“Four Guardians and Three Sets of Laws” (The Proprietary Rights of the Renter and the Borrower as a Basis for Classifying the Types of Guardians)

2006

A. Introduction: The renter as a fourth type of guardian

The Mishnah in Bava Metzia 93a (see also Shevuot 49a-b) states:

There are four types of custodians: the unpaid custodian, the borrower, the paid custodian, and the renter. The unpaid custodian swears regarding everything, and the borrower pays for everything, while the paid custodian and the renter swear regarding breakage, capture, and death, and pay for loss and theft.

Thus, the Mishnah counts four different types of guardians. The Talmud (ibid. and ibid.) cites a tannaitic dispute (R. Meir and R. Yehuda) regarding the renter: are his laws those of an unpaid guardian or of a paid guardian? It concludes in the following formulation:

If so, four? There are only three! Rabbi Naḥman bar Yitzḥak said: There are four custodians, but their laws are three.

The law follows the anonymous Mishnah, that the renter has the law of a paid guardian. But, in simple terms, the concluding statement — ‘four guardians and three sets of laws’ — is agreed by all. Both according to R. Meir and according to R. Yehuda there are only three halakhic categories for the four guardians. This is indeed what Tosafot write, s.v. ‘If so,’ ad loc., and below we shall spell this out more fully.

Maimonides writes likewise at the beginning of the Laws of Rental:

Four types of custodians are stated in the Torah, and they have three sets of laws. And these are the four custodians: the unpaid custodian, the borrower, the paid custodian, and the renter.

It is quite clear from the totality of halakhic sources that Jewish law indeed sees four different types among the guardians. Most enumerators of the commandments count only three guardians, but most of them note (see Maimonides’ Sefer HaMitzvot, positive commandment 243, and Sefer HaChinukh, commandment 59, among others) that the commandment of the paid guardian deals with both the paid guardian and the renter. That is, the renter is also on the map. An exception is Saadia Gaon, who in his parashiyyot (see his Sefer HaMitzvot, vol. 3, ed. R. Yerucham Perla, sections 23-25) appears to count the renter separately.

The question that arises here is why the Mishnah and the halakhah define four guardians at all. Why do we not classify the laws of guardians by their actual laws, that is, into three types? In the background one should note that the Torah itself contains only three sections dealing with guardians, and the Talmud (see the entire discussion at the beginning of the chapter HaSho’el in Bava Metzia) associates each of them with one of the three halakhic categories (unpaid guardian, paid guardian, and borrower). Rashi (s.v. ‘How does he pay,’ Bava Metzia 80b) even bases the Gemara’s hesitation there about the renter’s laws on the fact that there is no separate Torah section for the renter, or at least that the renter’s laws do not appear there explicitly.[1] See also Rashi on Exodus 22:14; we shall return to his comments below.

For example, the Talmud in Bava Kamma 107b cites the statement of Rami bar Hama:

The four types of custodians require partial denial and partial admission; and these are they: the unpaid custodian, the borrower, the paid custodian, and the renter. Rava said: What is the reason of Rambaḥ? In the case of the unpaid custodian it is explicitly written, “for this is it”; the paid custodian derives it through a verbal analogy of “giving” from the unpaid custodian. The borrower—“and if a man borrow”—the vav adds to the previous subject.

And what about the renter? He is counted among the guardians, but the Talmud brings no source for him at all. It does not even bother to say that he is derived from the paid guardian. The Talmud simply ignores him, and it appears that in its eyes the renter is nothing more than a subtype of paid guardian. A similar treatment appears with respect to misappropriation in the sugya of Bava Metzia 41b, and likewise with respect to other laws in the sugyot at the beginning of the chapter HaSho’el (Bava Metzia 95-96) and elsewhere.

So too in Tosafot, s.v. ‘Thirteen,’ Bava Kamma 4b, where the four guardians are discussed as part of the principal categories of damages, Tosafot write that there are only thirteen, because the renter is not a separate category: he is either like a paid guardian or like an unpaid guardian.

This picture sharpens even further the difficulty raised above. If there are in fact only three halakhic categories, both in the Torah’s sections and in the operative laws of guardians as practiced, then why do the Sages and the medieval authorities (such as the commandment-counters mentioned above) treat the renter as a fourth type of guardian, instead of placing him under one of the three existing categories? In what sense are there four types of guardians at all?

At first glance, the classification into four types might be based on the situation that generated the duty of guarding. In other words: the halakhic categories are indeed only three, but there are four situations in which a duty of guarding arises, each falling under one of those three categories. The renter is one of those situations, and the situation of the ordinary paid guardian is different from it; hence the renter counts as a separate type.

But this explanation is implausible. There are a number of additional situations in which duties of guarding arise that fall under those same three categories. For example: the finder of lost property (about whom Rabbah and R. Yosef disagree; see Shevuot 44a and parallels), who according to the law is like a paid guardian; the artisan (see Bava Metzia 80b and elsewhere), who is likewise like a paid guardian; and a creditor holding collateral (see Bava Metzia there), who according to most views is also liable like a paid guardian. So what is unique specifically about the situation of the renter that turns it into a fourth type of guarding? If the situation is decisive, there should be many more types (artisan, collateral, lost property as well). And if the laws are decisive, then there should be only three types.

We should stress that there are very few differences between the laws of the ordinary contractual paid guardian and the laws of those other, non-contractual, types just mentioned. But with respect to these differences, the renter resembles them rather than the ordinary paid guardian. For example, the level of care required of a paid guardian is above ordinary care over one’s own property (which is all that an unpaid guardian is required to exercise; see Bava Metzia 93b). By contrast, for the artisan, the creditor holding collateral, and the finder of lost property, ordinary care is sufficient (see Minchat Chinukh, commandment 59, sec. 1, s.v. ‘And so too the finder of lost property’). What, then, is the renter’s law? Presumably he is like them, since the reason a paid guardian is obligated in extra care is that he received payment for his guarding (see Bava Metzia there), unlike these others, who did not receive direct payment for guarding.[2] If so, what is special about the renter among paid guardians, that he is counted as a fourth type?[3]

In this article I shall attempt to explain the map of the various types of guarding, and especially why there are four different types of guardians, in light of the nature of the different liabilities. We shall see that the four types express four different kinds of obligation, even if most of the laws of two of them (according to the halakhah: the paid guardian and the renter) converge.

At the outset we shall deal with the proprietary rights held by the renter and the borrower, in contrast to the unpaid and paid guardians. From this we shall see the different character of their liabilities, as opposed to those of the unpaid and paid guardians. Later we shall distinguish between the renter and the borrower on the one hand, and the paid guardian on the other, and we shall see why the renter indeed constitutes a separate type.

To conclude the introduction, I note with some surprise that although this topic is highly fundamental to the laws of guardians, I did not find any direct discussion of it in the commentaries. After completing this article I found two sources that deal with it directly: the Lubavitcher Rebbe, in Likutei Sichot,[4] and Rabbi Lichtenstein, in his lectures on the chapter HaSho’el (especially in the first three chapters).[5] Their conclusions are very close to mine, and I have therefore incorporated what emerges from their remarks into the discussion.

B. The proprietary rights of guardians in the deposited object: renter and borrower[6]

The basic contradiction

A number of later authorities[7] discussed whether the rental payment is purchase money for a proprietary interest in the body of the object, such that the renter’s use of the object is use of his own object (that is, he has a proprietary interest in the object itself for its usufruct), or whether the payment is only for the uses, while he has no proprietary interest at all in the body of the object (which seems to fit better with the rule that rent accrues from beginning to end).

One practical implication is the disputed law in the sugya of a house that collapses in the middle of the rental period (Bava Metzia 103a). The decisors there disagree whether the renter must pay the entire sum because he has already acquired the house, or whether he must pay only for the period he already lived in it (below I shall detach this practical implication from the present discussion).

But the root of the matter is already present in a contradiction between different passages in the Talmud itself, and in disputes among the medieval authorities around them, over whether the renter has a proprietary interest in the rented object or not.[8] The first source appears in the sugya of Bava Metzia 56b, which discusses whether the law of price fraud applies to rental:

Rabbi Zeira asked: Is there a law of overcharging in rentals, or is there no law of overcharging in rentals? The Merciful One said “sale”—but not rental; or perhaps there is no difference. Abaye said to him: Does it say “sale” forever? It simply says “sale,” and this too, for its day, is a sale.

At first glance, this seems clearly to imply that the object is sold to the renter for the days of the rental, and that he has a proprietary interest in its body (for usufruct). As we shall see, a number of medieval authorities indeed took it that way. But the medieval authorities noted that in Avodah Zarah 15a, which discusses the prohibition on renting one’s house or animal to a gentile lest he use them for transgression, it seems clear that rental does not effect acquisition:

And everywhere they may not sell large livestock, etc. What is the reason? Granted, we are not concerned about bestiality, but we are concerned that one may have it do labor. But let him have it do labor—once he sold it, he has acquired it! It is a decree because of borrowing and because of renting. Borrowing acquires it, and renting acquires it! Rather, Rami son of Rabbi Yeiva said: It is a decree because of a trial run, for sometimes he sells it to him near sunset on Friday, and he says to him: Come, try it out; and it responds to his voice and goes because of him, and he is pleased that it goes, and it turns out he is driving his animal on the Sabbath; and one who drives his animal on the Sabbath is liable to bring a sin-offering. Rav Shisha son of Rav Idi objected: Does renting effect acquisition? But we learned: Even in a place where they said one may rent out a house, they did not say so regarding a residence, because he brings an idol into it. And if you say that rental effects acquisition, then when he brings it in, he is bringing it into his own house! Idol worship is different, because it is especially severe, as it is written: “You shall not bring an abomination into your house” (Deuteronomy 7). Rav Yitzḥak son of Rav Mesharshiya objected: Does renting effect acquisition? But we learned: An Israelite who rented a cow from a priest may feed it vetches of terumah, but a priest who rented a cow from an Israelite—even though its sustenance is his responsibility—may not feed it vetches of terumah. And if you say that rental effects acquisition, why may he not feed it? The cow is his! Rather, conclude from this that rental does not effect acquisition.

The Talmud gives two implications of this: (1) it is forbidden to rent a house or animal to a gentile, since if he commits a transgression with it, he does so using the object of the Jewish owner; (2) a priest who rented a cow from an Israelite may not feed it terumah, because it is not the priest’s property.

Thus, apparently, there is a contradiction between the Talmud in Bava Metzia regarding price fraud and the conclusion of the Talmud in Avodah Zarah. The medieval authorities split, following this contradiction, into two camps: some rule in accordance with the sugya in Avodah Zarah, that rental does not effect acquisition, and reconcile the sugya in Bava Metzia in various ways; others rule in accordance with the sugya in Bava Metzia, that rental does effect acquisition, and reconcile the sugya in Avodah Zarah in various ways.

The view of Tosafot and those who follow them: rental does not effect acquisition

Tosafot consistently rule that rental does not effect acquisition. For example, Tosafot, s.v. ‘And this,’ Bava Metzia 56b, write (and similarly Tosafot, s.v. ‘And now,’ Avodah Zarah there):

This is difficult, for in the first chapter of Avodah Zarah (15a, and there in Tosafot, s.v. Vehashta) it is proven that rental does not effect acquisition, from the case of an Israelite who rented a cow from a priest and from a house rented for dwelling. And one may answer that here, regarding overcharging, the Torah uses the extra term "sale" to include rental.

Thus, according to Tosafot, the law of price fraud is a special law; but as a general matter rental does not effect acquisition. We should note that even according to Tosafot there is apparently some element of acquisition in rental, since Tosafot too agree that rental is included by the term sale in the Torah. But in their view rental is a weaker kind of sale.

One might have understood this as a sale of usage rights (in halakhic terminology: usufruct). Unlike a mere temporary permission to use, rental obligates the lessor to provide the right of use; so something is indeed acquired here. But this is not acquisition in the body of the object itself (not even the body for its fruits), but only acquisition of the rights of use. Indeed, the Taz (Hoshen Mishpat sec. 315, and see the commentaries there) explains rental as acquisition only of the object’s usufruct (unlike other areas of Jewish law, where acquisition of usufruct alone is generally impossible, and only acquisition of the corpus for its fruits exists).

However, in Tosafot, s.v. ‘This is what he means,’ Arakhin 21b, it appears that Tosafot did not mean what the Taz says. Rather, in their view the renter has no proprietary interest in the rented object at all. It seems they mean that the lessor subjects part of the house to the renter in return for the renter’s money, and the renter thereby has a lien-like right (which is somewhat akin to acquisition), and this is what the acquisition of rental applies to.[9]

Now, Tosafot, s.v. ‘Thus they enacted,’ Bava Metzia 99a, write that the enactment of meshikhah for guardians was enacted with respect to the possibility of retraction, not with respect to the liabilities of guarding. They then prove the same with regard to both renter and borrower: the meshikhah enacted there is only for retraction. Their proof is based on the fact that with respect to the liability of guarding in the case of a borrower, Torah law itself requires meshikhah. At first glance this is difficult: why distinguish between borrower and renter on the one hand, and the unpaid and paid guardians on the other? And further, how do Tosafot know that a borrower truly requires meshikhah by Torah law?

It might seem that Tosafot mean that borrower and renter have a proprietary interest in the body of the object, and therefore meshikhah is required by Torah law to acquire the body of the object. There was thus an initial thought that the enactment of meshikhah in their case concerns the liabilities of guarding (for with respect to retraction there would no longer be room to enact anything, since they have already performed meshikhah and acquired the object and its liabilities by Torah law). Tosafot therefore come to teach that even for borrower and renter the enacted meshikhah concerns retraction. According to this understanding, that Tosafot would hold that borrower and renter do have a proprietary interest in the body of the object — against the Tosafot cited above.

But it seems this is not so. For, if so, it would follow that were it not for the enactment there would be a possibility of retracting from the contract even after performing meshikhah on the object and acquiring it by Torah law — which is difficult (though according to some medieval authorities in that sugya that indeed seems to follow). It therefore seems better to explain that Tosafot intend precisely to reject the initial thought that borrower and renter have a proprietary interest in the object, and for that very purpose they bring their proofs. According to this, in conclusion Tosafot indeed hold that even borrower and renter do not have a proprietary interest in the object, exactly like the unpaid and paid guardians, and this fits Tosafot’s consistent view elsewhere.

We also learn from here that in this respect the renter is like the borrower. This also follows from the sugya in Avodah Zarah, which explicitly equates borrower and renter, and concludes that in both cases the guardian has no proprietary interest in the object. By contrast, it should be noted that the sugya of price fraud in Bava Metzia, which speaks of acquisition in the object, certainly deals only with the renter and not with the borrower (for the borrower pays nothing). It is not clear what that sugya’s position is regarding borrowing. See below, when we discuss the views of the medieval authorities who ruled like the sugya in Avodah Zarah.

We thus learn that Tosafot’s view in a number of places is that the renter has no proprietary interest in the body of the object or the land, and uses it only by the owner’s permission.[10]

Now, from the language of the Talmud in Kiddushin 47b (and Bava Metzia 99a) it appears that the borrower acquires the object, as it says there:

Rabbi Huna said: If one borrows an axe from another and splits wood with it, he acquires it; if he did not split wood with it, he does not acquire it.

However, the continuation of the passage there explains that ‘he acquired it’ means not for liability in unavoidable accidents, but only that neither side can retract. Similarly, in Ritva on Kiddushin there, citing his teacher the Ra’ah, we find the view of Tosafot: renter and borrower have nothing in the body of the object, and the enacted meshikhah serves only to acquire the lien (and this is a rabbinic enactment, since ordinarily meshikhah acquires only the bodies of movable items).[11]

Later authorities brought proof for this view from the Mishnah in Bikkurim 1:2, which distinguishes between a sharecropper and a leaseholder, who do not bring first-fruits, and one who acquires the corpus for usufruct, who brings but does not recite (according to Reish Lakish, Bava Batra 136b). At first glance, this proves that the renter does not acquire the body of the object for its usufruct.

Now, all these views face a difficulty from the Tosefta, chapter 9 of Bava Metzia (cited by Tosafot above in Arakhin), according to which the renter can consecrate the object and the lessor cannot. That seems to imply that the renter has a proprietary interest in the object. True, as Tosafot note, this is the opposite of what the Talmud itself says in Arakhin there. Tosafot therefore distinguish, in light of the Jerusalem Talmud in Pesachim, between a renter who paid his rent in advance — in which case he has a proprietary interest in the body of the object for its usufruct — and a renter who pays at the end (as rentals are generally paid at the end), in which case the renter has acquired nothing; and it is this latter case that the Talmud in Arakhin discusses.[12]

According to this, even on the view of Tosafot and those who follow them, that the renter has no proprietary interest in the object, this is only when he pays at the end, such that the payment is consideration for use. But if he paid the entire sum at the beginning of the period, then the payment is purchase money for the corpus-for-usufruct, and the right to use follows automatically (he is using his own property). Even so, Tosafot apparently held that with respect to price fraud there is no difference, and therefore the sugya in Bava Metzia 56 also deals with rental that does not effect acquisition (for they did not resolve the contradiction between the sugyot by distinguishing according to the time of payment). So too it appears in Tosafot, s.v. ‘You may not,’ Bava Metzia 79b.[13] This distinction, however, is not accepted by all the medieval authorities.[14]

The view of Maimonides and those who follow him: rental does effect acquisition

Now, Nimmukei Yosef on Bava Metzia 56b also cites Tosafot’s difficulty from the sugya in Avodah Zarah. But he answers that rental acquires the corpus for usufruct, while ownership of the corpus itself remains with the lessor. The law regarding selling a house to a gentile in the sugya in Avodah Zarah depends on ownership of the corpus (which remains that of the Jewish lessor), and therefore it is forbidden to lend or rent it to him, even though the gentile does in fact acquire the corpus for usufruct.

