Ketubot 19b – 20a Two Against Two Regarding A Deed - The Practical Significance
The Baraita asserts that if a second set of witnesses comes to disqualify the witnesses signed on a deed that had been properly authenticated, they are not believed. The Gemara notes that this is a situation of "two against two" (conflicting testimony of two sets of witnesses; see previous shiur), and therefore the lender cannot collect his debt with the deed. What then is the practical significance of the fact that the disqualified witnesses are not believed? The Rishonim offer several answers, the most prevalent of which is that the plaintiff can forcibly seize from the defendant that which is due him according to the deed. Why is seizure effective in a case of uncertainty such as this?
In the next shiur we shall continue in the Gemara, and see the practical significance of the fact that we are dealing with a situation of "two against two" with respect to a deed, and we shall focus on the most common understanding of the passage that such a deed allows for seizure. Continue in the Gemara until 20a, "u-ke-she-hu shoteh zavin." See also:
1) Rashi, s.v. oki; Tosafot, s.v. ve-oki – what is the rationale underlying the two distinctions that they propose?
2) Ramban 16b, s.v. ve-hai, "ve-ha de-ka amrinan hakha… be-chezkat lekuchot kaimi"; Ramban, Bava Metzia 6a (see below). What is the Ramban's rationale?
3) [Challenging source: Chidushei Rabbi Akiva Eiger on our passage (see below)].
Chidushei ha-Ramban, Bava Metzia 6a:
"אבל י"ל דכל תרי ותרי לא אמרינן אוקי ממונא בחזקת דמרה קמא, אלא מאן דתפס תפס במטלטלי, דספיקא דאורייתא הוא, וכל ספיקא דאורייתא אפילו גבי איסורי נמי לא אמרינן אוקמינן אחזקה, כדפרישית בקידושין פרק האומר (סו ע"א)".
Chidushei Rabbi Akiva Eiger, Ketubot 20a:
"אוקי תרי להדי תרי כו' מידי דהוי אנכסי דבר שטיא. תמיה לי דלאיזה ענין מביא ראיה מבר שטיא וכי עדיין לא ידענו דמספק לא מפקי' ממונא והרי פרכינן בפשיטות ומגבי' ביה כבשטרא מעליא ואמאי תרי ותרי נינהו הרי דמפשט פשיטא ליה דבתרי ותרי אין גובין, ואי דהראיה על הא דמקרע לא קרעינן והיינו דאם תפס מהני ומביא ראיה מבר שטיא דמהני תפיסה, הא באמת בבר שטיא לא מהני תפיסה ונהי דהטעם דבקרקע ל"ש תפיסה אבל במטלטלין מהני עכ"פ ליכא ראייה משם על תפיסה דמהני, וזהו תמיה עצומה לכאורה. ואף דנראה דרש"י ד"ה ואוקי ממונא נזהר מזה. ודקדק לפרש ואם קרקע היא תעמוד בחזקת המוכר ועל זה קאי הראייה מנכסי דבר שטיא. אבל מ"מ אין זה תעלה דהא זה ל"ש להאוקימתא דאין נאמנים דאוקי תרי להדי תרי ואוקי ממונא בחזקת מריה דזה פשוט דבספק אין מוציאין ממוחזק ובקרקע מי מקרי מוחזק זה אינו שייכות להאוקימתא, וזה עינן ודין בפני עצמו בקרקע מי מקרי מוחזק. וי"ל למה דמבואר ביבמות (דף לא) דההיא דבר שטיא הטעם דתרי ותרי ספיקא דרבנן ומדאורייתא בכל תרי ותרי מוקמי אחזקה מש"ה מוקמי הקרקע בחזקת מרא קמא אבל למ"ד תו"ת ספיקא דאורייתא דתרי ותרי מסלקין לכל החזקות ה"נ לא מוקמי הקרקע בחמ"ק, והנה בפשוטו היה נראה כיון דאסתלק חזקת מ"ק מוקמי הקרקע ביד המוחזק דא"א להוציא מהמוחזק, אבל בכדי ליישב קו' הנ"ל י"ל דלמ"ד תו"ת ספיקא דאורייתא מסלקין לכל החזקות ואפי' חזקת מוחזק ולא אלים חזקת המוחזק כלל במקום תו"ת והדין דיחלוקו, ולזה בתחלה פרכי' ומגבי' ביה כבשטרא מעליא היינו דמגבי' ביה כולא הא תו"ת נינהו, דעכ"פ אין גובה כולו, ור"נ הו"מ לשנוייה באמת לא מגבי' ביה כולו אלא יחלוקו אבל ר"נ ס"ל לקושטא דמלתא דבתו"ת אין מגבי' כלל דקיימי' חזקת ממון דתרי ותרי ס' דרבנן ועל זה מביא ראיה מבר שטיא דבתו"ת לא מסלקין החזקות".
