With God's help

Misharim – 5765

One of the central tools of analytic study is the 'inquiry.' In yeshiva terminology, this usually means an examination of some law, intended to clarify which of two explanations of it is correct.[2] Such a scholarly inquiry assumes that there are two possible ways of explaining the law under discussion, and we must determine which of them fits the data. A basic assumption implicit in this kind of inquiry is that only one of the two explanations is correct. In this article we will try to examine that very assumption. In other words: we will examine the proposal that there is, in fact, a third explanation that is the correct one, and that it is constructed as a logical conjunction of the first two explanations.

There are several foundational inquiries that serve as cornerstones of yeshiva-style analysis. As we shall see, in each of them the inquiry itself, as well as the question we ask here, may take on a somewhat different character. We will begin the discussion with one of the best known among them: the ground of liability for damage caused by one's property.

A. The Ground of Liability for Damage Caused by One's Property

The Inquiry of the Later Authorities and the Dispute between the Chazon Ish and the Pnei Yehoshua

The later authorities discuss whether a person's obligation to pay when his property causes damage stems from negligence in guarding it, or whether it derives from the very fact that his property caused damage. Let us note that according to all views, both conditions are required: that the damager be his property (or at least under his responsibility, in some sense), and that there be negligence in guarding it. The question under investigation is which of these is primary, and which serves only as a secondary condition for liability.[3]

Several halakhic ramifications are presented as consequences of this inquiry, but in most of them that connection can be challenged. One of the clearest among them is the question upon whom the burden of proof rests in a case of doubt. On this matter the Pnei Yehoshua and the Chazon Ish disagree.

The Chazon Ish, Bava Kamma §7:7, writes that the burden of proof rests upon the damager. He must prove that he guarded properly in order to exempt himself.[4] By contrast, the Pnei Yehoshua, Bava Kamma 56b, s.v. In the Talmud: let us say that it should be, writes that the burden of proof rests upon the injured party. He must prove that the owner did not guard properly, since he comes to extract money from the damager.

Several later authorities understood that this dispute depends upon the above inquiry.[5] The Chazon Ish, who holds that the burden of proof rests upon the damager, apparently maintains that once his property caused damage, he is liable to pay. True, he can exempt himself by claiming that he guarded properly, but that is a claim of exemption, and therefore the burden of proof rests upon him.[6] By contrast, the Pnei Yehoshua holds that the burden of proof rests upon the injured party, and apparently in his view the known fact that his property caused damage is not, in itself, a sufficient basis for liability. In order to prove an obligation of payment, he must also prove that there was negligence in guarding it. Therefore the injured party is the extractor, and the burden of proof rests upon him.[7]

Thus, according to the explanations of these later authorities, the Chazon Ish's position is that the very fact that his property caused damage obligates payment, whereas proper guarding is only exculpatory. The Pnei Yehoshua's position is that negligence in guarding is what obligates payment, while the fact that it is his property is merely a condition for the duty of guarding to fall upon him.

An Apparent Contradiction in the Pnei Yehoshua's Position

Yet we find in the Pnei Yehoshua on Bava Kamma 24b a truly astonishing innovation. The Talmud there establishes that one who sets another person's dog upon someone else is exempt. With respect to the owner of the dog, the Talmud there discusses whether to obligate him or not. The Pnei Yehoshua ibid. writes that according to the side that the dog's owner is also exempt when the dog was set on, this the law applies even if the one who set the dog on was the owner himself (and later authorities have already expressed surprise at this. See Otzar Mefarshei HaTalmud here, note 99, and Ayelet HaShachar ad loc.). I later saw that the Chazon Ish as well, Bava Kamma §5:7, linked this to two formulations in Rashi later on 53b.

These remarks of the Pnei Yehoshua and the Chazon Ish are truly baffling. If a person does not guard his animal and it causes damage, he must pay. But if a person sets his animal on someone and thereby directly causes it to do damage, he goes free. There could hardly be a greater negligence in guarding than this. If passive negligence (failure to guard) obligates payment, then active negligence (setting it on) should all the more obviously obligate payment.

In other words: if a person sets another person's dog upon someone else, then there is room to exempt them both from payment. The instigator was negligent in guarding, but the vicious dog that caused the damage is the owner's property (and therefore the owner should not have kept it). Each of the two sides has one liability-generating parameter, but in order to obligate payment we need both of the parameters mentioned in the above inquiry: that it be his property and that there be negligence in guarding. But when the instigator is the owner, then one and the same person satisfies both conditions. If so, what is lacking in order to obligate him?