So too it appears in Nimmukei Yosef on the sugya of meshikhah for guardians (Bava Metzia 99a), where he writes that since renter and borrower use the actual body of the vessel, it is considered like a sale for the day, and it is fitting that their meshikhah should acquire for them, since they acquire the body of the thing. Indeed, this seems to follow from the language of the sugya there, which compares acquisition by meshikhah in rental to acquisition by meshikhah for buyers in a sale.[15]

Nachmanides wrote likewise (printed among the omitted passages to Rashba on Ketubot 59a) that the renter can transfer, subrent, and consecrate, because the corpus-for-usufruct of the object is literally his property. This is an additional novelty: the renter can even transfer the object in his possession.[16]

And so too Kovetz He’arot there cites Rivash, sec. 510, that if a person ends his rental in the middle of the term, it is not enough simply to vacate the house; he must reconvey it to the lessor. Apparently, this proves that he has a proprietary interest in the house, perhaps even full ownership of the corpus itself (and not only the corpus for usufruct).[17]

Maimonides too explicitly states in several places that the renter acquires the corpus for usufruct (see Laws of Borrowing 1:5; Laws of Sale 23:1-8; Laws of Rental 6:5 and 7:1, among others). However, he disagrees with Nachmanides and holds that this ownership does not give the renter the right to subrent the house onward (see Laws of Sale, chapter 23), unlike one who purchases for a fixed term, who can sell his rights.[18]

Yet in chapter 9 of the Laws of Terumot, law 7, Maimonides rules as halakhah the law that a priest who rents a cow from an Israelite may not feed it terumah, which seemingly indicates that he has no proprietary interest in it. So too regarding renting a house or animal to a gentile, Maimonides rules (see Laws of Sabbath 20:3) that this is forbidden, and this also seems to indicate that the renter has no proprietary interest in the object.

It may be that the explanation is what Maimonides writes in his Commentary on the Mishnah, Terumot 1:9: even though he has a proprietary interest in it, this is not called property acquired with his money (and not because it is not his for its usufruct).[19] But regarding renting a house or animal to a gentile, Maimonides also rules that it is forbidden, and according to this we would have to say that there too the matter depends on property acquired with his money — which is implausible, since that is a criterion relevant only to terumah. It therefore seems that feeding with terumah and renting to a gentile depend on ownership of the corpus, and the corpus belongs to the lessor (as in Nimmukei Yosef cited above). By contrast, for the law of price fraud all that is required is that there be something bought and sold, and for that ownership of the corpus for usufruct is sufficient; this too can be called ‘a sale for the day.’ Thus Maimonides follows the Nimmukei Yosef: the renter has ownership of the corpus for usufruct, and the lessor has ownership of the corpus itself.

We should note that the Mordechai also belongs to this camp, except that he agrees with Nachmanides that for this purpose the renter is fully like a purchaser, and can even transfer his rights to a larger family, just like one who purchased for a fixed term (see Darkhei Moshe, Hoshen Mishpat sec. 316, sec. 1).

A note on the practical implication discussed by the later authorities

It should be noted, contrary to the view of the later authorities mentioned above, that one need not infer from such a position that the renter must pay the entire sum immediately, or that if the house collapses he must pay the entire sum rather than only for the time he already lived there. The proof is that we find here a complete contradiction from one extreme to the other: Ritva on Bava Metzia 103a cites Nachmanides as ruling that if the house collapses, the renter pays only for the time he already lived in it. That is apparently contrary to Nachmanides’ view (which we saw in the omitted passages to Rashba cited above) that the renter has a proprietary interest. And conversely, Ritva there cites the Ra’ah as ruling that he pays the entire amount, and this too is contrary to what we saw above from the Ra’ah.[20] There are many other proofs that there is no necessary connection between the form of payment and the question whether rental effects acquisition, but this is not the place to elaborate.[21]

The explanation is that one can think the renter has a proprietary interest, but that the payment is determined by the uses that this proprietary interest enables. The acquisition itself is effected not by money, but by meshikhah (for movables) or hazakah (for land). Conversely, one may think that the renter has no proprietary interest, but that the payment must nevertheless be given for the whole transaction, because he purchased all the uses — especially if what he acquires is the lien that guarantees them, as we saw above in Tosafot’s view.

C. The difference between renter and borrower

Introduction

Until this point we assumed that the question whether the renter has a proprietary interest in the rented object applies equally to borrower and renter, and the answers were given in relation to to both together. We saw this from the Talmud in Avodah Zarah and also from the medieval authorities. True, we noted that the sugya in Bava Metzia speaks only of the renter, since the borrower pays nothing and price fraud is inapplicable. But it is certainly possible that this is only a technical constraint, and that the Talmud’s conclusion there, that rental does not effect acquisition, is equally valid with regard to the borrower.

Now, in Rashi on the sugya in Avodah Zarah, we find that precisely in that sugya (which explicitly links rental and borrowing) he says things that point toward a distinction between them. The same can be seen in Rashi in other sugyot in Bava Metzia. From these remarks we shall try to characterize the rental contract as opposed to borrowing, and return to the question posed at the outset: why is the renter a fourth guardian?

Rashi’s view in Avodah Zarah 15

As noted, the sugya in Avodah Zarah 15a does not distinguish between rental and borrowing. Throughout, it speaks of rental and borrowing acquiring, or not acquiring, in one breath. Why is there, in truth, a side in the Talmud that both rental and borrowing effect acquisition? Presumably because in both of these contracts the guardian has rights of use, unlike the unpaid and paid guardians (who, if they use the deposited object, are guilty of misappropriation). When someone has rights of use, there is room to think that these are proprietary rights, and not merely permitted use. Therefore this possibility arises both with regard to borrowing and with regard to rental.

Precisely against that background, and precisely in that sugya, we must consider why Rashi, when he comes to explain the Talmud’s initial assumption that the borrower has acquisition, writes as follows:

Borrowing acquires it—for during the days of the borrowing, since he is liable for unavoidable accidents, it is considered his.

Rashi explains the view that borrowing effects acquisition on the basis that the borrower is liable for unavoidable accidents. At first glance he ignores the fact that in the very same sentence the Talmud says (see the citation above): Borrowing acquires, and renting acquires — that is, rental too effects acquisition, even though in rental there is no liability for unavoidable accidents.[22]

Beyond that, it is not clear why the fact that he is liable for unavoidable accidents means that he has a proprietary interest in the object. Why can there not be liability for unavoidable accidents on an object that is not his, merely as a guarantee to the lender, who did him a favor and allowed him to use the object?[23]

Presumably Rashi means to cite liability for unavoidable accidents as an indication of the proprietary interest that the borrower has in the object. If the borrower is liable for unavoidable accidents, then he apparently has a proprietary interest in the object. We know other examples of such an indication. For example: a robber has an acquisition in order to render him liable for unavoidable accidents; a creditor holding collateral who is liable for unavoidable accidents (according to some views) acquires it by virtue of the law of R. Isaac. Our claim here is that the borrower belongs to that same list: his liability for unavoidable accidents is based on a proprietary interest that he has in the object.

Already at this stage we should note that our point can be understood in two different ways: (1) the proprietary interest is given to him in order to render him liable for unavoidable accidents (as with a robber); (2) the proprietary interest exists for other reasons (because of the use rights, or because of the contract with the owner), and liability for unavoidable accidents is a consequence of that proprietary interest. Below we shall explain this more fully.

In any event, these words of Rashi imply that the borrower’s liability for unavoidable accidents is not part of a guardian’s responsibility, or of his guarding obligations. The borrower’s liability for unavoidable accidents is a result of his ownership. Everything that happens to the object happens to the borrower, not to the owner, and yet the obligation to return the object to its owner remains in force. That is what generates the borrower’s liability for unavoidable accidents.

Rashba on Bava Metzia 36b

Rabbi Lichtenstein, in the first chapter of the lectures cited above, brings support for this point from Rashba on Bava Metzia 36b. The Talmud there proposes an explanation for the liability of a guardian who was negligent with a cow, it went out to the marsh, and died there in its normal manner. This is a case of ‘it began in negligence and ended in unavoidable accident,’ and the Talmud says about it:

…It need not be said according to the one who holds that if it began with negligence and ended with an unavoidable accident he is exempt—that he is exempt; rather, even according to the one who holds he is liable, here he is exempt. What is the reason? For we say: what difference does the Angel of Death make whether here or there…

Rashba asks from here against the borrower’s liability. If the rationale ‘what difference does the angel of death make whether here or there’ is considered a ground of exemption, why is the borrower liable for unavoidable accidents? Apparently this consideration could be raised in his case as well. Rashba explains that the borrower’s liability for unavoidable accident is not the result of any blame, and therefore the argument ‘what difference does the angel of death make whether here or there’ is irrelevant as a ground to exempt him. Rashba also rejects the possibility of a pure responsibility-liability unconnected to negligence in guarding, something like an insurance company. He grounds the matter in ownership:

And a borrower is liable even if it died naturally, because since all the benefit is his, the Merciful One placed it entirely in his domain as though it were his—except for death due to labor…

And that is exactly our point. Rabbi Lichtenstein there cites Rabbi Soloveitchik, who compared this to a robber (who is also liable for unavoidable accidents; see Bava Kamma 105b) and to one who holds collateral (who according to some medieval authorities is likewise liable for unavoidable accidents; see Rashi and Ba’al HaMa’or on Kiddushin 8b). He explains that the foundation of the liability is like that of a person who bought an object and paid for it, and after leaving the store the object was destroyed in an unavoidable accident. He obviously cannot return to the store and ask for his money back, because the object was lost to him and not to the storekeeper. The same is true of the borrower. His liability is not due to negligence in guarding or responsibility for unavoidable accident. Rather, he is under an obligation of return; and if the object is lost in an unavoidable accident, it was lost to him himself.

What is the renter’s law according to Rashi?

It is fairly clear that Rashba’s remarks apply only to the borrower. The renter is not liable for unavoidable accident, and it therefore seems that his liabilities are the result not of ownership, but of obligations of guarding, like the other guardians.

Let us now return to Rashi in Avodah Zarah. These remarks raise a question about the Talmud’s conclusion there. We saw that at the beginning of the sugya the Talmud says: Borrowing acquires, and renting acquires. But now we see that according to Rashi the reasons are different: in rental, because of the acquisition of rights of use; in borrowing, because of liability for unavoidable accidents. The Talmud’s conclusion is: Rather, conclude from this that rental does not effect acquisition. The proofs brought in the sugya are also only about rental (from the case of a cow rented to a priest). What, then, is the law of borrowing? According to Rashi’s explanation, it seems that the Talmud’s conclusion concerns rental alone, whereas with regard to borrowing the conclusion is that borrowing does acquire, for it is obvious that the borrower has liability for unavoidable accidents. The proof from feeding terumah to a cow rented by a priest says nothing about the argument that borrowing acquires by virtue of liability for unavoidable accidents. In rental there is no such liability, and therefore rental does not effect acquisition.

It may be that this is exactly why Rashi needed a different explanation regarding borrowing. Already at the beginning of the sugya he wished to hint to us that borrowing differs from rental. Borrowing does indeed include the rationale relevant to rental as well (for the borrower too may use the object), but that rationale is rejected as a basis for acquisition of the object (by force of the proof from an animal rented by a priest), and therefore, as halakhah, rental does not effect acquisition. But in the context of borrowing there is also the reasoning that the object is vested in him in order to render him liable for unavoidable accidents, and that may remain in force even in the conclusion.

If so, according to Rashi the conclusion of the sugya in Avodah Zarah is that rental does not effect acquisition, but borrowing does — and the proof is that the borrower is liable for unavoidable accidents.

Two further proofs for distinguishing between borrower and renter

A similar distinction appears in the Jerusalem Talmud, Shevuot, chapter 8, law 1. The Jerusalem Talmud discusses a case where the object was stolen from the guardian’s house, the guardian paid its value to the owner, and the thief was then found. The guardian acquires the double payment if the object was ‘stolen from the house of the man.’ According to the baraita there, this verse excludes the borrower:

It was taught: “And steals from the house of the man”—not from the house of the borrower. And should one then say: “And steals from the house of the man”—not from the house of the paid custodian and the renter? Since its safekeeping rests upon him, it is as though it is upon him.

The baraita teaches that the verse and steals from the house of the man excludes the borrower, whereas the renter is not excluded because the duty of guarding rests upon him. Tzafnat Pane’ach on Bava Kamma 11a and Ketzot HaHoshen at the beginning of sec. 349 explain that the borrower is the owner of the object, and therefore this is not a theft from the depositor’s house. Tzafnat Pane’ach explains that the borrower’s liability for unavoidable accidents derives from his obligation of return, and is not payment for negligence in guarding, unlike the other guardians. By contrast, the renter follows the same law as the other guardians: he is obligated in guarding, while the object belongs to the owner. If so, the renter is guarding another’s property, not using his own. According to this view, rental does not effect acquisition, while borrowing does.[24]

In the Mekhilta too we find an obscure exposition that separates the borrower from the other guardians (Masekhta deNezikin, parashah 16, s.v. ‘And when a man borrows’):

“And if a man borrows from his fellow”—Scripture separated the borrower from the general category of custodian and stated it as an independent matter.

The author of Mareh HaPanim on the Jerusalem Talmud cited above brings this exposition and links it to our issue.

In the next paragraph we shall see a similar possibility in Tosafot on Yevamot 66.

The baraita in Yevamot 66b

In Yevamot 66b it emerges that the baraita concerning a priest who rented a cow from an Israelite is composed of two parts. The first clause, which is also cited in Avodah Zarah, deals with an Israelite who rented a cow from a priest and a priest who rented a cow from an Israelite. As halakhah, a priest who rents an animal may not feed it terumah, because rental does not effect acquisition. The latter clause of the baraita, which appears only in Yevamot, deals with an Israelite who took a cow from a priest on valuation terms and a priest who took a cow from an Israelite in that manner; there it is ruled that a priest who took a cow from an Israelite may feed it terumah.

Rashi there explains that the case is one of a priest who rented an animal from an Israelite and stipulated that he would assume liability if anything happened to it or to its value. Tosafot, s.v. ‘But,’ object that this would be forbidden as usury, and answer that the case is one of borrowing.

The simple meaning of Tosafot is that it is a case of borrowing because in borrowing there is no monetary payment, and therefore no usury problem. So too Maharri Korkos understood them on Maimonides, Laws of Terumot 9:8. But in the Tzafnat Pane’ach cited above (and parallels), Tosafot are understood differently: they established the case as a borrower because in borrowing the cow is considered to belong to the priest, unlike rental.[25] That is, Tzafnat Pane’ach understood Tosafot to mean that when a priest borrows an animal from an Israelite, he may indeed feed it terumah, for borrowing effects acquisition, unlike rental, which does not.

The Tzafnat Pane’ach continues and says that Maimonides himself also held this way. In all those passages he explains that Maimonides (both in Laws of Terumot 9:8 and in the Commentary on the Mishnah to Terumot 11:9) understood the Talmud as speaking of a borrower, because borrowing effects acquisition. If so, according to the Tzafnat Pane’ach, Maimonides and Tosafot agree with the distinction we saw in Rashi on Avodah Zarah: the Talmud’s conclusion is indeed that rental does not effect acquisition, but borrowing does.

Clarifying the views of the medieval authorities according to the author of Tzafnat Pane’ach

However, the words of the Tzafnat Pane’ach are very difficult, both with respect to Tosafot and with respect to Maimonides. In Tosafot it appears quite clearly the opposite of what he says. Tosafot object that to arrange matters in this way would constitute usury, and answer that it is a case of borrowing. Clearly, then, in their view, were it not for the usury problem the law of the renter would be exactly like that of the borrower; thus it is clear that they held that borrowing likewise does not effect acquisition (Tosafot here remain consistent with their general view throughout the Talmud, as we saw above, that rental does not effect acquisition).

The Tzafnat Pane’ach is also difficult to reconcile with Maimonides. A look at Maimonides’ language, both in the Laws of Terumot and in the Commentary on the Mishnah cited above, shows that he never offered such an interpretive restriction in the Talmud (that the case is one of borrowing). In general, the whole discussion of Tosafot concerns the latter clause, which deals with one who took the cow on valuation terms, not with one who rented a cow. And if there is indeed a difference between renter and borrower, why did these authorities not also write in the first clause that it deals only with a priest who rented a cow from an Israelite, and not with borrowing?