I. The Gemara's conclusion
In the previous shiur, we dealt with the foundations of the law that when a second set of witnesses come and disqualify witnesses signed on an authenticated deed, they create a situation of "two against two" – conflicting testimony of two sets of witnesses. Today we will advance in the Gemara and deal with the practical significance of the doubt that is created with respect to the deed in question. The Gemara states as follows:
Our Rabbis taught: [If] two [witnesses] were signed on a deed and they died, and two [witnesses] came from the marketplace and said: "We know that it is their handwriting, but they were under duress, they were minors, they were disqualified [from giving] evidence," they are believed. But if there are witnesses that it is their handwriting, or where their handwriting came out from another place – from a document that had been contested and had been confirmed in court – they are not believed.
But do we collect with it as with a valid document? But why? They are two against two!…
Rather, Rav Nachman said: Set two against two, and leave the money in the possession of its owner. [It is] something like [the case of] the property of a [certain] madman. For a [certain] madman sold property. Two [witnesses] came [and] said: "When he was mad he sold [it]," and two [witnesses] came and said: "When he was sane he sold [it]." Rav Ashi said: Set two against two, and leave the money in the possession of the madman. (Ketubot 19b-20a)
Rav Nachman's answer to the argument that we are dealing with a situation of two against two is that indeed the law is "Leave the money in the possession of its owner." The simple understanding is that the law was decided in favor of the borrower, and that the lender cannot collect his debt based on the deed. But if this is the case, it is difficult to understand the Baraita, which distinguishes between a situation in which the invalidating witnesses are believed and a situation in which they are not believed. For it turns out that even in the situation in which they are not believed, they effectively nullify the deed. Owing to this difficulty the Ri Migash (cited by the Shita Mekubetzet, ad loc.) went as far as to propose that Rav Nachman essentially rejects the words of the Baraita, though he himself recoils from such a suggestion.
Another explanation in the Rishonim limits the Baraita to a particular case. Thus write the Tosafot (s.v. ve-oki) in the name of the Rash from Sens:
And it seems to the Rashba to limit the Baraita to the case of a receipt (shover), and therefore [the disqualifying witnesses] are not believed. For we set the money in the possession of its owner, and we presume that the witnesses on the receipt are fit and [therefore] he does not pay.
A shover (receipt) is a deed that testifies to the fact that a certain debt was repaid. Were we to accept the testimony of the second set of witnesses who come to invalidate the witnesses signed on the receipt, we would be obligating the borrower to repay his debt. But since we define the situation as one of two against two, they are not believed, and the borrower is exempt from repaying the loan, because we "set the money in the possession of its owner."
But the focus of our interest in this shiur shall be the answer given by Rashi and most of the Rishonim who followed in his footsteps:
And leave the money in the possession of its owner – in the possession of the borrower, and if it is landed property, it stands in the possession of the seller. And when it says in the Baraita: "They are not believed" – not that it should be a valid document with which to collect, but rather that we do not tear it up. And if he [the lender] seizes something, and then subsequently the borrower makes a claim against him, we do not remove it from him [the lender], for we say: Set two against two, and leave the money in the possession of him by whom it is.
