Q&A: Majority–Presumption–One Witness
Majority–Presumption–One Witness
Question
In tractate Gittin 2b it is explained that one witness is believed against a majority view regarding prohibitions (for even according to the initial assumption that most people overseas are not expert in the requirement of “written for her sake,” the witness is believed when he says that it was written for her sake).
On the other hand, it is explained in the Talmud that where there is a presumption of prohibition, one witness is not believed to permit the matter.
This is difficult, since a majority is stronger than a presumption—and all the more so stronger than a presumption of prohibition. So certainly one witness should be stronger than a presumption of prohibition?
Thank you very much.
(By the way, I seem to remember that the Rabbi mentioned several times a structure of difficulty like this, including an analysis of how to break it down. I’d be glad for a reference.)
Answer
There is a property of relations called transitivity. A relation P is transitive if it satisfies the following property: if relation P holds between A and B, and also between B and C, then necessarily relation P holds between A and C. For example, the relations “equal to” and “greater than” are transitive.
Therefore, you are quite reasonably assuming transitivity among legal proofs, but surprisingly that assumption is mistaken. The commentators discussed this, for example in Kovetz Shiurim on Bava Batra (if I remember correctly, in the passage about majority and proximity, among other places), where he deals with several cases of non-transitivity among proofs in Jewish law. In general, X may be preferable to Y in one respect, and Z may be preferable to X in another respect. In such a case, Z will not necessarily be preferable to Y—that is, transitivity need not hold.
Another example of this surprising lack of transitivity is what Rabbi Avraham, owner of the stone, writes regarding the forbidden labor of building on the Sabbath and its derivatives. The primary category is assembling parts and creating a space (building a house). It has two derivatives: assembling parts without a space (making cheese curds) and creating a space without assembling parts (a tent). There is no similarity at all between those two, and yet both are derivatives of the same primary category. In other words, the relation of similarity is also not transitive.
This is especially true when we are dealing with a rule that is not evidentiary but behavioral, such as the original presumption (a presumption of prohibition). A presumption has no evidentiary weight; it is only a rule for conduct, and therefore it has no superiority in credibility over other proofs. Rather, that is simply how Jewish law rules, for its own reasons. So there is no necessity at all to assume that transitivity should apply here.
We discussed this at length in the book on loops, the fifth in the Talmudic Logic series: http://www.collegepublications.co.uk/stl/?00006
Discussion on Answer
What you wrote—that one witness is not believed to change a known status—that itself is changing a presumption. It may perhaps depend on how one understands the oath triggered by one witness: whether it exempts from an obligation created by one witness, or whether its absence creates the obligation. The dispute is whether one witness creates a monetary obligation or only an oath.
And in the commentators you can also see different explanations—for example, whether the need for two witnesses is in order to cross-check testimonies (Netivot HaMishpat), in which case it sounds like one witness does have some degree of credibility, just not full credibility.
So your explanation does not seem right to me. Simply speaking, one witness is entirely evidence, not merely a practical rule—but insufficient evidence. Therefore a woman is permitted on the basis of one witness, and he is believed in matters of prohibition, etc.
As for your question, one can think of several explanations. I’ll suggest something that occurred to me just now:
1. The credibility of one witness against a majority is very straightforward. True, most people are not expert in the “for her sake” requirement, but the witness testifies that this particular scribe was expert in it. That does not contradict the majority; it clarifies that here the writer belonged to the minority. So we no longer have a doubt, and there is no need to resort to presumption. This is exactly Rabbi Shimon Shkop’s reasoning in the law that we do not follow the majority in monetary law in the case of animals sold for plowing. His claim is that although most are sold for plowing, the person claims that he belongs to the minority, so what relevance does the majority have against him? He brought a proof there from Nachmanides regarding gifts sent before betrothal. In other words, one witness against a majority is a case of “fully permitted,” not merely “overridden.” He completely neutralizes the majority against him.
But one witness is not believed against a presumption, because here there is something standing against him. The presumption is not changed by one witness. And the rule is that only on the basis of two witnesses do we change presumptive statuses.
It should also be remembered that a majority is statistical evidence, whereas one witness is substantive evidence. (See my columns on statistical evidence in law.)
Shev Shema'teta 6, chapter 7
Nice. Thanks. I didn’t remember, and now I saw that at the beginning of chapter 7 the author had the merit of arriving at my more expansive view. He writes exactly what I wrote. Good for him.
Now I was thinking that based on this one can explain Rabbi Meir’s view that the minority can be joined to a presumption. He holds that a presumption has the power to identify the minority out of the majority, as we saw with one witness. But according to his view it comes out fifty-fifty, and it seems that the presumption only arouses doubt and neutralizes the majority, but does not decide the matter. And in this Rabbi Meir agrees with the fundamental halakhic position that a presumption does not truly clarify reality the way one witness does.
Thank you very much for the answer and for getting into the details!!
A. In the Rabbi’s explanation there is still a missing rationale for why a transitive relation does not hold here from the other side of the coin. That is, why in general is majority stronger than presumption? Presumption should be stronger, since the majority points against it, whereas the presumption does not point against the majority—it only says that the case before us belongs to the minority.