It seems from the words of the Tzafnat Pane’ach that he understood those authorities as reading the baraita itself as distinguishing between borrower and renter: the first clause deals with the renter, and the latter clause (concerning valuation of the cow) deals with the borrower. Indeed, from Maimonides it appears that he did not read the baraita like Rashi (who understood the latter clause as speaking of a renter who accepted liability for depreciation of the animal’s value). Thus he writes (Laws of Terumot 9:8):

If an Israelite appraised a cow from a priest in order to fatten it and share the profit, he may not feed it terumah, even though the priest has a share in its increase. But if a priest appraised a cow from an Israelite in order to fatten it, even though the Israelite has a share in the profit, since the animal itself belongs to the priest, for he accepted it upon himself, this one may feed it terumah.

It seems that Maimonides understands the latter clause as referring to a priest who took the cow under a profit-sharing arrangement, that is, an investment partnership. The priest took the cow from the Israelite in order to fatten it, and the gain would be divided between them. As is known, such an arrangement is half loan and half deposit. Here the matter is not money, and therefore it is not literally half a monetary loan. Rather, apparently half the cow is lent to the agent in order to improve it, and half remains in his hand as the owner’s deposit, such that the profits are split between the borrower and the depositor as in a standard partnership investment. However, from Maimonides’ language it appears that the whole thing is lent to the priest, and only the profit is divided between them. Maimonides writes in the latter clause that when a priest takes a cow from an Israelite in this way, the Israelite has a share in the increase, but the body of the cow belongs to the priest; therefore he may feed it terumah. The reason is that this cow is lent to the priest, and borrowing effects acquisition. From here it is proved that, according to Maimonides, the borrower indeed acquires the body of the object. In the reverse case, when an Israelite takes such a cow from a priest, for that same reason the body of the cow belongs to the Israelite (while the profits are divided between them), and therefore he may not feed it terumah.

If so, according to Maimonides what is the difference between the first clause of the baraita (ruled in law 7 there) and the latter clause (ruled in law 8 there)? The first clause is rental, and rental does not effect acquisition (at least for the purpose of feeding terumah); the latter clause is borrowing, and borrowing does effect acquisition. We thus learn that the Tzafnat Pane’ach was correct: according to Maimonides, borrowing effects acquisition, and only rental does not. The reason Maimonides did not need an interpretive restriction for the baraita itself is that the baraita is itself divided between its first and latter clauses in order to teach this very point: the first clause deals with rental and rules that rental does not effect acquisition, while the latter clause deals with borrowing and rules that borrowing does effect acquisition.

And what about Tosafot? It is possible that the Tzafnat Pane’ach understood their answer precisely in the same way as Maimonides. In their initial assumption they understood, like Rashi, that the first clause deals with a valuation arrangement of a renter, and that if not for the usury problem rental would have been like borrowing. But Tosafot rejected that understanding because of the difficulty from the laws of usury, and therefore established that the latter clause speaks of a borrower — and perhaps they meant thereby what Maimonides meant. If so, even according to Tosafot it is possible that the understanding of the latter clause is that the borrower acquires, and therefore the latter clause of the baraita rules that the priest may feed it terumah. This would also make the words of the Tzafnat Pane’ach fit Tosafot (though there this is by no means necessary, as noted).

Still, with respect to Tosafot this is hard to say, for we have seen that their position everywhere is that rental does not effect acquisition, and that they liken borrowing to rental. And with respect to Maimonides too this requires clarification, for we saw that he writes that both borrowing and rental effect acquisition, whereas here the Tzafnat Pane’ach wishes to say, in Maimonides’ view, that borrowing does effect acquisition while rental does not. See below for an explanation of Maimonides’ view on this point.

The law of appraisal for the borrower in Bava Kamma 11a, and the Talmud’s comment in Sanhedrin 72a

The Tzafnat Pane’ach ties this understanding to the sugya in Bava Kamma 11a concerning appraisal for the borrower (and the same is found in Rabbi Lichtenstein’s lectures there). The sugya there explains that we do not appraise for a thief or a robber: if the object deteriorated while in the thief’s or robber’s possession, they must return money equal to the value of a full object, and the broken remains are theirs. By contrast, for damages we do appraise (for it is said, and the carcass shall be his). The Talmud then wonders: do we appraise for the borrower? The Tzafnat Pane’ach explains that the question is whether the borrower’s liability for unavoidable accidents is a duty to return the borrowed item, or whether it is a kind of damages-liability (like the liability for theft and loss in a paid guardian). If the duty is to return the object, then the guardian must return the object, and we merely appraise the depreciation so that he makes up the missing value in money. But if the duty is to compensate the owner for the loss of the object by way of liability, then the duty is to pay full price and the broken remains stay with the borrower.[26]

As halakhah, we rule that we do appraise for the borrower — meaning that the borrower’s duty is to return the borrowed object, not to pay by virtue of a separate liability for unavoidable accidents. The Tzafnat Pane’ach proves the same from the language of Sanhedrin 72a, which, in the course of discussing an object in the possession of a burglar who entered through a tunnel, says:

But that is not so. When the Merciful One placed it in his domain, that was only regarding liability for unavoidable accidents; but regarding acquisition, it remains in the domain of its owner, just as with a borrower.

And Rashi explains there:

But that is not so—this is Rava’s final conclusion, and “but that is not so” means that Rav’s reasoning here is not persuasive; for even though where he must pay its value they bear liability for it, that does not mean it is acquired to him to be like his other property. For when the Merciful One placed it in his domain and obligated robbers to pay, that was only regarding payment, so that he cannot say to him, “They were lost through unavoidable accident.” For we find with custodians that Scripture obligates them to swear that they did not lay a hand on their fellow’s property—but if he did lay a hand on it, then even if he claims it died or broke, that is, an unavoidable accident, he is liable, because by misappropriating it he became a robber with respect to it.

But regarding acquisition, he does not acquire it—if the items still exist with him and he wanted to withhold them, saying, “I will return the money to you,” he has not acquired them, and he must return them intact, as it is written (Leviticus 5), “and he shall restore the stolen item.” And the reason that when they are lost through unavoidable accident he returns money and cannot say to him, “They were lost through unavoidable accident,” and be exempt, is because a robber is no less than a borrower: just as, because all the benefit is his, the Merciful One placed them in his domain to require payment if lost through unavoidable accident, while so long as the item exists it is returned intact—so too here, since all the benefit is his, he is obligated to pay.

Thus Rashi explains that the acquisition in the case of the borrower is meant to ensure that the borrower cannot say to the depositor that the object was lost in an unavoidable accident. If the object suffered an unavoidable accident, it happened to the borrower himself, and the obligation to return therefore remains in force; hence he is liable even in unavoidable accidents.

The Tzafnat Pane’ach explains that from here Maimonides derived his conclusion that the borrower acquired the object. The duty to return the object means that the object is acquired to the borrower, and at the end of the borrowing period he has a duty to return it to the owner. If the object is lost in an unavoidable accident, the borrower must still pay for it — not because we impose upon him a liability for unavoidable accidents, for the Torah generally exempts in unavoidable accident, but because a duty to return the object rested upon him and he has still not fulfilled it.

We should note that Tosafot, s.v. ‘We do not appraise,’ in that sugya explain exactly the opposite. Tosafot say that a thief and a robber are not appraised because they acquired the object from the moment of theft or robbery. Therefore they become liable immediately in money, and we do not appraise the object. By contrast, one who did not acquire the object becomes liable only when the object deteriorates, and therefore he returns only the difference. As for the borrower, in the initial assumption they thought that we do not appraise (like a thief and robber), and Tosafot explain that this is because the borrower acquired the object, since he is liable for unavoidable accidents. But as halakhah it seems from Tosafot that because we do appraise for the borrower, it is clear that the borrower did not acquire the object.

It is entirely possible that Tosafot remain consistent with their view that, as halakhah, the borrower did not acquire, just like the renter; we already noted above that the Tzafnat Pane’ach is difficult to reconcile with Tosafot. But Maimonides, who holds that borrowing does effect acquisition even with respect to terumah (where even according to Maimonides rental does not effect acquisition), may have learned the sugya of appraisal for the borrower according to the Tzafnat Pane’ach’s explanation. As noted, the Tzafnat Pane’ach cites another dispute touching the same point, in Bava Kamma 112a, where two versions appear regarding a father who left his children a borrowed cow; but this is not our concern here.

The Tzafnat Pane’ach likewise explains the Jerusalem Talmud at the beginning of chapter 8 of Shevuot, which we cited above, according to which the borrower does not receive the double payment because the object is his, whereas the renter and paid guardian do receive it because its safekeeping rests upon him. His meaning is that in the case of the renter the object was stolen from its owner, because rental does not effect acquisition. The renter is not an owner of the object, but only obligated to guard it, and his liabilities arise from that responsibility of guarding. By contrast, the borrower acquired the object, and his liabilities arise from the duty of return.

Back to Rashi in Avodah Zarah

The distinction between renter and borrower brings us back to Rashi’s words in Avodah Zarah 15. As we have seen here, liability for unavoidable accidents is not what causes ownership, but serves as an indication that the borrower owns the borrowed object. Liability for unavoidable accidents is not a punitive consequence of negligence in guarding, but the implementation of the duty to return the object to its owner. Since the object is acquired to the borrower, whatever happens to the object is borne by the borrower. If it is damaged, stolen, or lost in an unavoidable accident, then it is the borrower’s own object that was damaged, stolen, or lost; therefore, when the time for return arrives, the borrower must return the object or its value to the depositor (if the object was lost in an unavoidable accident, he must return money in its place).

This may also explain the mechanism that obligates the borrower in unavoidable accidents, even though throughout the Torah a person is generally exempt in unavoidable accident. Since this is not a punishment, or punitive liability, but a duty to return an object to its owner, there is no exemption here on the ground of unavoidable accident (and this is explicit in Sanhedrin 72a and Rashi there). Even if the borrower failed to return the object because of unavoidable accident, he still did not carry out the duty to return it, and therefore he must return its monetary value.[27]

It should be noted that according to the Tzafnat Pane’ach, this apparently is not the explanation, for a robber too is liable for unavoidable accidents, yet with respect to a robber we do not appraise, unlike the borrower. If so, according to the Tzafnat Pane’ach, in the case of a robber the duty is indeed one of liability, and nevertheless he is liable for unavoidable accidents.[28]

In any event, in Rashi on the sugya in Avodah Zarah it seems that this is what he meant: liability for unavoidable accidents is an indication that the borrower acquired the object (and this also appears from Rashi’s language in Sanhedrin 72a, cited above). As we noted, according to this it seems that in the Talmud’s conclusion rental did not acquire but borrowing did. The proof brought by the Talmud there, from the baraita concerning a priest who rented a cow from an Israelite, concerns rental alone and not borrowing (as the Tzafnat Pane’ach understood Tosafot in Yevamot 66 above).

From this it would seem that Rashi also sides with Tosafot in holding that rental does not effect acquisition, except that in his view borrowing does. And so too follows from the Jerusalem Talmud at the beginning of chapter 8 of Shevuot cited above. But this is somewhat difficult in terms of the plain meaning of the Talmud there, for the Talmud links borrowing and rental together. Below we shall explain Rashi differently.

Back to the view of Maimonides and his school

Conversely, we saw above that Maimonides holds that both rental and borrowing effect acquisition, and that the conclusion of the sugya in Avodah Zarah that neither borrowing nor rental acquires receives a special explanation. If so, we now need to explain why the Tzafnat Pane’ach writes that according to Maimonides borrowing acquires but rental does not.

We have already seen that Maimonides apparently held, like the Nimmukei Yosef, that the renter and borrower have an ownership of the corpus for usufruct, while the corpus itself remains the owner’s. Only within this framework can we try to understand the difference between borrower and renter according to Maimonides. We shall do so in the next chapter, where we will examine the nature of the differences between them.

A marginal note: the opposite possibility

We find in Ba’al HaItur (see Mishnah Berurah, sec. 649, sec. 14, and in Pitchei Choshen, Rental, chapter 1, note 18) a possibility completely opposite to what we have said so far: namely, that specifically rental effects acquisition, while borrowing does not. For with respect to a borrowed lulav, the halakhah is that one does not fulfill the obligation where the Torah requires that it be ‘yours.’ From this it follows that borrowing does not effect acquisition. Yet Ba’al HaItur is uncertain whether a rented lulav is effective, and whether one can thereby fulfill the obligation. This suggests that he entertained the possibility that borrowing does not acquire, whereas rental does.

Indeed, it requires explanation why a borrowed lulav is not considered ‘yours’ if borrowing does effect acquisition. A number of later authorities wrote that a rented lulav likewise is not considered ‘yours,’ no worse than a borrowed one, even if rental does effect acquisition. So writes the Mishnah Berurah there in the name of Olat Tamid; see also Orhot Chaim, sec. 649, in Maharsham’s glosses there, in the name of several later authorities.

D. Renter and borrower: two kinds of acquisition for a term

Rashi’s distinction in Bava Metzia 80b

The Mishnah in Bava Metzia 80b states: ‘All artisans are paid guardians.’ And the Talmud there says:

Shall we say that our Mishnah is not in accordance with Rabbi Meir? For it was taught: How does a renter pay? Rabbi Meir says: like an unpaid custodian; Rabbi Yehuda says: like a paid custodian. — You may even say it is Rabbi Meir, for by virtue of that benefit—that he bypasses everyone else and rents it to him—he assumes the status of a paid custodian. — If so, then a renter too, by virtue of that benefit—that he bypasses everyone else and rents it to him—should assume the status of a paid custodian! Rather, you may even say it is Rabbi Meir: by virtue of that benefit—that he gives him a little extra—he assumes the status of a paid custodian. But in the case of a renter as well, are we not dealing with a case where he gives him a little extra? Rather, you may even say it is Rabbi Meir: by virtue of that benefit that he holds the rental payment, so that he does not need to go in and out for money, he assumes the status of a paid custodian. And if you wish, say instead in accordance with the version that Rabbah bar Avuha reversed and taught: How does a renter pay? Rabbi Meir says: like a paid custodian; Rabbi Yehuda says: like an unpaid custodian.

The Talmud explains that the Mishnah follows R. Yehuda, and at the beginning it assumes that the artisan is like the renter. It then explains that the artisan differs from the renter, and therefore even according to the view that the renter is an unpaid guardian, the artisan is a paid guardian because of the benefit that he ‘holds his wages.’ According to Rabbah bar Avuha, however, the initial assumption remains, that the artisan is like the renter.

Rashi there, s.v. ‘R. Yehuda says,’ writes as follows:

Rabbi Yehuda says: like a paid custodian—since it is with him for his benefit, even though he pays the wages for its work, he is a paid custodian; for if he paid no fee, he would be a borrower and liable for unavoidable accidents. Now that he pays him compensation, he is not a borrower, and he is a paid custodian.

Rabbi Meir says: like an unpaid custodian—because he is paid for his labor and does not receive wages for his guarding. And an artisan is comparable to a renter, for the object was with him for the benefit of the wages of his craftsmanship, but he does not receive wages for guarding, only wages for the work.

Rashi’s words require explanation. At first glance he means to say that we count relative deficiencies: one who has the object for his own benefit is considered a borrower. But the renter pays for the rental, and so his liability for unavoidable accidents is reduced, and he receives the laws of a paid guardian. It is not clear why such ‘bookkeeping’ should determine the laws of guarding, or why the result should be specifically the laws of the paid guardian.

Further, Ma’ayanai HaChokhmah asks there that according to Rashi’s view one should have the same difficulty even according to R. Yehuda: why is the artisan a paid guardian? The artisan is not like the renter, for if the renter paid no rent he would become a borrower, whereas the artisan, even if he did not do his work, would still be a guardian who receives payment and would therefore be a paid guardian. If so, since he does work for pay, according to Rashi’s ‘counting of deficiencies’ he should be considered only an unpaid guardian.

Explaining Rashi

It seems that Rashi’s view can be explained as follows: the renter is not a paid guardian, but rather a borrower who is exempt from unavoidable accidents. According to this, it is only accidental that the renter’s laws come out the same as those of a paid guardian, since a borrower without liability for unavoidable accidents happens, from a halakhic standpoint, to resemble a paid guardian. The reason he is like a borrower is that he too has permission to use the object, and that permission places the object in his domain. However, since the renter pays, he is exempt from unavoidable accidents.

The explanation is based on what we saw above in Rashi’s view, that the borrower is liable for unavoidable accidents because the Torah placed the object in his domain. Above we explained this to mean that the Torah vested the object in him in order to render him liable for unavoidable accidents (as is commonly understood regarding a robber). Already there we noted that there are two ways to understand the relation between ownership and liability for unavoidable accidents, and now we see that the second possibility is the correct one: because he has a right of use, the object is his,[29] and therefore he is also liable for unavoidable accidents. Liability for unavoidable accidents is an indication that ownership exists, not the reason for it. The reason is the right of use.