It seems that in order to understand this surprising innovation, even if only somewhat forcedly, we have no choice but to assume that what generates liability for damage payments is the fact that one's property caused damage. Alongside this, one may say that there is a duty of compensation resting upon the damaging property itself—the animal or other creature that caused the damage—except that such obligations are transferred, once they arise, to the owner of the damaging property.[8] If so, we must ask whether a duty of compensation falls on the dog itself (rather than on its owner) in this case. If it does, we will transfer it to the owner and obligate him to pay.

The Pnei Yehoshua apparently understands that the dog itself is exempt, since someone set it on, and it is not to blame; rather, the blame lies with the instigator. The dog is coerced by someone else, and therefore no obligation rests on it. But if no payment obligation arose for the dog, then automatically there is no obligation that can pass to the owner. True, the owner is guilty, but he is only an indirect cause of the dog's damage, and therefore on his own account (that is, under the law governing a human being who causes damage) we have no way to obligate him. And from the perspective of the dog no obligation arose, since it was coerced, so we also have no way to transfer liability to the owner under the law of damaging property. For this reason the Pnei Yehoshua exempts the instigator. As noted, the Chazon Ish also raises such a possibility.[9]

We must now return to the above dispute between the Pnei Yehoshua and the Chazon Ish in the passage regarding the burden of proof in cases of doubt. There we saw that the Chazon Ish's position is that what obligates is the very fact that his property caused damage, while negligence is only a condition for liability. If so, the ruling here accords with his position there.[10] But the Pnei Yehoshua's position is that what obligates payment for one's property having caused damage is negligence in guarding, and therefore the burden of proof rests upon the injured party. Yet here it is necessarily demonstrated that his position is that what obligates payment is the very fact that his property caused damage. The Pnei Yehoshua thus appears to contradict himself.

Resolving the Contradiction: Giving Up the Dichotomies

The obvious solution to this problem is that the Pnei Yehoshua's position is that both parameters together are required for liability to pay in cases of damaging property: both that his property cause damage and that there be negligence in guarding. One should note that such a position fully resolves the above contradiction. Regarding the burden of proof, even if both elements are required, it is clear that the burden of proof rests upon the injured party, since in the final analysis, even if only one of them is missing, it has not been proved that the condition of negligence in guarding—which is also required in order to obligate—was fulfilled. By contrast, regarding one who sets his own dog upon someone else, there we understand the liability as falling upon the dog that caused the damage, and such liability never arose at all; therefore the instigating owner is exempt.

Let us emphasize that the logic behind such a position in cases of damage caused by property is very clear. The fact that the property caused damage is required in order to establish liability on the damaging animal itself (as we saw, this is a liability resting upon the dog). But when we wish to transfer the liability to the owner, that requires justification in the form of negligence in guarding on his part. Thus both conditions are required, but each has a different role from the other.[11]

Accordingly, it is clear why, according to the Pnei Yehoshua, no liability arose for an owner who set his dog upon someone else. No liability arose on the dog itself, because the condition that the dog cause damage through its own 'negligence' was not fulfilled. Consequently, even if the condition of negligence in guarding was fulfilled—the condition that allows liability to be transferred to the owner—here no liability at all arose on the dog, and therefore there was nothing to transfer to the owner. For this reason, in such a case the owner is exempt from payment.

The inquiry above assumed that these two conditions are required in order to obligate someone for damage caused by his property. He must be the owner, and he must have been negligent in guarding. We saw that according to each side of the inquiry both conditions are required, and the question was only which of them is primary and which is secondary. If so, the same basic logic that led to these two conditions appearing in the inquiry at all can also lead to the conclusion that both of them are conditions of liability to the same degree. This is the point of departure for the question we will ask below: if so, why did we assume in the first place that only one of these two sides is correct, and only retrospectively, because of pressure, are we prepared to accept a position that sees them both as conditions of equal status? In other words: why did we see any problem in the Pnei Yehoshua's position at all?

Before we move on to the general methodological question, let us look at an example in which the Pnei Yehoshua himself does not accept such a conjunction of two sides of an inquiry.

B. This one benefits and that one does not lose ('one benefits and the other does not lose')

In the Bava Kamma 20a passage, we find an example of a 'yeshiva inquiry' carried out almost explicitly by the Talmud itself. The Talmud discusses the case of a person who lives in another's courtyard without the owner's knowledge: must he pay the owner of the courtyard or not? There are several possible cases to discuss here. When the courtyard is available for rent (is up for rent), the owner of the courtyard incurs a 'loss' from the intrusion. When the occupant needs lodging and cannot find a place for free (a person who is apt to rent), he is defined as one who 'benefits.'