That is to say, even in the case of a deed of loan, not believing the witnesses who come to invalidate the deed has an important ramification - namely, that even though the lender cannot collect his debt in court (because the situation is defined as two against two), if he seizes payment, the seizure is effective. The Tosafot Rid (printed below on p. 20b) disagrees, arguing that seizure does not help, and this also seems to be the position of the Rash of Sens and the Ri Migash, cited above, but as was already noted, most of the Rishonim accept the view of Rashi.
II. Seizure with a claim put forward with certainty and with a claim put forward with uncertainty
We must now consider some of the laws governing seizure in cases of doubt. By the nature of things, in the framework of a weekly shiur on Ketubot we cannot fully exhaust a topic that is the focus of some of the weightiest passages in Bava Metzia, and the subject of an entire literature, including "Tokfo Kohen" of the Shakh and "Kuntrus ha-Sefekot" of the brother of the author of the Ketzot ha-Choshen. We shall therefore content ourselves with a brief discussion of some of the main issues.
The passage of "Tokfo kohen" (Bava Metzia 6a-7a) relates to a clean animal whose status as a firstborn is uncertain; owing to the uncertainty the kohen cannot remove it from its Israelite owner. According to most Rishonim, the Gemara concludes that even if the kohen forcibly seizes the animal, we do not say that now the burden of proof falls upon the Israelite who wishes to remove it from him, but rather "if the kohen seizes it, we remove it from his hands." To explain this position we can say that not only does the law of chazaka in monetary matters bar the court from removing anything from the possessor's hand without proof, but it also decides the matter that for us the article belongs to the person in whose possession it is found. Hence, a person who forcibly removes that article from him is from our perspective a robber, even though we know that the possibility exists that he is merely taking that which belongs to him.
Against this conclusion, various talmudic passages imply that seizure is effective, one of them being our passage (according to the understanding of most Rishonim). The Kuntrus ha-Sefekot (2, 1) argues that our passage underlies the Rambam's exceptional ruling (Hilkhot Bekhorot 2:6) that if the kohen seized the animal, we do not remove it from him. On the other hand, the Rishonim who disagree with the Rambam suggest various distinctions between the different passages.
Let us open with the first suggestion made by the Tosafot on our passage:
The law regarding the firstborn animal is different, because the kohen seizes it based on an uncertainty, for he does not know whether it is a firstborn. But here where he claims with certainty, seizure is effective.
The Rosh in Bava Metzia formulates this distinction as follows:
If the kohen seized [the animal], we remove it from his hand, because seizure is only effective for someone who claims [it is his] with certainty. But the kohen, even after seizing [it], can only put forward his claim with uncertainty. Therefore we remove it from him, for we leave it in the hands of the mara kama… The rule is that in the case of a claim put forward with certainty, seizure is effective; since he claims [it is his] with certainty and he is [now] in possession of it, we don't have the power to remove it from his hands. But someone who is in doubt whether or not it is his, his seizure does not help to remove [the article] from the possession of the mara kama. (chap. 1, no. 13)
This answer implies that fundamentally seizure is effective; the seizer is not necessarily regarded as a robber, and now he is regarded as having present possession. And even though the other party enjoys the chazaka of the mara kama, for he is the last person recognized by us as the owner, the present possession of the article wins out against the chazaka of the mara kama. When, however, the present possessor puts his claim forward with uncertainty, his seizure is not sufficiently meaningful, and the chazaka of the mara kama is strong enough to remove the article from him. For an uncertain claim weakens his possession, as is explicitly stated by the Tosafot later in our masekhta:
Only here, where the possessor himself is in doubt about the cow, and he does not have any clear claim, and his possession is as if it did not exist – do we follow the chazaka of the mara kama, and it is like the case of a kohen who seized [an animal], where we remove it from his hand. (76b, s.v. al)
Of course, the words, "and his possession is as if it did not exist," do not mean that the chazaka does not exist at all, but rather that it is weakened, and therefore the chazaka of the mara kama removes the article from him. We can explain this by saying that the decisive level in a monetary chazaka that determines that the article belongs to the person enjoying the chazaka does not exist when the chazaka itself is weak and not based on a definitive claim that the article belongs to the person with the chazaka. All that we are left with is the practical level that states that a court may not remove an article from a person's possession without good cause to act, but restoring the article to the other party based on his chazaka of the mara kama is a sufficiently good reason to act.