If ownership were granted only in order to create liability for unavoidable accidents, then with respect to the renter there would be no room to speak of ownership (for he is exempt in unavoidable accidents). But according to the view that rental does effect acquisition, the renter too has ownership, because he too has a right of use (like the borrower). Thus ownership may exist even where there is no liability for unavoidable accidents. The converse, however, is true: liability for unavoidable accidents exists only where there is ownership (just as we saw in the case of a robber; though perhaps there the liability for unavoidable accidents is the reason for the ownership — see above on the sugya in Sanhedrin 72a).

But now a difficulty arises. According to what we explained above, liability for unavoidable accidents is a result of ownership. Since the borrower is under a duty to return the object that belongs to him at the end of the term to the owner, if the object is lost in an unavoidable accident he is not guilty, but still an unavoidable accident is not as though he had acted. At the end of the day he still has not fulfilled his duty to return the object, and therefore he must pay its value. If so, the renter too acquires the object for his uses, and thus he too ought to have a duty to return the object to the lessor. If so, we should also obligate him in unavoidable accidents. Even apart from all our prior arguments, this is difficult for those views holding that rental effects acquisition: why is the renter not liable for unavoidable accidents?

We are therefore forced to conclude that the renter differs from the borrower. The borrower received the object, but at the end of the period he must return it to the lender, and from here comes liability for unavoidable accidents. This can be understood in two ways: (1) borrowing is like a gift on condition that it be returned;[30] (2) borrowing is a permanent gift, but at the end of the period there is a duty of return (which is not a condition of the very validity of the borrowing itself). As halakhah, a gift on condition to return is a gift; certainly a gift with a duty of return at the end, where that duty is not a condition of the gift itself, is also a gift. Therefore the borrower is the owner of the object (borrowing effects acquisition), and he is correspondingly liable for unavoidable accidents.[31]

By contrast, the renter did not receive the object as a gift on condition to return, but as a gift or sale for a fixed term. At the end of the term the ownership lapses, and the object reverts automatically to its original owner. At the end of the period the object becomes, of itself, the property of its original owner once again; therefore if the object is lost in an unavoidable accident, the loss is the lessor’s, and the renter is exempt.[32] I later found that this is also written in Ayelet HaShachar on Bava Metzia 79a, s.v. ‘And it requires consideration’ (regarding the renter’s use of a carcass).

This also explains very well why the Talmud in Avodah Zarah 15a linked the acquisitions of borrowing and rental. The fact that the borrower is liable for unavoidable accidents is an indication that he has ownership, not the cause of ownership. The cause of ownership is the right of use. Therefore, if the borrower has ownership, the renter too has ownership, for both have rights of use.[33]

We can now understand Rashi’s words cited above. The renter has permission to use the object, just like the borrower; therefore the object is acquired to him (at least for its usufruct). However, there is no liability for unavoidable accidents, because he pays for the use; but ownership is certainly present. With respect to the artisan, Rashi wrote that the object is with him for his own benefit (so as to profit from it), and therefore he resembles the borrower and the renter, and is as though the owner of the object. Yet with respect to liability for unavoidable accidents he is like the renter and not like the borrower, because he works on the object and improves it for the owner. This is not an exercise in counting deficiencies, but a substantive analysis: the right of use determines that there is ownership. Liability for unavoidable accidents depends on whether all the benefit is his or not. Hence there is likewise no room to compare him to the unpaid guardian according to R. Yehuda, as Ma’ayanai HaChokhmah had asked.[34]

According to R. Meir, who says that the renter is like an unpaid guardian, it may indeed be that the renter has no proprietary interest in the object. He disagrees with the very premise that the right of use grants ownership. Presumably, according to this view there are two separate transactions between lessor and renter: use in exchange for rental payment, and a contract of deposit with unpaid guarding. This also resolves the difficulty raised by Ma’ayanai HaChokhmah.

We can now also understand Rashi’s position in the sugya in Avodah Zarah in another way. Above we saw that he explains that the borrower acquired the object in order to be liable for unavoidable accidents, which implied that even after the Talmud proved from the priest who rented an animal that the right of use does not constitute acquisition, the borrower still acquires, because in his case there is an additional reason — beyond the right of use — in order that he be liable for unavoidable accidents. Thus the law concerning feeding terumah to an animal rented by a priest would apply only to rental and not to borrowing (as the Tzafnat Pane’ach understood in Tosafot on Yevamot cited above). We noted that this is somewhat forced in light of the language of the sugya, which links borrowing and rental to one another.

But according to what we have explained here, there is no need for that. The borrower’s liability for unavoidable accidents is only an indication that he has ownership, but that ownership exists in the renter as well. The reason for ownership is the right of use (which exists in the renter too), not liability for unavoidable accidents (which exists only in the borrower). Therefore there is no need to distinguish between renter and borrower regarding ownership. As we have seen, there is nevertheless a difference between them in the character of that ownership (whether there is a duty of return or whether the ownership lapses automatically).

If so, Rashi follows the approach of Maimonides: both rental and borrowing effect acquisition. Liability for unavoidable accidents is an indication of the borrower’s ownership, and from it there is consequently proof of the renter’s ownership as well. There is, however, a difference between them regarding the expiration of that ownership at the end of the term: in the borrower it requires a return to the lender, whereas in the renter the lessor’s ownership revives automatically. The practical implication is liability for unavoidable accidents.

The view of Maimonides

We can now also understand Maimonides. We saw that he holds that both rental and borrowing effect acquisition, at least an acquisition of the corpus for usufruct, and therefore price fraud applies to them. The Tzafnat Pane’ach distinguished, regarding feeding with terumah, between rental (which does not acquire) and borrowing (which does acquire), and we said that this distinction must be understood within the framework that Maimonides himself established, namely that both rental and borrowing do acquire.

According to what we have now said, we can understand Maimonides like Rashi: borrowing effects acquisition permanently, except that there remains a duty of return. Such an acquisition (a usufructuary acquisition in perpetuity)[35] is very close to ownership of the corpus, and therefore allows a priest to feed his borrowed animal terumah. By contrast, rental is an acquisition for a term, and that is a usufructuary acquisition that does not allow a priest who rented an animal to feed it terumah.

It thus emerges that in general, according to Maimonides, both rental and borrowing effect acquisition (for example, with respect to price fraud). Regarding feeding an animal terumah, we saw that Maimonides apparently distinguishes between rental and borrowing, ruling that borrowing acquires while rental does not. We explained this through the distinction made by Rashi between the acquisitions of the renter and the borrower: both do acquire, but one is permanent and the other is for a term. Regarding handing an animal over to a gentile, Maimonides forbids both rental and borrowing (see Laws of Sabbath 20:3), and this seems to show that even the borrower does not have complete acquisition in the object; something remains to the owner, and that alone is enough for the gentile’s labor to be considered labor done with the Jew’s animal.[36]

A general note about gratuitous and paid acquisitions: ‘a gift on condition to return’ and ‘a sale for a term’

Usually in the Talmud we find the categories of ‘a gift on condition to return’ and ‘a sale for a term.’ We do not find ‘a sale on condition to return’ and ‘a gift for a term.’[37]

It may be that whenever the transfer is gratuitous rather than for money, it is a transfer in perpetuity on condition to return, and not a transfer for a term; whereas when the transfer is for money, it is a transfer for a term and not a perpetual transfer on condition to return. Therefore, when discussing ‘on condition to return’ in the abstract, the reference is ordinarily to a gift; and when discussing ‘for a term’ in the abstract, the reference is ordinarily to a sale.

The point can be explained in several ways. For example (this explanation was suggested to me many years ago by my friend R. Haim Stand to R.): in a sale, the money determines the duration of the ownership. For the time beyond that, he paid nothing, and therefore he is not the owner of that further period. Thus in a sale for a term, when the term expires the ownership lapses. But in a gift there is nothing to define the duration of the ownership, since no money was paid. Therefore the transfer is perpetual. True, the donor conditions the gift on the recipient’s returning the object at the end of a certain time. Thus what limits the time is the condition, which establishes a duty of return, and not money. Fulfillment of the condition must be by an act of returning, for the object belongs to the recipient in perpetuity and, without return, it does not revert to the owner. By contrast, in a sale for a term, the ownership lapses at the end of the term, and the object automatically returns to its owner; hence there is no duty of return.[38]

Accordingly, it is very clear why in rental there is no duty of return at the end of the period, while in borrowing there is. Rental is for money, and the money defines the period. At the end of the period the ownership lapses and the object returns automatically to the lessor. By contrast, in borrowing there is nothing to delimit the period, and therefore the ownership is perpetual, except that there is a duty of return (whether as a condition or merely as a duty, as discussed above).[39]

Indeed, Maimonides in Laws of Borrowing 1:5 likens rental to sale and borrowing to gift, writing as follows:

If one borrows a vessel from another, or an animal, without specification, the lender may demand its return whenever he wishes. But if he borrowed it for a fixed period, once he has taken possession and acquired the right, the owners cannot reclaim it from him until the end of the borrowing period. Even if the borrower dies, the heirs use the borrowed item until the end of the term. And this is logical: a purchaser acquires the corpus in perpetuity for the money he paid; and one who receives a gift acquires the corpus in perpetuity though he paid nothing; and a renter acquires the corpus for its use until a fixed time for the money he paid; and a borrower acquires the corpus for its use until a fixed time though he paid nothing. Just as a giver, like a seller, cannot retract forever, so too a lender, like a lessor, cannot retract during the term.

These points have additional implications as well; we intended only to draw the reader’s attention to them.

We can now proceed to explain, in light of all that we have said thus far, the four types of guardians.

E. Solving the problem: the renter as a borrower exempt from unavoidable accidents

Introduction

In this chapter we shall explain why the renter is a unique type of guardian. We shall examine the halakhic relation between his laws and those of the borrower, and argue that his resemblance to the paid guardian is merely incidental. Our main claim is that the renter is not a paid guardian, but rather a borrower who is exempt from unavoidable accidents. In the following chapters we shall see additional halakhic implications of this conception.

The source of the borrower’s liabilities

We explained that the borrower is liable for unavoidable accidents because he acquired the object. His liability for unavoidable accidents is not part of his laws of guarding, but rather the fulfillment of the duty to return the object to its original owner. The borrower must restore the object to the owner, and any event that prevents him from doing so is his responsibility. Therefore he must return money in place of the object.

But if that is so, then his liabilities for theft and loss, and perhaps even for negligence, may likewise be derived from his ownership of the object. If the object was stolen or lost, then the borrower did not return it to its owner, and therefore he owes the owner the value of the object as a fulfillment of the duty of return. There is no reason to distinguish these liabilities from the liability for unavoidable accidents.[40] The conclusion is that all the borrower’s liabilities as a guardian are derivatives of his rights of use and his duty of return, and not the result of a separate guarding contract between him and the lender.

Is the borrower a type of guardian at all?

Above we defined the contract of borrowing as containing two components: the right of use and the undertaking of guarding. We now reach a different conclusion: in the case of the borrower, the liabilities of guarding are an outgrowth of the right of use. The right of use is based on the proprietary interest that the borrower has in the object, and that proprietary interest underlies all of the borrower’s liabilities, for at the end of the period a duty to return the object to its owner comes into force. That duty creates the guardian’s responsibility, all the way up to liability for unavoidable accidents and beyond.

Indeed, R. Yerucham Perla, in sections 23-25, elaborates in showing that at least according to Saadia Gaon and Tosafot Rid, the borrower is not considered a guardian at all. He acquired the object, and all his liabilities express the duty of return to the owner. This is how he explains the fact that, according to those authorities, there is no guardian’s oath at all in the case of a borrower. In every case he must pay, and therefore the question of an oath does not arise. His only exemption is where the animal died in the course of its work; there it would seem that he is exempt by oath, but R. Yerucham Perla explains that according to those views one must say that this is only a rabbinic oath and not the oath of guardians.

True, in the Sages the borrower is certainly treated as a type of guardian. But according to these views his liabilities derive from a different source than the liabilities of ordinary guardians. In his case there is also a difference in the laws themselves, since he is liable even for unavoidable accidents, unlike all the other guardians. It is therefore clear why the halakhah treats the borrower as a separate type of guardian. But the renter is not unique in his laws, for from the standpoint of the laws of guarding he is identical to the paid guardian, while from the standpoint of ownership he resembles the borrower (to some extent). If so, our question concerning him seems to remain intact.

The renter as a different kind of guardian

However, the direction we have taken thus far opens the door to understanding why the renter too is a different kind of guardian. First, the very fact that his guarding liabilities are like those of a paid guardian, while his ownership resembles that of a borrower, already places him in a different position from all the others. Yet this distinction seems not to touch the laws of guarding as such, for from that standpoint he is identical to the paid guardian.

We explained that in the case of the borrower there are not two aspects to the contract. It is a contract that grants rights of use and imposes a duty of return, and from these there are also derived the obligations ‘to guard’ — or more precisely, to return the object to its owner. What, then, about the renter? It is possible that in his case too the liabilities of guarding are derived from the rights of use, and therefore he is a different kind of guardian from the paid guardian. But he is also exempt from unavoidable accidents, and is therefore also different from the borrower. If this is indeed the renter’s status, then our puzzle about why he is counted as a fourth type of guardian disappears on its own.

To sharpen the point, let us take as an example the exemption of an animal that dies in the course of its work. It is clear that this exemption exists in the case of the borrower and the renter, but not in the case of the paid and unpaid guardians (for if they use the animal, they are treated as robbers, since they have misappropriated the deposit). At first glance, this is a clear difference between renter and paid guardian, and thus an answer to our question why the renter is a separate type from the paid guardian. But on more superficial reflection this difference might seem irrelevant, since it belongs not to the laws of guarding but to the laws of ownership. After all, the borrower too is exempt when the animal dies in the course of its work, and the reason for that (see Bava Metzia 93b) is that using the animal is the very purpose of the borrowing contract: Did he borrow it in order to set it up in a niche? If so, the exemption where the animal dies through its work is a product of the laws of ownership — that is, the right of use — and not of the laws of guarding. It is therefore reasonable that the renter should also enjoy such an exemption, since from the standpoint of ownership he resembles the borrower. But this difference does not, apparently, help us, because we are looking for a difference between renter and paid guardian that belongs to the laws of guarding, not merely to the laws of ownership.

But in light of what we have said here, it emerges that in the borrower all the liabilities of guarding are derivatives of the ownership contract. If so, if in the renter too the laws of guarding are part of the laws of ownership, then the exemption where the animal dies in the course of its work is indeed a clear difference between him and the paid guardian. It is entirely relevant, because like all the other liabilities of guarding of renter and borrower, it too is derived from ownership. It is not part of the rights of use themselves, but another duty of guarding that stems from the right of use, like all the other duties of guarding.

We are thus proposing an answer to the question why there are four types of guardians: the renter is a special type of guardian because he resembles the borrower in that his liabilities of guarding derive from his ownership and not from a guarding contract (as in the case of the paid and unpaid guardians). Therefore he too is exempt where the animal dies in the course of its work, and in this respect he differs from the paid guardian. Yet he also differs from the borrower, because although his liabilities derive from ownership, he is exempt from unavoidable accidents, unlike the borrower.

Why is the renter exempt from unavoidable accidents?

The question that now arises is: how can there be a type of liabilities that is derived from ownership, and yet there is an exemption from unavoidable accidents? At first glance, a duty of return rests upon him, and an unavoidable accident is not as though he acted; he has not fulfilled the duty of return.

But in light of what we said in the previous chapter, this problem does not arise at all. We saw that the renter’s ownership lapses automatically at the end of the term. There is no duty of return upon him, for the animal reverts automatically to the ownership of its original owner. The borrower, of course, must not be negligent with the animal — that is, he must allow it to return to its owner — but he need not reconvey it to them. This, then, is the basis of the renter’s exemption from unavoidable accidents. If an unavoidable accident occurs, the renter is exempt, because he is not responsible at that level for restoring the animal to its owner. Theft or loss, and certainly negligence, do obligate him, because a duty rests upon him to see to it that the animal returns safely to its owner — physically, not by way of a juridical reconveyance.[41]

Perhaps this can be explained even more simply. Contrary to what we proposed above, maybe in the renter there is no duty of return at all. This is something like a double contract (rights and liabilities of guarding), but in fact all the renter’s guarding liabilities — which are exactly like those of the paid guardian, not like those of a borrower exempt from unavoidable accidents — are imposed upon him as part of the ownership contract, and not, as we suggested above, as the implementation of some duty of return. They are part of the consideration for the right of use that he received. According to this, the renter has no duty of return at all, and it is therefore obvious why he is exempt from unavoidable accidents.

An example from the Shakh’s view regarding collateral

Above we cited the dispute among the medieval authorities regarding the liabilities of a creditor who holds collateral. Rashi and Ba’al HaMa’or hold that he is liable even for unavoidable accidents. Nachmanides and Tosafot hold that he is liable only like a paid guardian. Nachmanides sharply attacks the opposing view, explaining that in the case of a borrower all the benefit is his, and therefore he is liable for unavoidable accidents. But with one who guards collateral not all the benefit is his; how can one say that he is liable for unavoidable accidents?