The Talmud's hesitation can be presented as follows:[12] In the case of 'one benefits and the other loses,' it is clear that the intruder must pay. There two conditions are met: the occupant benefits, and the owner also incurs a loss. And in the case of 'this one does not benefit and that one does not lose,' it is obvious that there is no obligation to pay. There neither of the conditions is met. But we must ask whether in the first case payment is required because the owner of the courtyard suffers a loss, or because the occupier himself benefits. The practical ramification that the Talmud discusses directly is the case of 'one benefits and the other does not lose.' In this case there is benefit but no loss. If what obligates is the benefit, then here too we should obligate the intruder to pay. But if what obligates is the loss suffered by the owner of the courtyard, then here he is exempt (for there is no loss). This is a classic practical ramification of the yeshiva-style inquiry carried out by the Talmud itself, and in this case the Talmud itself is formulated in an almost fully 'Acharonim-like' manner.

Already among the medieval authorities, and certainly among the later ones, one can see different formulations of the hesitation, and as noted they also discuss whether this is indeed the Talmud's hesitation at all. But in light of what was said above (regarding the dichotomy of the inquiry into what generates liability in damage payments), it is especially interesting to consider the Pnei Yehoshua's words in this passage. Let us examine this specifically in the fourth case, one not discussed at all by the Talmud: 'this one does not benefit and that one loses.'

As a practical ruling, we hold that in a case of 'one benefits and the other does not lose,' the occupant is exempt. Seemingly, this is a conclusion that what generates payment is not the occupier's benefit but rather the owner's loss.

And in Tosafot, s.v. This one does not benefit on 20a, they ruled that in a situation of 'this one does not benefit and that one loses,' the occupant is exempt from payment (the Rif and his school disagree with them on this). Seemingly, it follows from this that what obligates is benefit rather than loss. But if so, it is not clear why we rule as a matter of Jewish law that in a case of 'one benefits and the other does not lose' he is exempt. If benefit is what obligates, we should have held him liable here, as explained above.

Now the Pnei Yehoshua there, s.v. 'In Tosafot,' is uncertain about this. He formulates it as a doubt regarding the case of 'one benefits and the other loses': why is the intruder liable to pay there? From Tosafot's position it emerges that loss alone does not obligate payment, and from the halakhic ruling regarding 'one benefits and the other does not lose'—that he is exempt—it emerges that benefit alone likewise does not obligate payment. The Pnei Yehoshua writes as follows:

And it is forced to say that since this one benefits, that helps to render him liable for the loss on that account; even though it is only indirect causation, the two rationales combine to obligate him, even though with either one alone he would not be liable. But this is forced… (And it is forced to say that since this one benefited, that assists in obligating him on account of the loss; even though it is only indirect causation, the two considerations combine to obligate him even though neither one by itself would obligate. But this is forced…)

Thus, the Pnei Yehoshua refuses to accept a solution that proposes that two parameters, neither of which is sufficient by itself to impose liability, can combine to create an obligation of payment.

Seemingly, this stands in tension with his position, as it emerges in the passage about the factors that generate liability for damage caused by one's property. Above we saw that there he takes the view that two parameters, each of which is insufficient by itself to obligate, combine to create an obligation of payment.

Let us note that the Pnei Yehoshua himself proposes another solution in the passage of 'one benefits and the other does not lose' (and he links the passage to the question of the 'measure of Sodom'). However, quite a number of medieval and later authorities specifically adopt the very approach rejected by the Pnei Yehoshua—namely, that the conjunction of the two factors is what obligates. For example, R. Shimon Shkop, in Shiurei Bava Kamma §§19–20, proposes in explaining the view of Tosafot cited above that what obligates is the benefit, but there is a condition that there be loss. In other medieval authorities the opposite seems to appear: what obligates is the loss, but there is a condition that the one causing the loss benefit.

Seemingly these are solutions that combine two different parameters in order to create liability. But if we look carefully at the words of the commentators, it appears that there is not really a true conjunction here. Each commentator chooses a dominant parameter that obligates payment, and treats the second as a secondary condition. For example, R. Shimon Shkop, in his understanding of Tosafot, argues that what obligates is the benefit, except that there is a condition for liability that there be loss.[13]

It seems to me that a more careful examination of the passage and the commentators reveals that nevertheless there are some who do not mean such a conjunction, but rather a conjunction of equal status; but this is not the place to elaborate.