III. Seizure before the uncertainty arises
Another answer is brought in the continuation of the Tosafot on our passage:
It may also be argued that here we are dealing with a case where the seizure took place before the uncertainty arose.
That is to say, if the plaintiff collected his debt with the help of a deed before the second set of witnesses came and disqualified the witnesses signed thereon, we do not remove the money from the plaintiff and return it to the defendant, even if the defendant had claimed from the outset that the witnesses signed on the deed are disqualified from giving testimony. The rationale seems to be that it is only at the time that the uncertainty arises that we determine who is regarded as the person enjoying the chazaka of possession. Since at the time when the uncertainty arose, the plaintiff was already in possession of the money, the burden of proof falls upon the defendant.
We shall not expand upon the matter at this time, but merely note that this view is not accepted by all (see Kuntrus ha-Sefekot 2,2 and on); and that there is disagreement about how to define the moment when the uncertainty arises, whether it is the moment that the question arises (Tosafot, Bava Batra 2a, s.v. lefikakh) or the moment that the parties appear before the court (Rabbenu Yona, ibid. 4a).
IV. Seizure in a case of two against Two
The Ramban in Bava Metzia 6a and Ketubot 16b adopts a different approach to explain the efficacy of seizure implied by our passage. Let us examine his words in Bava Metzia:
But we can say that in every case of two against two, we do not say "leave the money in the possession of the mara kama," but rather he who seized it seized it in the case of movables, for it is an uncertainty by Torah law. And regarding every uncertainty by Torah law, even with respect to prohibitions, we do not say, "Leave it with its chazaka," as is explained in Kiddushin, chap. Ha-omer (66a).
The Ramban argues that the uncertainty created in the case of two against two is different from the usual case of uncertainty. The term that he uses – "an uncertainty by Torah law" – is taken from the Gemara in Yebamot 31a, where it is used not in the context of monetary laws, but rather in the context of prohibitions. The simple explanation is that by Torah law we do not decide doubts based on the existing situation (chazaka deme'ikara) in the case of two against two. The Gemara's conclusion there, however, seems to be that the case of two against two is not "an uncertainty by Torah law," but merely "an uncertainty by rabbinic decree." That is to say, by Torah law we rely on a chazaka even in the case of two against two, and it is the Rabbis who introduced a stringency with respect to prohibitions not to rely on a chazaka for leniency. The Ramban, however, did not understand the matter in this way; perhaps he had a different reading.
We shall not explain here at length why the case of two against two is different than a regular uncertainty with respect to the chazaka. We shall content ourselves with the following explanation: According to Halakha, witnesses are the ultimate evidentiary tool, and they create absolute certainty. Even when they clash with another set of witnesses, they cannot be nullified, and they still provide absolute certainty, only that now we also have absolute certainty in the opposite direction as well. Such a situation creates an uncertainty that does not follow from any lack of knowledge on our part, but rather from excessive knowledge – a clash of contradictory knowledge. In such a situation, the chazaka is irrelevant, for it is a tool that is used when we do not know the truth, but not where we have all the knowledge we need. The opposing witnesses make the chazaka irrelevant and the supporting witnesses make it unnecessary.
How does the Ramban apply this to seizure in monetary cases? The Ramban says that "in every case of two against two, we do not say 'leave the money in the possession of the mara kama.'" It seems, however, that he is not referring only to the chazaka of the mara kama, but also to the chazaka of ownership enjoyed by the person from whom the money was seized, which had defined him from our perspective as the definite owner, so that seizing from him is considered robbery. A situation of two against two nullifies this chazaka, by way of a kal va-chomer argument that "even with respect to prohibitions we do not say, 'Leave it with its chazaka.'" The chazaka of the mara kama, according to certain understandings, is a type of chazaka deme'ikara regarding ownership (see Kuntrus ha-Sefekot 1,5), and therefore we are dealing with a simple application of the term "uncertainty by Torah law." But even with respect to the chazaka of ownership, even though we are dealing with an entirely different concept, we can say that our assumption that we are dealing with the owner as long as it has not been proved otherwise, is irrelevant when we have at our disposal the strongest proof, i.e., witnesses. This being the case, we do not restore the money to the party who enjoys a chazaka, and seizure is effective.