The Shakh, Hoshen Mishpat sec. 72, sec. 9, rejects Nachmanides’ difficulty and says that in the borrower’s case the liabilities are under the laws of guardians, and therefore his liability for unavoidable accidents is derived from the fact that all the benefit is his. But in the case of one guarding collateral, he acquired it by virtue of the law of R. Isaac, and liability for unavoidable accidents therefore follows automatically from that acquisition, even if not all the benefit is his.

Nachmanides here remains consistent with his own view that the borrower did not acquire, like Tosafot. But according to Maimonides and those who follow him, who hold that both borrower and renter acquired the object, we can make a similar claim. In order for the borrower to acquire the object, the consideration that all the benefit is his is required. The meaning of that claim is that the borrower acquired the object (or, according to some views, that therefore the Torah places the object in his domain so as to obligate him in unavoidable accidents). But in the case of collateral there is acquisition by virtue of the contract itself between creditor and debtor, and liability for unavoidable accidents follows from that acquisition on its own. Therefore it is not necessary that all the benefit be his.

Nachmanides apparently understands that the borrower acquires in order to become liable for unavoidable accidents, whereas the holder of collateral acquires irrespective of that. According to Maimonides and those who follow him, however, we explained here that the borrower too acquires the object by virtue of his contract with the lender, not in order to become liable for unavoidable accidents. Therefore his liability for unavoidable accidents is like that of the holder of collateral. The expression ‘all the benefit is his’ does not mean a condition for ownership, but rather expresses the ownership itself: all the benefit is his because he is the owner. Rabbi Lichtenstein likewise inclines in this direction in explaining Rashba’s view cited above.

Rashi in the sugya of Avodah Zarah

Whether we understand it this way or another, Rashi’s explanation in Avodah Zarah now fits very well, for he cites liability for unavoidable accidents as an indication that the borrower acquired the object. We asked: why does the Talmud there link the question of the borrower’s ownership to the ownership of the renter, who is not liable for unavoidable accidents? According to what we have now said, the answer is straightforward, since all the liabilities in the case of both borrower and renter are a product of ownership. True, in the borrower’s case the proof is unequivocal, since liability for unavoidable accidents cannot arise from any other source. But once we understand this in the case of the borrower, we can apply it to the renter as well. His liabilities too arise from ownership, even though he is exempt from unavoidable accidents, as explained. If so, Rashi’s words there concern both borrower and renter, in full accordance with the Talmud itself. The expression ‘all the benefit is his’ with respect to the borrower expresses the fact of ownership itself, and not merely a condition of liability for unavoidable accidents. Therefore even in the renter’s case, where not all the benefit is his, there is still ownership. He is, however, not liable for unavoidable accidents, as explained above.

The renter’s laws in the Torah are part of the section of the borrower

We may conclude by noting that there is indeed a source in the Torah for the renter’s laws (as cited by R. Yerucham Perla above, and perhaps as implied by Maimonides’ language at the beginning of the Laws of Rental that we cited at the opening of the article — though see the qualification we raised above from his words in chapter 1 of the Laws of Rental, law 3), namely the verse And if it is a hired animal, it comes with its hire. That verse appears within the section of the borrower, and it strongly suggests that the renter’s law is some limitation upon the law of the borrower. That is: the renter is a borrower with an exemption from unavoidable accidents, and not a paid guardian.

Accordingly, the resemblance to the laws of the paid guardian is only an accidental result of the exemption from unavoidable accidents, not the result of an essential similarity to the paid guardian. If so, the renter is indeed a fourth type of guardian: a borrower who, because he pays for his use of the rented object, is exempt from unavoidable accidents. This is also what emerges from the simple meaning of the Torah, which placed the renter within the section of the borrower.

And so Rashi writes on that verse (Exodus 22:14):

If it is a hired animal—if the ox was not borrowed but rented, it came into this renter’s possession through rental and not through borrowing, and all the benefit is not his, since he uses it by virtue of his payment; therefore he does not have the legal status of a borrower to be liable for unavoidable accidents. And Scripture did not specify what his law is, whether like an unpaid custodian or like a paid custodian; therefore the sages of Israel disagreed about him: How does a renter pay? Rabbi Meir says: like an unpaid custodian. Rabbi Yoḥanan says: like a paid custodian.

Thus Rashi here too, like the other Rashi passages we cited from the Talmud, explains the renter as a borrower with an exemption from unavoidable accidents. He is written in the section of the borrower because he belongs to that section. The novelty in his case is his exemption from unavoidable accidents. I later found in the aforementioned Likutei Sichot that he wrote similarly, and even brought additional support from several Rashi passages (see Bava Metzia 80b, s.v. ‘And R. Yehuda’ and s.v. ‘R. Meir,’ and there 81a, s.v. ‘Lend it to me’).[42]

As for the other paid guardians who do not arise out of contract, in simple terms they resemble the paid guardian and not the borrower, and therefore there is no reason at all to define them too as separate types. The finder of lost property is an ordinary guardian, except that his wage is given only indirectly (according to R. Yosef, who grounds it in exemption from charity — ‘the penny of R. Yosef’), or else the Torah itself imposes upon him the liabilities of a paid guardian (according to Rabbah). In any event, the liabilities of guarding there are not due to ownership. From the standpoint of the laws of guarding he is a regular paid guardian (though there is of course a novelty here), and certainly not a different type of guardian. Regarding collateral and artisan the matter is somewhat more complex, but this is not the place to elaborate.[43]

The view of the Jerusalem Talmud and of Maimonides

Rabbi Lichtenstein, in the lectures cited above, chapter 2 (pp. 26-27), notes that the Jerusalem Talmud is explicit against Rashi. The Jerusalem Talmud, Shevuot, chapter 8, law 1, states:

You thus find that there are three passages: the lowest one concerns the borrower, the middle one concerns the paid custodian and the renter, and the upper one concerns the unpaid custodian…

It emerges from the Jerusalem Talmud that the section of the paid guardian also includes the renter.

Rabbi Lichtenstein continues and says that Maimonides too does not appear to have learned like Rashi. Rashi understood the verse If it is a hired animal, it comes with its hire as referring to the renter, that is, to the borrower who comes with payment. But from Maimonides it appears that he understood that verse as referring to the owner and not to the guardian, for he writes as follows (Laws of Rental 1:3):

If one deposits with another, whether without payment or for payment, or lends him an item, or rents it to him—if the custodian borrowed the owners along with their item or hired them, the custodian is exempt from everything… as it is said: “If its owner is with him, he shall not pay. If it is hired, it comes with its hire”…

It should be noted, however, that nothing here forces us to conclude that according to Maimonides the renter is not a type of borrower. What one can infer from here is only that according to Maimonides one is not compelled to adopt the position that the renter is a borrower exempt from unavoidable accidents; but there is also no proof that he denies it.[44]

Implications: a borrower from whom the lender receives benefit

The Minchat Chinukh, commandment 59, sec. 15 (following Ketzot HaHoshen, sec. 340, sec. 5), opens with a sentence that says explicitly that the renter is a borrower exempt from unavoidable accidents:

Now, the reason that a renter is exempt from unavoidable accidents even though he uses it, whereas a borrower is liable for unavoidable accidents, is that since a borrower has all the benefit, he is liable for unavoidable accidents. But a renter does not have all the benefit, since he pays rent to the owner; therefore he is exempt from unavoidable accidents and has only the status of a paid custodian.

The author of Ketzot HaHoshen there adds that if the renter gives the lessor less than a perutah, even though a paid guardian is not created for less than a perutah, he still becomes a renter. The reason is that in a loan where the lender receives some benefit — even less than a perutah — then not all the benefit is the borrower’s, and therefore he is like a renter. We thus see that the renter is not a paid guardian, but rather a borrower in a case where the lender also receives benefit, and that is why this ‘borrower’ is exempt from unavoidable accidents. So too follows from Rashi in Bava Metzia 80, whose words we cited above. The Minchat Chinukh, however, remains uncertain on this point; see the conclusion of his discussion there.

Even where the lender receives a non-monetary benefit, such as the benefit of a commandment, several decisors wrote that the borrower is not deemed a borrower but a renter (see the Responsa of Ran, sec. 19, cited by the Shakh, Hoshen Mishpat sec. 72, sec. 29).[45] This too proves that the renter is a borrower with an exemption from unavoidable accidents, and not a paid guardian.

In the next chapter we shall see that this emerges from the plain meaning of the sugya with which we began.

A dispute among the medieval authorities on this matter

Rabbi Lichtenstein, in chapter 2 (p. 28 ff.), notes that there is a dispute among the medieval authorities over whether the renter is literally like a paid guardian, or rather a borrower exempt from unavoidable accidents. He relates this to the foundation of the renter’s liability. Tosafot, s.v. ‘Let us say,’ Bava Metzia 80b, explain that the renter becomes liable like a paid guardian because, after all, he receives benefit in return for his guarding (even though he pays for it, and this is not the place to elaborate).[46] Tosafot, of course, remain consistent with their view, as we saw, that the renter has no ownership in the object (and perhaps even in the borrower’s case they would learn that this too is a guarding contract, unless borrowing differs from rental).

By contrast, the Ra’avad (in Shitah Mekubetzet there, s.v. ‘Let us say the Mishnah’) explains why the artisan is like the renter, and writes as follows:

And an artisan is comparable to a renter, for he performs work on the object itself like a renter.

From his language it appears that the foundation of the artisan’s and renter’s liability is not that they receive payment or some other benefit, but the very fact that they work upon the body of the object. It is quite clear that the foundation of the liability is like that of the borrower, with only an exemption from unavoidable accidents, as we explained above.[47]

F. ‘Four guardians and three sets of laws’

The sugya of Bava Metzia 93a and Shevuot 49b

On the Mishnah we cited at the beginning of our discussion — ‘There are four guardians’ — the Talmud brings the dispute between R. Yehuda and R. Meir as follows:

Who is the tanna who taught four custodians? Rav Naḥman said in the name of Rabbah bar Avuha: It is Rabbi Meir. Rava said to Rav Naḥman: Is there anyone who does not accept four custodians? He said to him: This is what I am saying to you: Who is the tanna who holds that a renter is like a paid custodian? It is Rabbi Meir. But we have heard Rabbi Meir say the opposite, for it was taught: How does a renter pay? Rabbi Meir says: like an unpaid custodian; Rabbi Yehuda says: like a paid custodian. Rabbah bar Avuha taught it in reverse. If so, four? There are only three! Rav Naḥman bar Yitzḥak said: There are four custodians, but their laws are three.

The Talmud’s question is puzzling: why does the Talmud use the phrase ‘if so’? At first glance, according to every opinion there are only three guardians, and not only according to the view that the renter is like a paid guardian. This difficulty exists however one construes the opening or closing clauses. Moreover, at the beginning of the sugya the assumption is the very opposite: both for Rav Nahman and for Rava it is obvious that according to everyone there are four guardians, and no one disagrees with that. If so, how can the continuation of the sugya then ask that there are really only three? The question itself is true and real, but it is not clear why at the beginning they simply assumed the exact opposite.

Indeed, Rashi in the sugya in Shevuot wrote that the words ‘if so’ are not to be read (and in the text of Shevuot as we have it they do not appear). Tosafot, s.v. ‘If so,’ in the sugya in Bava Metzia, do read them, but write that they are not precise usage (and they bring parallels there). Thus Rashi and Tosafot both agree that there are only three types of guardians, according to all tannaitic opinions. The question of the first clause therefore remains: why was it obvious at the outset that according to everyone there are four guardians, with no dissent? At first glance we should have assumed precisely the opposite: according to everyone there are only three types. Beyond that, Tosafot’s disregard of the plain meaning of the words ‘if so’ in Bava Metzia is itself somewhat forced.

The view of Tosafot HaRosh and Maharam

In fact, in Tosafot HaRosh on the sugya in Shevuot, in the name of Maharam of Rothenburg, we find a version that reads ‘if so,’ and he also writes that the Talmud used that phrase precisely. His language is as follows:

Our teacher Rabbi Meir explained: Granted, if you say that this is what it means—there are four custodians; regarding three, everyone agrees, and concerning the fourth Rabbi Meir and Rabbi Yehuda disagree—then that is why it counts them as four, even though according to both this master and that master their laws are three. And the fact that in the latter clause it explains that a renter is like a paid custodian is not because it intends to rule anonymously like Rabbi Meir, for since in the first clause which taught four custodians it informed us that this is a dispute, it is not the way of the Talmud to teach a dispute and then an anonymous ruling in the same breath. Rather, it was not concerned in the latter clause to teach anything but the words of Rabbi Meir, because the anonymous Mishnah follows him. But once you say that the entire Mishnah is Rabbi Meir, then four are really only three!

He explains that in the initial assumption they thought the first clause of the Mishnah went according to everyone, and then there are indeed four guardians: on three everyone agrees, and about the fourth — the renter — R. Meir and R. Yehuda disagree. It appears from his words that there are four guardians not only according to the editor of the Mishnah, who leaves the renter’s law open, but also according to R. Yehuda and R. Meir themselves, for the Talmud says, ‘Is there anyone who does not hold that there are four guardians?’ In other words, both R. Meir and R. Yehuda agree that the renter is a different kind of guardian, and therefore they dispute to whom he should be likened. At that stage the Talmud understood that the Mishnah’s statement in the latter clause that the renter is like a paid guardian was only because that is the halakhah, while the Mishnah as a whole goes according to everyone. At that stage it is indeed obvious that there are four different types of guardians, because the renter is a different type according to all views.

Maharam explains that in the Talmud’s conclusion the latter clause is specifically according to R. Meir, and it follows that the first clause as well is according to him and not according to everyone. Presumably this is because Rav Nahman’s opening language is, ‘Who is the tanna of [the Mishnah] “four guardians”?’ — implying the whole Mishnah, and not only the latter clause. Hence he understands Rav Nahman’s answer not as saying that he was speaking only of the latter clause, but as an explanation of his opening statement. The latter clause testifies regarding the first clause that the whole Mishnah is according to R. Meir.

The question then arises why the Mishnah writes that there are four guardians; therefore the Talmud uses the expression ‘if so.’ That is, once we understand that the Mishnah is according to R. Meir, it becomes difficult why there should be four guardians rather than three. To this the answer is given: there are indeed four guardians, but their laws are only three.

Thus, although Maharam’s reading may at first glance seem somewhat forced in the simple meaning of the Talmud, it appears to offer the most straightforward reading. According to him, no stage of the sugya is difficult at all, and Rav Nahman’s words are in fact more precise than they would be according to the reading of Rashi and Tosafot.

The Talmud’s conclusion according to Maharam and the Rosh: according to R. Yehuda there are only three types of guardians

Still, according to Maharam one may raise a slight question: why in the initial assumption was it obvious to the Talmud that both R. Meir and R. Yehuda held there to be four guardians (and that they disagreed only about the renter’s laws), whereas at the end the question is raised in the language ‘if so’? That is: once we reach the conclusion that the Mishnah is according to R. Meir, then the difficulty arises why there are four guardians, whereas in the opening stage we assumed that even R. Meir himself held there to be four.

One might suggest that Maharam means only that the editor of the Mishnah holds there to be four guardians, because he looks from a standpoint outside the dispute of R. Meir and R. Yehuda (that is, there is another type of guardian in the sense that there is an additional type whose law is disputed), but not that R. Meir and R. Yehuda themselves agree there are four. But that cannot be, for we already explained that according to this Rava’s words become difficult: Rava assumes there is no one in the world who disagrees that there are four guardians, implying that this includes R. Meir and R. Yehuda themselves.

We are therefore forced to say that Maharam and the Rosh read the Talmud as follows: it turns out that Rav Nahman interpreted the whole Mishnah, including the first clause, according to R. Meir. Presumably what drove him to this is that, in his opinion, even the first clause does not fit R. Yehuda, because R. Yehuda apparently holds that there are three guardians and not four (the renter is simply one and the same category as the unpaid guardian). The Talmud then asks: if so, why assume that according to R. Meir there are four guardians and not three? Just as, according to R. Yehuda, we have seen that there are only three, so too according to R. Meir there are only three, because the renter is like the paid guardian and they constitute one category. To this Rav Nahman answers that there are indeed four types of guardians, though their laws are only three. In other words: there are four types of guardians even though, from a halakhic perspective, these amount to only three legal groupings.

But now let us ask: after the conclusion, according to whom does the first clause go? From the flow of the sugya it appears clearly that even in the conclusion the whole Mishnah, including the first clause, remains according only to R. Meir. That is, even in the conclusion R. Yehuda himself still cannot be read into the first clause, even if we interpret it as saying that there are four types of guardians but only three sets of laws.

If so, the Talmud’s conclusion according to Maharam and the Rosh is that according to R. Yehuda, who holds that the renter is like an unpaid guardian, there are only three types of guardians; whereas according to R. Meir, who holds that the renter is like a paid guardian, there are four types, though only three sets of laws.