C. The Problem of Dichotomies in Yeshiva Inquiries

Thus far we have seen two inquiries in both of which the basic assumption was that one of two explanations is correct. Thereafter another possibility emerges: perhaps a third explanation, which is in some sense a conjunction of the first two, is specifically the correct explanation. This is an expression of the general question we set at the outset: why is the simple assumption that the inquiry is dichotomous?

It seems, however, that there is a difference between the two examples brought above. With respect to what generates liability in cases of damage caused by property, the inquiry could have been formulated in a more starkly dichotomous way: is what obligates negligence in guarding, or the very fact of ownership? In that formulation, there is here an 'inquiry' in the sense described above: which of the two explanations of the law is correct. But above a more accurate formulation was presented, according to which both elements (my property that caused damage and negligence in guarding) exist in each of the two explanations, and the question is which of them is the primary liability-generating element and which serves only as a secondary condition. In this formulation, the inquiry is not exactly of the classic type—namely, which of two explanations is correct—but rather: which of two elements within a given explanation is dominant and which is secondary. Many yeshiva inquiries are of precisely this kind.

If so, the examples brought above represent two different types of inquiry. But in the end even the second (more accurate) formulation of the inquiry into what generates liability in damage caused by property produces two explanations that compete with one another: 1. What obligates is my property that caused damage, with negligence in guarding as a condition. 2. What obligates is negligence in guarding, except that there is a condition that it be my property (I am only required to guard my property). The inquiry now becomes one in the ordinary sense, asking which of these two (complex) explanations is the correct one.

However, it seems that despite the similarity, the difference between the two types of inquiry nevertheless has an implication, specifically with regard to the alternative proposal we are examining in this article. The proposal that there is a third explanation that is the correct one, and that it is a conjunction of the two dichotomous explanations, is not valid in inquiries of the second type. In the non-dichotomous formulation, the complex explanation is not a simple logical conjunction of the two simple explanations, but the creation of a third explanation, different from the previous two. Instead of liability being generated by a dominant element and a secondary element, one proposes an explanation according to which both elements are dominant. This is not a simple logical conjunction of the two original explanations.

By contrast, if the inquiry were which element is the determining one regarding liability (as the Pnei Yehoshua understands in the passage of benefit and loss), then the proposal that the determining parameter is the conjunction of the two elements—benefit and loss—would be a simple logical conjunction of the two sides of the inquiry.

True, one may say that even in the non-dichotomous formulation there is a problem from a similar family, since there is a question which element generates liability (the second being only a secondary condition), and the alternative proposal is that both generate liability to the same degree. That is, there is a dialectic here, except that it concerns not two explanations but two elements within an explanation. Between the two elements there is a structure similar to that of an inquiry of the first type: initially we assume that only one of them generates liability, and afterward we propose that only both together generate liability. We will return to this point below.

D. what is common to the two cases?

There is a question that looks very similar to the one discussed here: the structure of the hermeneutical rule known as what is common to the two cases?.[14] By means of this rule we derive a new law from two known laws. Neither of them alone can teach us the additional law, because each source has a unique feature that does not exist in the case being learned (and therefore generates a refutation), and so we derive it from the conjunction of the common denominator shared by both.

This structure is related on several planes to our present discussion. First, the mode of derivation itself combines two factors, each of which by itself cannot do the work, and extracts the result (= the new law) from their conjunction. But beyond that, this mode assumes a logical premise that seems highly problematic. The question is why we should not say that in truth the new law cannot be derived even from the conjunction of the two source laws. If each of them has a unique feature that does not allow us to infer the conclusion regarding the new law, why is it not correct to refute in the same way even the derivation from both sources together? Seemingly, it could be that the law in source A depends on its unique feature, and the law in source B depends on its unique feature. In other words, why do we assume that the law is generated by the common denominator of the two sources, and not by each of their unique features separately? If the law is generated by each of the unique features, then in the new case—which is characterized by neither of those two features—we cannot derive the law with respect to it.

The obvious conclusion is that every law in the Torah derives from one reason only.[15] From the very structure of what is common to the two cases?, we implicitly reject the possibility that a given law can derive from either of two different reasons. Seemingly, here we have a source for the assumption underlying the yeshiva inquiry: that there are not two explanations for one and the same law.

But on the face of it this is incorrect. The case of what is common to the two cases? is logically different from the case of the yeshiva inquiry. In the question of what is common to the two cases? we examine the possibility that a given law may be derived from either of two different reasons. That is, there are two different factors, each of which alone can determine the same law. Such a possibility is indeed ruled out by the rule of what is common to the two cases?. But in our case we are examining another possibility: granted that there is one explanation for the given law, why do we assume that it contains only one of the two relevant elements, and not both of them?