The Ramban, however, adds a very important qualification:
"In every case of two against two, we do not say 'leave the money in the possession of the mara kama,' but rather he who seized it seized it in the case of movables." It is clear from the Ramban that with respect to landed property, seizure does not help even in a case of two against two. Why not? It stands to reason that the concept of seizure regarding landed property is much less significant than regarding movables. A person cannot seize landed property; at best he can stand on it and work upon it. He can also not distance it from its owner; at best he can temporarily bar him from entering it – "landed property remains in the possession of its owner."
But even if we accept this argument regarding landed property, we must still understand why the court removes the seizer and restores the property to the mara kama – surely we said that the situation of two against two nullifies the chazaka of the mara kama. This being the case, what special status does the mara kama enjoy?
It stands to reason that even according to the Ramban, who says that two against two is an uncertainty by Torah law, this does not mean that the concept of chazaka is entirely nullified in the case of two against two. This is already reflected in the Ramban's main discussion of the topic, to which he refers at the end of the passage that we saw, in his novellae to Kiddushin 66a. From there it follows that there are certain Torah laws, regarding which a chazaka helps, despite the fact that it is "an uncertainty by Torah law." It seems that according to the Ramban, the term "uncertainty by Torah law" means that by Torah law there is an uncertainty, but it is still possible to use a chazaka as a practical procedure regarding certain issues, and here we must decide regarding which issues it is effective. One expression of this relates to seizure: it stands to reason that the chazaka of the mara kama exists, but it only works as a practical procedure. Accordingly, the seizer is not regarded as a robber, and his seizure helps to cast the burden of proof onto the other party. When, however, the seizure is insignificant, as in the case of landed property, we recognize the chazaka of the mara kama and restore the property to him.
The Kuntrus ha-Sefekot (2,1) raises an objection against the Ramban's qualification regarding landed property from the Gemara in Yebamot (31a-b) with respect to the law mentioned also in our passage – the property of the madman, which the Gemara in Yebamot explicitly identifies as landed property. The Gemara there brings the law of the madman as an expression of the fact that two against two is "an uncertainty by rabbinic decree":
Even in the case of two sets of witnesses the uncertainty is only by rabbinic decree, for we say: Set the two against the two, and leave the woman in her original state. This is similar to [the incident with] the property of a [certain] madman. For a [certain] madman sold property. Two [witnesses] came and said: "When he was sane he sold [it]," and two [witnesses] came and said: "When he was mad he sold [it]." And Rav Ashi said: Set two against two, and leave the land in the possession of the madman. (31a-b)
What is the reality with respect to the madman? If the landed property is still in his hands, and the two sets of witnesses disagree as to whether he must hand it over to the buyer, surely it is obvious that in such a case the court would not remove the property from his possession. And this is true whether two against two is "an uncertainty by Torah law" or it is "an uncertainty by rabbinic decree." Perforce, we are dealing with a case in which the landed property is already in the buyer's hands, and now the witnesses disagree whether or not the seller had been sane at the time of the sale. And why does the ruling in the case that the property is returned to the madman prove that two against two is "an uncertainty by rabbinic decree"? Because had it been "an uncertainty by Torah law" we would have said that the seizure should have been effective, and that the buyer should not have had to return the property to the madman. It turns out then that the Gemara in Yebamot supports the Ramban's position that seizure is effective in the case of two against two if we see it as "an uncertainty by Torah law," but it goes further and applies it even to landed property!
As was stated above, the Kuntrus ha-Sefekot raises this objection, but leaves it unanswered. Later in the Kuntrus (7,8), he returns to the issue and proposes a resolution, according to which the situation of "an uncertainty by Torah law" by itself does not allow for seizure of landed property, but in the case of the madman – according to his understanding of the case – we are dealing with seizure with permission, for the transfer of the property was executed not by force, but by the madman, and therefore the seizure is effective.