Explaining the conclusion: back to our discussion

In light of what we have said in the previous chapters, this can be explained as follows. R. Meir holds that the renter is like a paid guardian, and so the halakhah rules. As we have seen, the renter is not truly a paid guardian; rather, this is a borrower with an exemption from unavoidable accidents, and the halakhic resemblance to the paid guardian is merely accidental. If so, the conclusion is indeed that there are four types of guardians and three sets of laws, exactly as we explained above. But according to R. Yehuda, the renter is like an unpaid guardian, and therefore it is not plausible to conceive him as a borrower with an exemption from unavoidable accidents. R. Yehuda understands every rental contract as containing two components: the transfer of rights of use and an undertaking of guarding as an unpaid guardian. According to him, therefore, from the standpoint of the laws of guarding the renter is not a fourth type of guardian at all, but simply an unpaid guardian. Hence in his view there are only three types.

Let us clarify this further. When we seek to understand R. Meir’s position, according to which the renter is like a paid guardian, there is no way to understand this as the result of a guarding contract, for he receives nothing in exchange for his guarding. He pays for the rights of use, so why should he be liable like a paid guardian to guard more than ordinary people, and to bear broader responsibility (for theft and loss)? Therefore there is no escape from the conclusion that according to R. Meir these liabilities are not the result of a guarding contract, but derivatives of ownership. The resemblance to the liabilities of the paid guardian is merely accidental, for the renter receives nothing for his guarding and there is no reason to view him as a paid guardian.

R. Yehuda’s position, however, is that the renter is like an unpaid guardian. To explain such a position there is no need at all to assume that the liabilities arise from ownership, for he is simply and truly an unpaid guardian. He received the animal for his own use, and obviously he must guard it with ordinary care; otherwise the owner would never have rented it to him. But he did not receive payment for guarding it, and therefore his liabilities are those of an unpaid guardian. If so, the natural and straightforward result is that he is an unpaid guardian; there is no need to explain this as a derivative of his ownership. Thus R. Yehuda understands the rental contract as containing two independent components: obligations of guarding and responsibility, and rights of use. Hence according to him the renter is simply an unpaid guardian, and there are no four types of guardians at all.

We should note that this is exactly what we said above in resolving the question of Ma’ayanai HaChokhmah on Rashi: one who holds that the renter is like an unpaid guardian does not accept the notion that he has proprietary rights in the object — that is, he does not agree that the renter is a borrower with exemptions. In his view there are apparently two separate transactions between renter and lessor: permission to use in exchange for rent, and an obligation of guarding as an unpaid guardian. Accordingly, there are not four types of guardians, but only three.

After writing this, I saw that in Torat Chaim on the sugya in Shevuot he wrote the opposite: according to R. Yehuda there are certainly four guardians, since the renter certainly is not literally an unpaid guardian. Only according to R. Meir, who holds that the renter is like a paid guardian, does the question arise why there are four guardians at all. His words imply that he read the Talmud like the Rosh — that the question arose only with respect to R. Meir (and that the reading ‘if so’ is exact) — but his conclusion is the reverse. In my opinion his approach is forced, and the Rosh and Maharam certainly did not read the sugya that way.[48]

The Lubavitcher Rebbe, in Likutei Sichot (vol. 31), at the beginning of Parashat Mishpatim, likewise elaborates in explaining the sugya and searching for the uniqueness of the renter. In practice he writes much like what we have said (see especially there, note 20 and its context), though he does not enter into the legal difference between the renter and the other guardians, focusing instead mainly on the psychological plane (that the renter’s concern in the transaction is to secure rights, unlike the unpaid guardian whose defining concern is guarding itself). He explains that the renter is a composite of borrower and paid guardian, and it somewhat sounds as though he sees two contracts operating in parallel, like the initial understanding that we rejected. As noted, according to that view it is difficult to see the renter as a different type of guardian, since from the standpoint of the laws of guarding he is simply a paid guardian, and what does it matter that he also has rights of use? That would only mean a different situation, like collateral, artisan, or lost property. Yet he writes that the renter is like a paid guardian in the quantity of the liability, but like the borrower in the quality of the liability (see there p. 114 and elsewhere). Perhaps his intention is what we have suggested, that the liabilities of guarding are an outgrowth of the rights of use; this still requires further clarification.

The view of the other medieval authorities

Now, in the language of the Rif on this sugya, the matter is almost explicit. He writes:

There are four types of custodians, etc. But these four custodians are really three, for a renter is like a paid custodian! Rabbi Naḥman bar Yitzḥak said: There are four custodians, but three sets of laws.

Thus the Rif understands the question on the Mishnah as directed only according to the view that the renter is like a paid guardian. Accordingly, the explanation in the conclusion — ‘four guardians and three sets of laws’ — was also stated only according to R. Meir. This is exactly the explanation we proposed in the Rosh and Maharam.

As noted, our interpretation is necessary for Maharam and Tosafot HaRosh. Tosafot, however, disagree, for they wrote that the words ‘if so’ are not precise. If so, they presumably disagree with the understandings we have proposed here. As we saw, in their view the renter has no proprietary rights in the rented object, so it is no surprise that they disagree here as well. According to them the classification into four types of guardians is not precise. That is, Tosafot apparently understand the rental contract as containing two independent components, both according to R. Meir and according to R. Yehuda. Nevertheless there are four types of guardians, because in practice there are indeed four patterns of law: the renter is exempt where the animal dies in the course of its work and in unavoidable accidents; the paid guardian is liable where it dies in the course of its work (for if he used it, that is misappropriation), and exempt in unavoidable accidents; and the borrower is liable in unavoidable accidents and exempt where it dies in the course of its work.

We should note that Tosafot here too ignore the fact that these differences belong not to the laws of guarding, but to the question of rights of use. It may be that this is another reflection of what we saw above, that Tosafot do not accept the Tzafnat Pane’ach’s view that the borrower’s liability for unavoidable accidents is due to the duty of return. We saw that Tosafot interpreted the sugya in Yevamot as referring to the borrower for an external reason (usury), and they probably do not agree with the substantive explanation proposed by the Tzafnat Pane’ach.

As for Rashi, we explained above that both borrower and renter have proprietary rights in the object. The fact that he too does not read the words ‘if so’ does not necessarily contradict what we have said. It may be that according to him both R. Yehuda and R. Meir agree that there are four guardians and three sets of laws, and therefore there is no place for the phrase ‘if so.’ In that respect he is like Tosafot, though not for their reasons.

It should be noted that Maimonides does indeed quote this Mishnah, in Laws of Rental 1:1 (and see also there at the beginning of chapter 2):

Four types of custodians are stated in the Torah, but they have three sets of laws; and these are the four custodians: the unpaid custodian, the borrower, the paid custodian, and the renter.

Yet it does not appear that he links the renter to the borrower, but rather to the paid guardian. So it seems from his words in law 2, where he shows that their laws are identical. But that may simply follow the Talmud, and not an essential similarity. True, the way he divides his code between the Laws of Rental and the Laws of Borrowing and Deposit appears problematic in light of what we have said. In the Laws of Rental he discusses the paid guardian and the renter, whereas in the Laws of Borrowing and Deposit he discusses the unpaid guardian and the borrower. Had he held like Rashi, it would have seemed more natural to divide between the laws of renter and borrower, who have proprietary rights in the object, and those of the paid and unpaid guardians, who are merely guardians.

On the other hand, we saw that Maimonides explicitly writes that borrower and renter have proprietary rights in the object. Above we explained that the laws in the sugya of Avodah Zarah are unique and do not contradict our issue. If so, what we have said in understanding renter and borrower can be said also according to Maimonides. It is entirely possible that the arrangement of the laws in his code is unrelated to the substantive understanding, but reflects only technical classifications: the Laws of Borrowing and Deposit deal with gratuitous transactions, whereas the Laws of Rental deal with paid contracts. If so, there is no contradiction in Maimonides to the conceptual account of the laws of guardians that we have presented here.

G. Differences between the renter, the paid guardian, and the borrower

Introduction

Up to this point we have seen that, at least according to a number of medieval authorities, the renter is an independent type, neither borrower nor paid guardian: he is a borrower exempt from unavoidable accidents, and the fact that most of his laws coincide with those of the paid guardian is only an accidental result of this fact. In this chapter we shall briefly examine what follows from our analysis in two directions: we shall see whether there are differences between renter and paid guardian that reflect the fact that the renter is a guardian of a different type; and, for the same reason, we shall also examine the differences between renter and borrower.

At first glance there are two independent components included in the contracts of rental and borrowing: rights of use and ownership, and the responsibility of guarding. One would expect that the differences between renter and paid guardian would be connected to the laws of ownership and rights of use (the renter receives rights of use while the paid guardian does not, though their laws of guarding are identical), and that the differences between renter and borrower would be connected to the laws of guarding (both are owners of the object for its usufruct, but they differ in guarding obligations and responsibility).

Yet, as we saw above, it is difficult to separate those two components. The reason is that, at least according to several views, in the case of borrower and renter the responsibility of guarding is derived from the contract concerning rights of use, and is not a separate contract. Therefore differences that arise from rights of use become relevant also to the classification and taxonomy of the types of guardians. For example, the fact that the renter is exempt when the animal dies in the course of its work while the paid guardian is liable stems, at first glance, from rights of use; but it may be a direct consequence of the fact that all the guarding obligations of the renter are derived from his rights of use, and so is this obligation. There are of course also technical differences concerning how the contract takes effect (the effective act of acquisition), and concerning the possibility of the owner terminating the contract within the term (which in rental is generally impossible).

We shall therefore try in this chapter to survey the differences between the renter and the other guardians. In light of what we said above, we will not distinguish here between duties of guarding and rights of use. The very collection of all these differences is important in itself, precisely against the background that ordinarily no distinction is drawn between renter and paid guardian as regards guarding, and between renter and borrower as regards rights of use. These differences, in all their variety, also illustrate more sharply the fact that the renter is a guardian of a different kind. We will not be able to discuss each aspect at length, and therefore in the course of our brief remarks we shall try mainly to indicate the connection between these laws and the characterizations of the renter that we have encountered thus far.

Differences we have already encountered

We have already noted above several differences among the various guardians. As halakhah, the law of price fraud applies to rental, whereas to the paid guardian it presumably does not. Rental is ‘a sale for the day’ (and even according to Tosafot and those who follow them, we saw that there is a lien in the rented object that is akin to acquisition), whereas a guarding contract is certainly not a sale in any sense. We also saw one leniency: a paid guardian must guard more carefully than ordinary people guard their own property,[49] whereas the renter apparently need guard only in the ordinary way (like the unpaid guardian). This distinction too depends on our thesis, for in the renter’s case there are no duties of guarding as such, but only liabilities of responsibility by force of the rights of use. Therefore, although his responsibility is like that of a paid guardian, his duty of care is like that of an unpaid guardian — ordinary care. Likewise, according to Ketzot HaHoshen, sec. 340, sec. 5, a person becomes a renter even if he paid less than a perutah, unlike the paid guardian. We connected this too to our explanation: in the renter’s case the contract is not about guarding. The guarding is derived from ownership, and wherever there is ownership there are guarding liabilities. By contrast, in the paid guardian’s case the contract is wholly and entirely a guarding contract, and if he receives less than a perutah there is no effective contract.[50]

With respect to the borrower, there is the basic difference that the borrower is liable for unavoidable accidents while the renter is not, and we have already explained this in terms of the different character of the time-limited ownership that each has. We also saw a difference regarding the law that there is appraisal for the borrower, regarding one who took a cow from his fellow on valuation terms according to the Tzafnat Pane’ach in Maimonides, and regarding a priest who borrowed a cow from his fellow according to the Tzafnat Pane’ach in Tosafot on Yevamot (though we noted that this does not appear to have been Tosafot’s intent). Both of those laws served us as illustrations of the distinctions we made.

Regarding the requirement that something be ‘yours,’ in the case of borrowing the Talmud explicitly states, and the halakhah rules, that one does not fulfill the obligation with a borrowed lulav. Regarding rental, true, we saw views according to which specifically in rental the requirement of ‘yours’ is fulfilled even though in borrowing it is not, but according to most views it seems that even in rental the requirement is not fulfilled. As for the paid guardian, no one disputes that the requirement of ‘yours’ is not met. It should be noted, however, that in the sugya of collateralized leaven (Pesachim 31a-b) it seems that even the paid guardian has some proprietary dimension in the object; but that is a special rule in the laws of leaven and this is not the place to elaborate.[51]

Changing the type of work

A borrower or renter may not change the type of work stipulated with the owner. If he did so and the animal died because of the change, he is liable. A number of later authorities distinguished here between borrower and renter. In Sema, sec. 341, sec. 20, a contradiction is noted in the Tur, the Shulchan Arukh, and the Rosh concerning a change to the same kind of work itself (for example, a spade borrowed for another orchard): in rental it is written that he is not liable to pay, whereas in the case of a borrowed spade it is written that he is liable. One cannot resolve this by invoking the difference in laws of guarding, namely that the borrower is liable for unavoidable accidents, since even the borrower is exempt where the animal dies in the course of its work.

However, in Netivot HaMishpat there, sec. 13, a distinction is drawn between borrower and renter: borrowing is gratuitous, and therefore there is room to assume the owner is particular, whereas in rental, where payment is made, we estimate that the lessor is not particular about this.

According to our approach, the explanation is simple. One might say that in the borrower’s case there is a duty of return, and therefore even if the change was to the same kind of work itself, still the duty of return did not lapse, for it is not like a case where the animal died in the course of the work for which it was loaned. Therefore, even though it died in an unavoidable accident, he must return money in its place. In the renter’s case, however, although it is not considered a case of ‘died in the course of the work’ for which it was rented, the acquisition lapses automatically at the end of the term, and therefore since this is an unavoidable accident he need not pay for it. The reasoning of the Netivot is merely the explanation of this very difference.

The Maggid Mishneh, in chapter 4 of the Laws of Rental, writes that if one rented an animal to carry a man and instead had a woman ride it, even though the woman is heavier, this is not deemed such a change as to incur liability. Yet at the beginning of the Laws of Borrowing he writes that even if one changed it to a lighter task, that is considered a change and he is liable. This too proves our point: in the borrower’s case such a change counts as a different labor, and therefore it is not a case of ‘died in the course of its work.’ In the renter’s case, although it is indeed not considered ‘died in the course of its work,’ he is nevertheless exempt in unavoidable accident, because no duty of return rests upon him.

A similar discussion appears around Maimonides’ ruling in chapter 5 of the Laws of Rental, law 5, where he states that the rule that a renter may not subrent applies only to movables, but in the case of land he may. The Gra there cites responsa of Rashba writing that the same is true of borrowing, and so too Mahaneh Ephraim, Laws of Rental sec. 19, as well as Ran and Nimmukei Yosef on Bava Metzia 79b. In Dvar Avraham, vol. 2, sec. 17, secs. 6-7, the issue is discussed at length, and a possible distinction is raised between them in accordance with the Netivot cited above. However, the conclusion there is like Mahaneh Ephraim and Rashba, that there is no difference between borrowing and rental on this point.

In any event, we should note that the difference we have cited here is not connected to differences between borrower and renter in the laws of guarding; rather, these are differences concerning rights of use and the duty of return.

The acquisition of the double payment

A sugya that explicitly raises the difference between renter and borrower appears in Bava Metzia 34a. The sugya deals with the law of a guardian who paid, or who said that he is willing to pay, and whether he receives the double payment when the thief is found. Let us briefly review the course of the sugya.

The Mishnah states that if the guardian paid, and only afterward the thief was found, then the thief pays the double payment to the guardian and not to the owner. The Talmud discusses the law of a guardian who merely said that he wishes to pay:

He paid and did not want to swear [etc.]. Rabbi Ḥiyya bar Abba said in the name of Rabbi Yoḥanan: “He paid” does not mean that he actually paid; rather, once he said, “I hereby will pay,” even though he has not yet paid… A baraita was taught in accordance with Rabbi Yoḥanan: If one rented a cow from another and it was stolen, and the former said, “I hereby will pay and will not swear,” and afterward the thief was found—the double payment is paid to the renter. Rav Pappa said: In the case of an unpaid custodian, once he says, “I was negligent,” he transfers the double payment to him, for if he wished he could exempt himself by claiming theft. In the case of a paid custodian, once he says, “It was stolen,” he transfers the double payment to him, for if he wished he could exempt himself by claiming breakage or death. A borrower who says, “I hereby will pay,” does not thereby acquire the right to the double payment. By what claim could he exempt himself? By claiming death due to labor. But death due to labor is uncommon. Some say that Rav Pappa said: A borrower as well, once he says, “I hereby will pay,” acquires the right to the double payment, for if he wished he could exempt himself by claiming death due to labor. Rav Zevid said to him: This is what Abaye said: A borrower does not acquire it until he actually pays. What is the reason? Since all the benefit is his, by mere words he does not transfer to him the double payment. A baraita was taught in accordance with Rav Zevid: If one borrowed a cow from another and it was stolen, and the borrower hastened and paid, and afterward the thief was found—the double payment is paid to the borrower.

