others to read it to the eidim. One is tempted to explain this view along the lines presented earlier: if we are only concerned with the veracity of the reading, the prospect of future scrutiny assures an honest reading. In fact, it appears from the Rashba that this condition justifies our relying upon one person, rather than two, to read the document to the witnesses. Any additional legal concerns of eid mi-pi eid would then be solved through the first idea the Ramban presented, that this information is only considered gilui milta.

 

            However, we might also view this second statement of the Ramban from a more legal perspective. If information is forthcoming and can be readily investigated and acquired, we might not consider it formal testimony. Halakha recognizes information which only eidim can supply as legal eidut. When the information can be established through alternate means, it no longer classifies it as formal eidut. As this information is no longer eidut, it can be relayed to the signatories through other means.

 

            Having established the Ramban's treatment of our sugya, we might consider a more extreme position. The Maggid Mishna offers a different and controversial understanding of the Rambam's requirement that the eidim understand the get. According to the Maggid Mishna, they do not have to understand the exact text of the get, neither on their own, nor through the public recital. Instead, it is sufficient to explain to them - in a general sense - that this bill is being employed by the husband to divorce his wife. Recognizing this purpose, they may sign their names, even without actually grasping the meaning of the specific text of the document. This statement is a more radical version of the answer that the Ramban and Tosafot articulated. According to these Rishonim, signatories do not witness the event but rather the document's composition. The process of creating a shtar allows the generation of testimony through this process. Having witnessed the preparation of a document and subsequently signing their names, the eidim generate their testimony. According to the Maggid Mishna, the shtar testifies without the witnesses ever knowing what precisely is written in the document. This view clearly recognizes the shtar as an independent source of testimony and the witnesses as mere accessories to lend the document a more formal status. We discussed this issue — whether a document contains the recorded voices of the eidim or provides an independent voice — in shiur # 3. The Maggid Mishna clearly writes the eidim out of the shtar by allowing them to sign a document whose precise contents they have never fully read.

 

Sources for the next shiur: Daf 19b - Part One: The Requirement of the Eidei Mesira to Read the Get

1. Reading the get
    A. Tosafot 19b s.v. Tzerichi
        Mordechai, Paragraph 343
        Avnei Milu'im 31:4
2. Delivering a get that appears to contain no text
    A. Tosafot 19b s.v. Ta'ama
        Rabbeinu Crescas s.v. Amar Shemuel
    B. Rashi 19b s.v. Hashta
Rabbeinu Crescas s.v. Ve-parkinehu amar Shemuel...

Questions:
Part One:
1. Why must witnesses to the delivery of the get read the document?
2. Why should they read it before the issue?
3. How does the Avnei Milu'im alter our understanding of the eidei kiyum?

Part Two:
1. How does the Tosefta contradict Shemu'el's opinion?
2. Ultimately what is the basis of Shemu'el's "safek"?

This website is constantly being improved. We would appreciate hearing from you. Questions and comments on the classes are welcome, as is help in tagging, categorizing, and creating brief summaries of the classes. Thank you for being part of the Torat Har Etzion community!