This rationale underlying the combination proposed by the Kuntrus ha-Sefekot is not clear to me. But more than that, it seems that he only had before him the Ramban's novellae to Bava Metzia, but not his novellae to Ketubot. For a careful examination of the Ramban's words in Ketubot refutes his attempted resolution of the difficulty. Let us see what the Ramban writes above, p. 16b:
It seems to me that in every case of two witnesses against two witnesses, we do not remove the [seized property] from his hand in the case of movables. This is not true in other cases of uncertainty, e.g., where the kohen seized [the firstborn]… for witnesses are stronger [proof], and we do not remove [property] against them… And this applies to movables, but in the case of landed property, seizure is not effective. But rather since the family members have a chazaka on the property, he who wishes to remove it from them must bring proof…. And below as well we say that since he has a chazaka from his father, we set the property in the possession of the madman, even if the buyer seized it. And there we are dealing with landed property, as Rashi explains. But in a similar case involving movables, they remain in the possession of the buyer.
The Ramban repeats what he had said that seizure does not help with respect to landed property even in the case of two against two, and the proof that he brings is from the case of the property of the madman, regarding which seizure does not help! Clearly then he does not maintain that there is some local element there that allows for seizure if two against two is "an uncertainty by Torah law," but rather this case proves that even if two against two is "an uncertainty by Torah law," and for this reason seizure works in a case of two against two, seizure does not work on landed property. While it is true, as noted by the Kuntrus ha-Sefekot, that the Gemara in Yebamot is explicitly against this, and the law of the madman is brought as support that the case of two against two is "an uncertainty by rabbinic decree," the Ramban clearly does not agree with this. According to him, apparently, the Gemara there brings a proof according to those who maintain that the case of two against two is "an uncertainty by rabbinic decree," but those who maintain that the case of two against two is "an uncertainty by Torah law" will deal with this proof and say that since we are dealing with landed property, seizure does not help (and this, as we explained above, is because even according to them the chazaka is not entirely canceled in the case of two against two).
V. The law of the property of a madman in our passage – Rabbi AKiva Eiger's novel explanation
Before concluding let us consider the problematic course of the Gemara in our passage. Rav Nachman's conclusion regarding the fact that there is conflicting testimony about the deed is: "Rather, Rav Nachman said: Set two against two, and leave the money in the possession of its owner."
What this means is, according to most Rishonim, that money cannot be collected by way of the deed, but seizure is effective, and according to the Ramban, this is because we are dealing with an uncertainty created by the conflicting testimony of two sets of witnesses. According to this, however, the proof that the Gemara brings is very strange: "[It is] something like [the case of] the property of a [certain] madman. For a [certain] madman sold property. Two [witnesses] came [and] said: 'When he was mad he sold [it],' and two [witnesses] came and said: 'When he was sane he sold [it].' Rav Ashi said: Set two against two, and leave the money in the possession of the madman." Surely regarding the madman – as we saw in the Gemara in Yebamot – the law is that seizure does not work. While it is true that there is no contradiction between the case of the deed and the case of the madman, for the Ramban distinguishes between movables and landed property, but how are we to understand the proof that is brought from the case of the madman? Surely there we are dealing with a case where there was seizure, only that it is not effective, whereas here we are dealing with a case where there wasn't any seizure, but had there been seizure, it would have been effective. So also asks Rabbi Akiva Eiger in his novellae, ad loc.
Rabbi Akiva Eiger's answer is exceedingly bold. According to him, the law of the madman – as opposed to the position of the Ramban according to our understanding – is only according to the authority who says that two against two is "an uncertainty by rabbinic law." This is because the authority who says it is "an uncertainty by Torah law" maintains that the situation of two against two entirely nullifies the chazaka of the mara kama. This, however, does not mean that seizure helps, because the new chazaka that the seizure tries to create is also nullified by the situation of two against two. Therefore, according to the authority that says that the case of two against two is an uncertainty by Torah law, the law should be that they divide the property between them (!!), and there is no difference between landed property and movables. According to this understanding, the Gemara initially agrees with the position that the case of two and two is an uncertainty by Torah law, and therefore it asks: "But do we collect with it as with a valid document? But why? They are two against two!", because the situation of two and two gives the plaintiff only half of what he wants. Rav Nachman answers that nothing at all can be collected with the deed, and even seizure will not help. This is because the situation of two against two is "an uncertainty by rabbinic law." Rav Nachman brings support for this from the law of the madman, which proves that the situation of two against two is "an uncertainty by rabbinic decree," and therefore seizure does not help.