R. Pappa explains that the depositor transfers the right to the double payment to the guardian in return for his willingness to pay, since the guardian could have exempted himself with various claims. As regards the unpaid guardian, the paid guardian, and the renter, the law is agreed. As regards the borrower, R. Pappa himself is uncertain, because the claim that the animal died in the course of its work is uncommon. Abaye, however, holds that in the borrower’s case a mere statement is never enough, and therefore he does not acquire the double payment until he actually pays. The reason is not the same reason given by R. Pappa (that the claim that the animal died in the course of its work is uncommon),[52] but rather that since all the benefit is his, by mere words the owner does not transfer the double payment to him. This is explicit in the baraita brought there in the Talmud’s conclusion, and so the halakhah rules.

Thus we have here another difference between renter and borrower, and this too is tied to the fact that in the borrower’s case all the benefit is his; therefore the owner is not inclined to give him anything beyond the right to use the object.

At first glance, however, this does not seem to be a difference connected to our discussion, for this is merely an estimate of the owner’s intention, depending on whether he is doing the borrower a favor, and not on the nature of the borrower’s proprietary rights and liabilities.

Still, we may recall what we suggested above from the words of the Netivot, namely that this very estimate determines the borrower's acquisitions and laws. The phrase ‘all the benefit is his’ indicates that he has ownership of the object, and because he acquires the borrowed object without payment, all the benefit is his — unlike the renter. Above we saw that for this reason he is liable for unavoidable accidents, not because there is an estimated intention that the parties meant this in the contract, but as a result of the acquisition that he has in the object.[53]

Now, Maimonides brings this law in chapter 8 of the Laws of Deposit, law 2. But he does not write the reason that all the benefit is his. He simply states that in the case of the borrower he acquires the double payment only if he actually paid, whereas in the case of the other guardians he acquires it even if he merely said he would pay. The commentators have already noted this, perhaps suggesting that Maimonides’ terse wording means he ruled like the first version of R. Pappa. But this is difficult in the sugya, for the Talmud clearly brings the baraita as proof for Abaye, and the halakhah follows him.

Now we find in the Jerusalem Talmud, at the beginning of chapter 8 of Shevuot, which we cited above, that it too adopts this distinction, but without the rationale ‘all the benefit is his,’ and on the basis of another reason:[54]

It was taught: “And steals from the house of the man”—not from the house of the borrower. And should one then say: “And steals from the house of the man”—not from the house of the paid custodian and the renter? Since its safekeeping rests upon him, it is as though it is upon him.

In this baraita, the reason for the difference between renter and paid guardian on the one hand, and borrower on the other, is not based on estimates of the owner’s intention. The reason is the verse and steals from the house of the man. True, Pnei Moshe there explained this like the Babylonian Talmud, but that is forced in the language of the Jerusalem Talmud.[55] Ridbaz there notes that the Jerusalem Talmud remains consistent with its own view, according to which the guardian’s acquisition of the double payment when he paid is learned from a verse, and not because the owner is presumed to transfer the double payment to him, as in the Babylonian Talmud (see the Jerusalem Talmud at the beginning of the chapter HaMafkid).

If so, the verse excludes the borrower from acquiring the double payment, whereas the renter and the paid guardian are not excluded. The explanation in the baraita in the Jerusalem Talmud is: Since its safekeeping rests upon him, it is as though it is upon him. We mentioned above that the Tzafnat Pane’ach on Bava Kamma 11a and Ketzot HaHoshen at the beginning of sec. 349 explain that the borrower is the owner of the object, and therefore he uses his own property; the theft from him is not a theft from the depositor’s house. By contrast, in the case of the renter the law is different, because the renter is merely guarding another’s property and not using his own; therefore the object still belongs to the owner, and the double payment goes to him.[56] Above we wrote that according to this baraita it would seem that borrowing effects acquisition while rental does not.

However, in light of our conclusions, this too can be understood differently: both rental and borrowing effect acquisition — and we have already seen that this is indeed the view of Maimonides. The difference between them is that rental acquires for a term, whereas borrowing acquires permanently together with a duty of return. If so, in borrowing the body-for-usufruct belongs to the borrower in perpetuity, and therefore it is as though the object was stolen from his own house. By contrast, in rental the body remains the lessor’s, and the renter has only a usufructuary acquisition for a term; therefore this is considered a theft from the lessor’s house, and the double payment goes to him.

According to the Jerusalem Talmud, the difference between borrower and renter is not because all the benefit is the borrower’s, but because of the difference in the nature of his ownership of the object. If this is Maimonides’ source, then it is very clear why he does not cite the Babylonian Talmud’s rationale that in the borrower’s case all the benefit is his.[57]

It should be noted that Ridbaz there explains the Jerusalem Talmud against our interpretation. He explains that the borrower’s duty to pay in unavoidable accident is not because he violated his guarding, like every other guardian, for the borrower is liable even in unavoidable accident. The borrower’s duty to pay is because he accepted responsibility toward the owner to restore the object, but there is no legal bond between him and the object. Therefore the object is considered the owner’s and not his. By contrast, in the renter’s case the payment is part of the duties of guarding, and therefore there is a bond between the renter and the object (like that of the paid guardian), and therefore the object is considered as though it were the renter’s for the purpose of the double payment.

He thus assumes that the borrower’s liability does not stem from the laws of guardians, but he does not regard the borrower as an owner — quite the opposite. This resembles somewhat the view of Ba’al HaItur that we saw regarding the requirement that something be ‘yours’: the renter is the owner, but the borrower is not. By contrast, the renter is responsible for the object like every guardian, and therefore is considered as though he were the owner with regard to the double payment. This reasoning probably relates to the sugya of collateralized leaven mentioned above, and the matter still requires clarification.[58]

Guarding liabilities at the end of the contractual term

The Shulchan Arukh, Hoshen Mishpat sec. 343, law 2, writes that if the period of the contract has ended, a paid guardian becomes an unpaid guardian so long as he has not returned the object, and in simple terms the same should be true of the renter. By contrast, a borrower, after the end of the term, becomes a paid guardian, not an unpaid guardian (see Minchat Chinukh, commandment 60, sec. 8).

At first glance, this fact indicates that the contract of guarding is not separate from the rights of use, but rather derived from them. Therefore, when the rights of use expire, the situation with respect to guarding changes as well. True, in the case of the borrower it is not entirely clear why he should not become an unpaid guardian, for if he no longer has rights of use, he ought to become an ordinary depositary without payment. It seems one must say that the benefit he derived from the borrowing is considered a form of payment that obligates him in paid guarding after the term has ended (though Rabbi Lichtenstein, at the end of chapter 2, ties this to his own innovation concerning paid guarding without payment as a result of trust). Let us note that according to our approach above, the duty of guarding during the period of borrowing itself is not the result of any payment, but derives from ownership and rights of use.

Terminating the contract in the middle

As for the ability of the guardian to terminate the contract in the middle, borrower and renter can terminate the contract in the middle, whereas the paid and unpaid guardians cannot do so (see Shulchan Arukh, Hoshen Mishpat sec. 293, law 1; Mahaneh Ephraim, Laws of Borrowing and Deposit, sec. 7; and Minchat Chinukh, commandment 60, sec. 6). In simple terms, this too indicates that the guarding liabilities of renter and borrower are derived from rights of use, for the ability to terminate the contract with respect to the rights also terminates the duties.

The imposition of guarding liabilities

As is well known,[59] there is a dispute among the medieval authorities regarding the formal acquisition involved in guarding. Maimonides’ view (Laws of Rental 2:8) is that one does not become liable as a guardian until he performs meshikhah, while the Rosh (Bava Metzia, chapter 8, sec. 15) and Rabbeinu Yitzhak in Tosafot, s.v. ‘They enacted,’ Bava Kamma 79a, hold that the guardian becomes liable from the moment the owner withdraws. The acquisition is needed only so that he may not retract.

What of the renter and borrower? In their case the act of acquisition also has proprietary significance. Minchat Chinukh (commandment 60, sec. 5) assumes that the acquisitions of the borrower and renter depend on that same dispute, and this is also the plain meaning of the Talmud and the decisors.

Accordingly, it is quite plausible that according to Maimonides the duties of responsibility depend on the rights, and therefore the act of acquisition with respect to the rights also imposes the responsibility of guarding. This fits his view, as we explained above, that borrower and renter have a proprietary interest in the object, and that the liabilities of guarding are derived from it.[60] By contrast, according to the Rosh, Tosafot, and those who follow them, the act of acquisition with respect to the rights does not determine the laws of guarding. It is possible that here they too remain consistent with their view that the renter and borrower have no real proprietary interest in the object, and therefore their liabilities of guarding cannot be derived from it, but resemble those of the other guardians.

The renter with the owner

R. Yerucham Perla, in his discussion of sections 23-25, addresses the question whether when guarding is done ‘with the owner’ one can nevertheless become liable for misappropriation, and according to what valuation one pays.

He raises possibilities according to which a borrower ‘with the owner’ is liable for misappropriation, whereas the unpaid and paid guardians ‘with the owner’ are exempt from misappropriation (see there fol. 152b, s.v. ‘Rather,’ and the surrounding discussion, and his explanation of that view on fol. 153a, s.v. ‘And now,’ and elsewhere). He objects to these possibilities from the Talmud in Bava Metzia 96b:

Rami bar Ḥama asked: With respect to a husband regarding his wife’s property, does he have the status of a borrower or of a renter? Rava said: According to your sharpness, your question is mistaken. In any event, whichever way you take it: if he is a borrower, then it is borrowing with the owner present; and if he is a renter, then it is rental with the owner present.

The Talmud thus shows that there is no difference between rental ‘with the owner’ and borrowing ‘with the owner.’ But according to those views, in rental ‘with the owner’ one would be exempt even from misappropriation, whereas a borrower ‘with the owner’ would be liable for misappropriation. He discusses this at great length.

According to our approach, the answer is that the law of the renter ‘with the owner’ is like that of the borrower ‘with the owner.’ Both are liable for misappropriation, and in this they differ from the paid and unpaid guardians. Especially according to our explanation that the renter is a borrower with an exemption from unavoidable accidents, his laws ought to resemble those of the borrower (except with respect to unavoidable accidents), not those of the paid guardian.

Payment in cases of misappropriation

The Talmud in Bava Metzia 43b establishes that in the case of the borrower the payment is assessed according to the value at the time of the unavoidable accident (or theft). By contrast, in cases of misappropriation by the other guardians, even if the object was later stolen or lost, the payment is according to the value at the time of the misappropriation, like a robber who pays according to the value at the time of the robbery, and not at the time of the later accident (see Shulchan Arukh, Hoshen Mishpat sec. 292, law 5).

What happens when the borrower himself misappropriates the object? In simple terms, there is no ordinary law of misappropriation in the borrower’s case, for the deposit was given to him for his use. True, it is possible that use in a manner contrary to the agreement with the lender would count as misappropriation (see above, in our discussion of changes in use). Apparently the medieval authorities dispute this in the sugya of Bava Metzia 41b (see R. Yerucham Perla there, fol. 151a, s.v. ‘And I saw,’ and onward, and Maharit’a, responsum 43, cited there): R. Hananel holds that the borrower can commit ordinary misappropriation, while the other medieval authorities hold that there is no such state in the borrower’s case. According to the majority, when the Talmud uses the expression ‘misappropriation’ regarding the borrower, it means his ordinary use, and in that context it is compared to the unpaid and paid guardians. Yet it is clear that payment in the borrower’s case is according to the time of the accident, whereas in the case of the unpaid and paid guardians the payment is according to the time of the misappropriation.

What, then, is the renter’s law? Here too he would seemingly be like the borrower, since he has rights of use. Therefore, in the renter’s case as well, there is no ordinary concept of misappropriation in its regular sense. If so, the renter too should pay for an accident according to the object’s value at the time of the accident, and in this respect he is not like the paid and unpaid guardians, but like the borrower. True, his guarding liabilities are like those of the paid guardian, but the rights of use that he has cause his laws regarding the valuation of payment to resemble those of the borrower.

Summary: four guardians and three sets of laws

We have explained, according to the view of Maimonides, Rashi, and those who follow them, why there are four types of guardians. The renter is a borrower with an exemption from unavoidable accidents: the foundation of his liability is like that of the borrower (by force of ownership), while the quality of his liabilities resembles that of the paid guardian (with the exceptions dictated by ownership).

According to Tosafot and those who follow them, however, the problem remains intact. If the renter is indeed like the paid guardian, and his liability does not arise from ownership but from an ordinary guarding contract, why does halakhah distinguish four types of guardians?

At the end of chapter 2, Rabbi Lichtenstein proposes a different mechanism. He innovates that there is a type of paid guarding that arises from the trust the depositor places in the guardian, even if there is no payment for the guarding. This is a different kind of paid guarding, and the renter belongs to it.

One should note, however, that according to Tosafot this is difficult to say, for as we saw in Tosafot, s.v. ‘Let us say,’ Bava Metzia 80b, they explain that the renter is liable like a paid guardian because he receives a benefit, and that is akin to wages. According to Tosafot, perhaps the fourth type is a guardian who receives compensation that is not given in exchange for guarding; in such a case it is the Torah that imposes upon him the liabilities of paid guarding (including theft and loss), and not the contract. A similar situation exists in the finder of lost property, whose ‘wage’ is the penny of R. Yosef, and likewise in the artisan (whom Tosafot there explain in the same way as the renter).

This bears upon the remarks of Mahaneh Ephraim, Laws of Borrowing and Deposit sec. 1, mentioned above, who raises the possibility that all liabilities of guardians in Jewish law are not the result of the contract and the undertakings of the parties, but rather liabilities that the Torah imposes upon one who accepted upon himself to enter the category of guardian (though, of course, he can stipulate otherwise). According to that approach, the laws of guardians do not belong to contract law, but are a regular Torah law. According to our discussion here, at least on the view of Tosafot this seems correct with respect to the renter, though not with respect to the paid and unpaid guardians, and perhaps not with respect to the borrower either (if we hold that his liabilities arise from ownership, unlike the renter as construed by Tosafot).

H. Summary

In this article we dealt with the question why halakhah distinguishes four different types of guardians, even though from a legal point of view the renter has the law of a paid guardian. Our point of departure was the apparent assumption that the renter is a composite of two different contracts: he is like the borrower with respect to ownership and rights, and like the paid guardian with respect to guarding liabilities.

However, our conclusion was that the contract of rental, like the contract of borrowing, contains one component and not two. The renter is indeed like the borrower, and in both cases the guarding liabilities (which differ between them) are an outgrowth of rights of use. But their ownership has a different character (in the borrower it is a permanent acquisition with a duty of return, whereas in the renter it is an acquisition for a term), and therefore their liabilities differ.

We noted that Rashi implies that the renter is written in the Torah in the verse (Exodus 22:14), And if it is a hired animal, it comes with its hire, though his specific laws are not detailed there. He is, however, placed in the section of the borrower, and therefore it is reasonable that he is not a paid guardian, but rather a borrower exempt from unavoidable accidents, as we saw in the various sources we cited. Accordingly, the halakhic resemblance between renter and paid guardian is only accidental.[61]

According to other approaches, he belongs to the section of the paid guardian, and therefore the difference between him and the borrower or the other guardians is of a different kind, as explained at the end of the previous chapter.

Our conclusion is that the halakhic differences between the renter and the borrower, on the one hand, and between the renter and the paid guardian, on the other, do indeed justify classifying him as a separate type of guardian. This distinction is not merely theoretical. It also has several practical halakhic consequences, as we described throughout the article, and especially in chapter 7.

[1] True, R. Yerucham Perla, in his commentary to Saadia Gaon’s Sefer HaMitzvot, at the beginning of sections 23-25, raises the possibility that the renter is in fact written in the Torah, in the verse (Exodus 22:14) If it is a hired animal, it comes with its hire, which appears within the section of the borrower. Indeed, Rashi on that verse also says that it deals with the renter (his remarks will be discussed below). So too seems to follow from Maimonides at the beginning of the Laws of Rental cited above (where he wrote: ‘Four guardians were stated in the Torah’).

R. Yerucham Perla there concludes that even Saadia Gaon merely used a conventional turn of phrase, and in practice he too does not count the renter separately. Below, when we see why the renter is a fourth guardian, the obstacle to understanding Saadia Gaon’s language literally will disappear.

In any event, it is clear that the renter’s laws do not appear in the Torah, unlike the laws of the other three guardians.

[2] On this see Rabbi Lichtenstein’s lectures, end of chapter 3.

[3] Presumably the renter is not counted as the principal category of all non-contractual paid guardians, for he differs from them in essence: they become paid guardians because they receive a benefit of monetary value for their guarding, and in that respect resemble the contractual paid guardian; whereas the renter becomes liable in return for the deposit of the animal for his own use, which is a different mechanism of paid guarding. Perhaps this itself is what turns him into a different type of guardian, as will be discussed further below.

[4] Likutei Sichot (vol. 31), by the Lubavitcher Rebbe, Otzar HaChasidim, Brooklyn 1992, eighth series, second volume, beginning of Parashat Mishpatim (based on a talk for Parashat Vayishlach, 1987).

[5] Shiurei HaRav Aharon Lichtenstein – Bava Metzia, HaSho’el, Yeshivat Har Etzion, Alon Shevut 2003.

[6] For additional sources from decisors, responsa literature, and later authorities on this topic, see Pitchei Choshen — Rental, chapter 1, note 2.