And regarding the definition of the credibility of one witness in matters of prohibition, two comments:
B. About the Rabbi’s assumption that one witness clarifies reality—it is very difficult, because then the witness is not testifying against the presumption at all; he is only testifying that the bill of divorce was written for her sake, and only as a second-order result would the presumption be uprooted, but not by his testimony itself, rather by giving the bill of divorce to the woman (as I noted above parenthetically).
C. What the Rabbi noted from the testimony of one witness in monetary law does not, in my humble opinion, relate to one witness in matters of prohibition. For one witness in matters of prohibition is believed even against majority and proximity, unlike in monetary law. And there one needs formal testimony in court and the other laws of testimony, unlike one witness in matters of prohibition.
As for the definition of “an established prohibition”:
D. I looked in Shev Shema'teta that Yareh cited above, and I saw that the Rabbi is very sharp in his words, and I felt confused because I too understood that his words are the same as what I suggested. For it is explained there, just as I wrote, that even in a place of “presumption,” one witness is believed to say that there never was a presumption. There is no obstacle to believing one witness against “the law of presumption”; rather, the problem is where, as a starting assumption, the legal status of the item under discussion is known. But where it has a presumption of prohibition, yet according to the witness’s statement it was never prohibited in the first place—there one witness is believed.
Thank you very much for the answer and for
A. I explained that the presumption does not clarify that the case before us belongs to the minority. It is a practical rule. That is unlike one witness.
B. The witness is not testifying against the presumption but rather comes to change the presumption. A presumption is not an evidentiary rule, and as such it does not contradict opposing evidence. Rather, according to Jewish law, only certain kinds of evidence (strong enough presumptions) can change an existing status. That is a practical rule, and it is not connected to comparing evidentiary strengths.
C. Why does that matter?
D. I was speaking about the beginning of his words. I did not go through the entire chapter.
Regarding B: it is still not understood why majority is stronger than presumption and one witness is not.
C: What I meant is the reasoning in Sha'arei Yosher (7:1, s.v. “u-mah”) that one witness for an oath is half of two witnesses who create a monetary obligation, and therefore he must be valid just like two witnesses. But for prohibitions he is not half of two witnesses, rather a separate law, and therefore the laws of witness-validity do not apply to him.
Thank you very much
B. Maybe because a majority changes the initial status and creates a new presumption (when the majority applies, the initial assumption about the truth is what fits the majority, not what is currently presumed), but if the presumption already exists, you need strong evidence to overcome it, and one witness is not strong enough.
Thank you very much for the analysis and the definition.
I tried to find a rationale for this distinction, and I thought as follows: when the Talmud says that one witness is not believed where there is an established prohibition, it does not mean because of the law of presumption as a rule for handling doubts. Rather, even without the law of presumption, it may still be that one witness is not believed where he comes to change our known understanding of the item in question (even though according to him there is a reason why our understanding should change).
But where he does not come to change prior knowledge regarding the item under discussion, one witness is believed even if he says that the item before us belongs to the minority.
(As an aside—this reasoning is based on a relatively complex prior assumption. We do not accept the statement of one witness that reality is in fact as he says, not even as a practical rule for how we are to conduct ourselves regarding the item in question. Rather, there is a disconnect between the question of reality and the question of conduct in matters of prohibition. It is only with respect to the question of conduct in prohibition that the witness is believed, but with respect to reality we do not accept his words. This is similar, for example, to the rule that in a case of definite claim versus uncertain claim, the definite claim prevails: this is not a question of credibility but a rule of legal conduct detached from the question of what actually happened in reality, and that is why the rule “the burden of proof is on the claimant” does not apply there.)
This is proven by the fact that the Talmud there says that a witness who testifies that the bill of divorce was written for her sake is testifying about “a matter of sexual prohibition” and against “an established prohibition.” But that is puzzling, because his testimony is not to permit the woman, but only about the validity of the divorce document—that it was written for her sake. That is not a matter of sexual prohibition, nor is there a presumption of prohibition in that point.
And it is proven that we do not accept his words about the reality and from there derive the rule for how we should conduct ourselves regarding the item in question; rather, the entire acceptance of his words is only with respect to the question of prohibition/permission.
If we assume this line of reasoning, it becomes possible to say that the witness is not believed where we already have prior knowledge regarding the prohibited status of the item under discussion—even without any law of presumption! For when we discuss whether to accept his words, we are not discussing the woman’s legal status because of the question whether the bill of divorce was written for her sake, such that we would need to say that the Talmud means to resolve the doubt by the rule of presumption. Rather, we are discussing whether the woman is permitted or forbidden. And since we have prior knowledge that the woman is forbidden, one witness is not believed.)
Is this also the Rabbi’s intent, or does the Rabbi hold that even though one witness is not believed against the practical rule of presumption, he is believed against a majority (that is how I understand the Rabbi’s meaning from the last paragraph of his answer)? If the Rabbi indeed means the second understanding, does the Rabbi have some rationale for it? What does it depend on? Why in practice is he stronger here and not stronger there?
Thank you very much!