Of course, the words of Rabbi Akiva Eiger go against almost all the Rishonim, and he is forced to join the camp that understands that seizure does not help in our passage, and that Rav Nachman had great difficulty explaining the Baraita, if he was able to explain it at all. Moreover, Rabbi Akiva Eiger's proposal is very novel. But I did not find any other explanation of the difficulty rising from the course of the Gemara. The matter requires further examination.
(Translated by David Strauss)
 For our purposes, we shall follow the simpler assumption that we are dealing here with seizure performed in the presence of witnesses, that is to say, a situation in which the seizer does not enjoy a miggo that he could have said that he did not seize the money. Regarding this, see Rashba, s.v. ein ne'emanin; Ritva, s.v. ha de-amrinan; Ramban 16b, s.v. ve-hai.
 This may also be understood from the words of the Rambam, Hilkhot Edut 7:7. The Rid does not explain how Rav Nachman's statement can be reconciled with the Baraita. It is possible that he maintains (this may also be the position of the Ri Migash in his second explanation) that the very fact that we do not tear up the deed constitutes a practical difference between the two parts of the Baraita. What this means is that if the witnesses who came to disqualify the witnesses signed on the deed are later turned into edim zomemim, the authenticated deed can then be used to collect the debt from the borrower's assets.
 Our passage deals with a monetary debt, regarding which the rule that "there is a presumption that whatever is in a person's hands belongs to him" would seem to have no validity in its plain sense. For there is no disagreement as to the ownership of the money, but only a question whether or not a debt exists. However, the power of a chazaka of ownership to decide an uncertainty (according to those who maintain that such a power exists) is not based merely on the realistic assumption that the chazaka indicates ownership, but on the solid legal assumption maintained by the court that a person is not obligated to return property or make a payment unless it is proven otherwise.
 Of course, seizure is only effective when a doubt arose as to the facts in a particular situation, as in our passage or in the passage dealing with the animal whose status as a firstborn is in doubt. It is clear, however, that in the absence of any uncertainty, a person who forcibly seizes property from his fellow is regarded as a robber, even if he claims that the property belongs to him.
 It is interesting to note that the Ran in his novellae formulates his position in a slightly different manner than the Ramban: "In every case of two against two, we do not say 'leave the money in the possession of the mara kama,' because we rely on each set of two witnesses as if what they say is certain. And it is not right that we should remove money from the seizer's possession, since there are two witnesses who testify in support of what he said." It seems to me that he had to say this in the wake of a certain disagreement that he has with the Ramban regarding seizure, which we cannot discuss here. In any event, the mechanism is different: He does not build on a challenge raised against the chazaka, but rather that the seizure has a solid foundation – witnesses who cannot be cancelled even by other witnesses, and it is difficult to remove money from a party supported by witnesses. It is possible that the Ramban agrees that his position is based on the foundation that "we rely on each set of two witnesses as if what they say is certain," because this is what undermines the chazaka. But according to the Ran the critical factor is not the undermining of the chazaka, but rather the fact that the seizer is supported by witnesses. It should be noted that the Ramban in Ketubot 16b, while referring to what he says in Bava Metzia, uses a formulation that is similar to that of the Ran: "For witnesses are stronger [proof], and we do not remove [property] against them…."
 Furthermore, seizure with respect to landed property is clearly not totally meaningless, for in a case where neither party has supporting witnesses that he is the mara kama, we invoke the principle of "kol de-alim gevar," "the strongest wins out," and whoever seizes it retains possession (Bava Metzia 34b)!