[7] See, for example, Kovetz He’arot, beginning of sec. 52 and sec. 53; and Kehillot Ya’akov on Bava Metzia sec. 46, Nedarim sec. 31, Avodah Zarah sec. 9, Shevi’it sec. 1, and elsewhere.

[8] I am offering only a brief survey here. For fuller detail, see Arakhin 20-21; Bava Metzia 56b, 99a, 103b; Bava Kamma 79a; Avodah Zarah 15a; Ketubot 59b; Nedarim 46b; Kiddushin 47b; and the medieval authorities on all those sugyot. Especially important are Ran on Nedarim there, the omitted passages at the end of Rashba’s novellae to Ketubot there (which in fact are by Nachmanides, as is well known), and Ritva on Kiddushin there, among others.

[9] I later found the same in Chazon Ish, Bava Kamma sec. 23, sec. 20, s.v. ‘And Riban answered,’ where he cites this in the name of Ritva. Below we shall see that Ritva in Kiddushin 47b indeed follows Tosafot’s view that the renter has no proprietary interest in the rented house.

[10] See also Tosafot, s.v. ‘An Israelite who rented,’ Yevamot 66b; Tosafot, s.v. ‘One who sells his slave,’ Bava Batra 50b (see the aforementioned Kovetz He’arot, who understood Tosafot this way, though it is not certain); Tosafot, s.v. ‘A husband,’ Bava Metzia 96b (who distinguish between one who purchases for a term and the renter with respect to the laws of guarding). The same can also be proved from Tosafot, s.v. ‘And his fellow,’ Bava Metzia 99a, that the renter has a lien in the rented object (and the readings of Kovetz He’arot and Kehillot Ya’akov in that Tosafot are not necessary, and are even difficult). The same may be proved from Tosafot, s.v. ‘This is what he means,’ Arakhin 21b, and from Tosafot, s.v. ‘And his fellow,’ Bava Metzia 99a; but this is not the place to elaborate.

[11] So too writes Avnei Milu’im, sec. 72, sec. 2 (in the gloss there, letter 9). See also Kehillot Ya’akov, Nedarim sec. 28.

[12] Some distinguish between ‘this house,’ in which case the renter has a proprietary interest in the house, and an unspecified house, where the renter does not. See the omitted passages to Rashba on Ketubot cited above, Tosafot, s.v. ‘Even though,’ Arakhin 20b, the Rosh as cited in Shitah Mekubetzet there, and R. Gershom there.

[13] Kovetz He’arot (end of sec. 53) and Kehillot Ya’akov (Bava Metzia sec. 46) brought proof for this distinction from the sugya of adding to the rent, see there. Perhaps this also resolves the difficulties of R. Akiva Eiger in Gilyon HaShas on Bava Metzia 64b and 79b regarding the Nimmukei Yosef; see there (see the collected novellae of R. Akiva Eiger in the last paragraph before fol. 80a).

[14] See Ran on Nedarim there, who explicitly rejects this. So too the omitted passages to Rashba cited above prove that Nachmanides did not hold this way, and the Rosh in his rulings to Nedarim 46b objected to this distinction. They could instead distinguish between ‘this house’ and an unspecified house, as noted in the previous footnote.

[15] According to Tosafot and those who follow them, what is acquired there is the inability to retract, and perhaps this is connected to the lien that, according to their view, the renter has in the object for his uses. The lien is what is acquired, and therefore the parties cannot withdraw from the transaction.

[16] This also seems to follow from the sugya in Bava Metzia that discusses whether a renter may subrent, which implies that as a matter of law he can in principle rent the object onward to another, and the only question is whether he is permitted to do so.

[17] All this depends on the question whether ownership of the corpus for a fixed term is possible — a matter on which the medieval authorities disputed. See on this Pri Moshe – Kinyanim, by R. Moshe Segal, and more; but this is not our subject here.

[18] Later authorities objected that in Laws of Rental 5:5 Maimonides writes that the renter too may transfer his rights to another. It has been answered that Maimonides also agrees that the renter may transfer his rights to another, though not to someone with a larger household who may damage the house. In such a case the renter is forbidden, whereas a purchaser for a term is permitted (see chapter 5 of the Laws of Rental). See below for an explanation of the difference between them.

[19] A similar idea is cited in Avnei Milu’im, responsum 17, from Shitah Mekubetzet on Bava Metzia, that a woman may eat terumah because she is called property acquired with his money even though she is not acquired to her husband. This is the other side of the same coin: what determines the right to eat terumah is the existence of an act of acquisition concerning the object, and not the fact that one has a proprietary interest in it. Therefore, in rental, where no act of monetary acquisition was performed, he may not feed it terumah, even though the animal is his for its usufruct.

[20] In Kovetz He’arot there he remains in doubt concerning Nachmanides, and did not notice the opposite contradiction in the Ra’ah. See there also his question against the Mordechai, and his difficulty for the medieval authorities here as to how rent accrues from beginning to end. According to our approach this is very straightforward, since the payment is for the uses and not for the acquisition. The acquisition in the body of the object can be effected by meshikhah and the like.

[21] See Iggerot Moshe, Hoshen Mishpat sec. 74.

[22] One might perhaps have said that on the side of rental, the fact that the renter acquires is obvious, because he paid money. But in borrowing it is not obvious why there should be a side that he acquires, and therefore Rashi says that the borrower too ‘pays’ something to the lender: liability for unavoidable accidents.

But this explanation is strained, for several reasons: (1) if this were indeed Rashi’s intention, he should first have asked why borrowing should acquire in contrast to rental, and only then answered that it is because of liability for unavoidable accidents. Instead, Rashi presents this as the fundamental reason for the borrower’s acquisition. (2) Beyond that, why should acquisition without payment be difficult to conceive? Even in permanent acquisitions there are gifts and sales; a gift too fully transfers the object to the recipient. (3) More generally, it is difficult to view liability for unavoidable accidents as the borrower’s ‘payment.’ First, this is an undertaking and not something of monetary value. Second, it is only an elementary undertaking in return for the favor the lender is doing him. The lender gains nothing from that undertaking as compared with a situation in which he would simply not have lent the object at all. How, then, can this be considered ‘payment’ for the borrowing? This is unlike rental, where the money remains in the lessor’s hands, so he has gained something concrete from the transaction.

[23] In the previous note we explained why it is implausible to see liability for unavoidable accidents as ‘payment’ for acquisition. The conclusion is that liability for unavoidable accidents is an indication, not a cause. See further our remarks below.

[24] A parallel distinction appears in the Babylonian Talmud, Bava Metzia 34a; see below.

[25] See Tzafnat Pane’ach on Maimonides, Arakhin 6:22, s.v. ‘But if’ (fol. 46a in that edition), and in his novellae to Bava Kamma, part 1, on fol. 11a, s.v. ‘The Gemara said in the name of Samuel’ (pp. 13-14 in that edition), and part 2 on fol. 112a, s.v. ‘The Gemara: their father left them’ (p. 186 in that edition).

[26] True, at first glance he might be able to return the broken remains as something of monetary value within the overall compensation. See Tosafot, s.v. ‘We do not appraise,’ in that sugya in Bava Kamma.

[27] In Likutei Sichot, the Lubavitcher Rebbe also wrote this way (see there p. 113). In note *16 he refers to the novellae of the Tzemach Tzedek on the Talmud, Ketubot, beginning of chapter 5, s.v. ‘Rather, he subjects his body.’ More from his remarks will be cited below.

[28] However, the Tzafnat Pane’ach’s words on this are very puzzling, for in the case of the robber it is obvious that there are robbery-acquisitions in the object, and the Talmud itself says that it is acquired to him in order to render him liable for unavoidable accidents. That is, the very principle of the Tzafnat Pane’ach regarding the robber is almost explicit in the Talmud, and yet specifically regarding the robber he does not say it, but only regarding the borrower. We already noted that Tosafot in the sugya of Bava Kamma 11 do indeed say so even regarding the robber; but specifically with regard to the borrower, for whom appraisal is the halakhah, it is not clear whether Tosafot say this. We also saw above from Sanhedrin 72a, which compares the borrower to a burglar regarding liability for unavoidable accidents and acquisition.

[29] So too follows from the Ra’avad cited in Shitah Mekubetzet on Bava Metzia 80b, who wrote that the artisan is like the renter because he performs work on the body of the object.

[30] See R. Yerucham Perla there, fol. 152b, where he discusses whether in a gift on condition to return the recipient bears liability for unavoidable accidents. There he explains that borrowing ‘with the owner’ is like a gift on condition to return. We, of course, are speaking here about every borrower.

[31] What we say applies both to the view of Ketzot HaHoshen and to that of Netivot HaMishpat regarding a gift on condition to return. Their dispute concerns only whether the acquisition is perpetual or for a term, but according to both of them the nature of the acquisition can be as described here.

Still, with respect to the borrower it seems more plausible not to view the duty of return as a condition of ownership. A duty of return rests upon the borrower at the end of the borrowing period. It does not condition his ownership, but it nevertheless exists as a duty upon him. In the renter’s case there is no such duty.

[32] Even if the unavoidable accident occurred within the term. Until the end of the term it is the renter’s ‘fortune,’ and therefore the lessor need not provide him with another animal. But at the end of the term it is the lessor’s ‘fortune.’ One must also distinguish here between ‘this donkey’ and an unspecified donkey. See Bava Metzia 79a (regarding a donkey) and 103a (regarding a house).

[33] According to what we have said, the difficulty of Nachmanides in Milhamot to Shevuot (25a in the pages of the Rif, cited by the Shakh, Hoshen Mishpat sec. 72, sec. 9, and in Ma’ayanai HaChokhmah, sec. 74) against Rashi is likewise resolved. Rashi held that a creditor holding collateral acquires it by virtue of R. Isaac’s law even with respect to unavoidable accidents. Nachmanides objected from the sugya ‘all artisans are paid guardians’: even according to the view that an artisan acquires through the improvement of the vessel, he is nevertheless not liable for unavoidable accidents. According to our explanation, this is very straightforward: in the case of collateral he acquires it like a borrower (as a gift on condition to return), and is therefore liable for unavoidable accidents. But the artisan, according to the view that he acquires through improvement, acquires it like a renter (as an acquisition for a term), and is therefore not liable for unavoidable accidents.

[34] As we saw, the view of Tosafot and those who follow them is that the renter did not acquire, that is, he has no ownership. If so, how do they explain the similarity between the renter and the artisan? Indeed, in Tosafot, s.v. ‘Let us say,’ Bava Metzia 80b, the artisan is compared to the renter because both receive payment, even though that payment is not given for guarding (proof for this can be brought from the penny of R. Yosef in the case of the finder of lost property, where too the ‘payment’ is not for guarding, and nevertheless he is a paid guardian). The explanation of this is found in Tosafot HaRosh there, and this is not the place to elaborate.

[35] See Laws of Sale 23:2, and the references brought there in the Frankel edition in Sefer HaMafte’ach, first entry, from Chakrei Lev; the discussion is well known.

[36] The Tzafnat Pane’ach explains this by saying that the renter or borrower there is a gentile, who does not acquire. In the case of an Israelite, the borrower acquires while the renter does not. But this is forced in both the Talmud (which brought proof from feeding a priest’s animal with terumah to the case of renting to a gentile) and in Maimonides.

[37] True, see Shulchan Arukh, Hoshen Mishpat sec. 207, law 5; but it does not appear to have a source. Also, betrothal by means of a gift on condition to return is seemingly like a sale on condition to return, but betrothal is different, for he does not acquire the woman herself. In any event, as a general rule what we said is certainly correct.

[38] Of course, all this is in the ordinary case. If they explicitly define the transaction otherwise, they may perhaps be able to do so.

[39] According to this, the Rivash, sec. 510, cited above, who requires the object to be reconveyed to the depositor, is speaking only of rental and not of borrowing. Indeed, the case there was rental.

[40] Theoretically, one might say that the borrower has two bodies of law: (1) duties of return, which obligate him in negligence, theft, and loss, and even in unavoidable accidents; and (2) duties of guarding, which obligate him like a paid guardian. But there is no reason to posit such duplication. If ownership is capable of generating all the guarding liabilities of the borrower, why assume that he has an additional source of liability as well (a double contract: guarding liabilities and ownership)?

[41] According to our discussion, it is possible that according to the views of Saadia Gaon and Tosafot Rid cited in R. Yerucham Perla above, the renter too would be exempt where the animal died in the course of its work by means of only a rabbinic oath, and not by the oath of guardians, for like the borrower he is not really a guardian. True, this is somewhat forced, since it would have implications also for misappropriation and for a renter ‘with the owner,’ and in any event these views are a minority among the medieval authorities (see R. Yerucham Perla there); so we leave the matter aside.

[42] I saw his remarks only after completing the writing of this article. Below I shall briefly note some differences between his conclusions and those presented here.

[43] There is a law that an artisan acquires through the improvement of the vessel, and that a creditor acquires collateral. But one must distinguish among different cases (whether the collateral was taken at the time of the loan or not; whether it was taken by the court, through the court, or by the creditor himself outside court; and more), and this is not our subject here. We should note that even if the liabilities of these two, or one of them, derive from ownership, there is still no need to define for them a different type of guarding, because in such a case they would fall under the renter (as derivative cases of that category. See my article, ‘Two kinds of ma ha-tzad: conceptual construction,’ Meysharim 2, Yeshivat Hesder Yeruham, 2003).

[44] True, in the Commentary on the Mishnah to Bava Metzia 7:9 it appears somewhat that the renter is included in the section of the paid guardian (see Rabbi Lichtenstein’s lectures, chapter 2, note 5), and this can be pushed aside, but it still requires further analysis.

[45] See also Ketzot HaHoshen, sec. 72, sec. 34, and Minchat Chinukh there, where they discuss whether these statements were said only according to R. Yosef’s explanation of the law of the finder of lost property, or also according to Rabbah.

[46] Even if this benefit is not payment for guarding, he is like the finder of lost property, who receives a benefit not in exchange for guarding and yet, as halakhah, is treated like a paid guardian.

[47] Rabbi Lichtenstein does not present this as a liability that stems from ownership, but as a guarding liability that stems from the special trust the owner places in the guardian and not from a guarding contract (he also proves this from Rashbam on Exodus 22:6). According to our discussion, it is entirely possible that the Ra’avad is following Rashi’s line.

[48] I should note that, in my opinion, even his inferences from the plain meaning of the Talmud’s language (see there around notes 8-9) are not correct, and his explanation of the sugya (see there p. 114) is also not precise, as emerges from our discussion above in explaining the sugya, and likewise from the Rosh and Maharam cited above; but this is not the place to elaborate.

[49] We hinted above that this is not entirely clear. It is possible that this is only a liability of responsibility and not a duty of guarding. A paid guardian may perhaps be liable to pay for theft and loss without there necessarily being a duty to guard specifically against them. See on this Rabbi Lichtenstein’s lectures, chapters 1 and 3.

[50] The problem is not specifically the formal act of acquisition, but the fact that there is no enforceable contract for less than a perutah.

[51] The proof is that even the Talmud in Avodah Zarah, which discusses whether rental and borrowing acquire or not, does not bring evidence from the sugya of collateralized leaven.

[52] Some medieval authorities there (see Rashi and Tosafot, s.v. ‘According to the first version of R. Pappa’) explained that according to that version of R. Pappa, in the borrower’s case even if he actually paid he still did not acquire the double payment.

[53] See Mahaneh Ephraim, Laws of Borrowing and Deposit sec. 1, who investigates whether the liabilities of guardians are the result of undertaking, or rather liabilities imposed by the Torah. He discusses the meaning of the expression ‘all the benefit is his,’ and concludes that it does not mean that we are dealing with reciprocal undertakings. These are merely the reasons why the Torah made the borrower liable for unavoidable accidents. This interpretation is very close to what we have said here.

[54] True, the Jerusalem Talmud does not mention that the case there is one where the guardian merely said he would pay and did not actually pay. Nor does it explicitly state that the issue is the acquisition of the double payment by the paying guardian. Nevertheless, the commentators there explained the Jerusalem Talmud in this way.

[55] See also Mareh HaPanim there.

[56] It is not entirely clear from the words of Ketzot HaHoshen why this distinction leads to the owner’s acquisition of the double payment, and the matter could be resolved; but that is not our concern here.

[57] As is well known, one rule in interpreting Maimonides is that at times he rules like the Jerusalem Talmud against the Babylonian Talmud.

[58] It is possible, however, that he ties this only to ownership of the corpus and not of the usufruct; but then the renter’s case would still require clarification.

[59] See Minchat Chinukh, commandment 57, sec. 5; commandment 59, sec. 2; and commandment 60, sec. 5; and Mahaneh Ephraim, Laws of Guardians sec. 7.

[60] True, Maimonides says the same even regarding the unpaid and paid guardians, who certainly have no proprietary interest in the object, and therefore this is not conclusive.

[61] See the aforementioned Likutei Sichot, which notes this as well (note 12 and its context). There he refers to the Levush, Hoshen Mishpat at the beginning of sec. 291.