In other words: the problem of what is common to the two cases? addresses whether to unite two explanations by performing a logical OR operation between them, whereas our problem asks why not to intersect them by performing a logical AND operation. In what is common to the two cases? we ask why there should not be two possible explanations, whereas here we ask why both explanations should not be required together as one comprehensive explanation of the law under discussion.

Had we asked why not think that loss too is a sufficient factor for liability, and that mere benefit can also generate liability, that would have been a problem relevant to the proof from what is common to the two cases?.[16] But here we are asking something else: why both should not be required together.

In fact, the derivation of what is common to the two cases? seemingly leads to the opposite conclusion. Even when it appears that a given law has two different explanations, we immediately look for the common denominator between them—that is, some conjunction of them that is what generates the law. If so, why should we not assume in yeshiva inquiries as well that the conjunction of the two elements/explanations, or some common denominator between them, is the correct explanation?

In an earlier article (Misharim 2) I argued that the principle of what is common to the two cases? is not only an interpretive rule relevant to the interpretation of Torah and Scripture. We saw there that it is a more general form of halakhic thought. When we have two halakhic elements, we can combine them and create a third concept made up of both, and draw halakhic conclusions from it. There we saw examples of constructing a new form of labor by combining two different primary categories of labor on the Sabbath (spitting as a conjunction of carrying four cubits in the public domain and winnowing). If so, here too we should apparently have done the same thing and assumed that if there are two elements relevant to the problem, we ought to combine them and create a complex explanation.

E. The Problem of Simplicity and Elegance in Philosophy of Science and in Halakhic Interpretation

At first glance it seems that there is a universal problem here. A theory with one component is simpler than a theory with two components. In other words, the simpler explanation appears to us more likely to be the correct one. This is a general human form of thought, and not necessarily a distinctive feature of halakhic thought. This is what is called in philosophy of science 'the problem of the elegance and simplicity of scientific explanation (theory).' We shall see below that the identification of the problem of dichotomies with the problem of elegance and simplicity is not so simple.

Simplicity and Elegance in Philosophy of Science

The problem of simplicity in philosophy of science asks: why do we in fact assume that the simpler explanation is also the correct one? In the context of the natural sciences, the question of simplicity and elegance has embarrassed philosophy of science for several generations. There are countless ways to generalize any given collection of experimental results into a general theory. Philosophers of science ask themselves in this context two questions: 1. Why are the generalizations we choose indeed the simplest ones (what is the criterion of simplicity)? 2. Why is the simplest generalization indeed the correct one (what is the connection between simplicity and truth)?[17]

For this reason, and several others, many philosophers of science conclude that scientific theory really contains no content beyond the collection of facts it generalizes. Theory is nothing more than a sophisticated description of a set of facts, and all theories that fit that set of facts are equally correct. Theory is a human cognitive tool, not a deeper acquaintance with the world. According to this approach, the theoretical entities defined within the framework of a theory are merely fictions that make convenient description of the world possible, not actually existing or real entities.

Accordingly, the choice of the simple and elegant theory is obviously the natural one. Not because it is truer—for all theories are equally true—but because it is the most convenient. If there is no difference in truth between theories, why not choose the simplest and easiest to use?

According to such a line of thought, analytic explanations of Jewish law likewise do not reveal something about the nature of Jewish law, or of the Torah. Rather, they are tools used by our thought in order to organize the totality of halakhic phenomena (the various laws, or the various opinions in Jewish law). Accordingly, there is no room for the question why we choose the less complex explanation. That is the explanation most convenient to use, and therefore we choose it. Such an explanation is not a claim about Jewish law, but a tool by which we grasp it. To the same degree, we could have used any other explanation (see below around note 23, where I reject this possibility).

In the second part of my book Shtei Agalot VeKadur Porach (Two Wagons and a Hot-Air Balloon: On Science, Religion, and Myth) I show that this approach in philosophy of science is fundamentally untenable. The fact that scientific theories are confirmed by experiments after they are proposed leads to rejection of this suggestion. But if so, the difficulty still remains: why is there a connection between simplicity and elegance and truth?

In that book I showed that there are two principal ways to solve this problem: 1. The accepted approach, according to which there is in fact no connection between simplicity and truth. According to this approach, theories are claims about us and not about the world. It may be that simplicity is not really a concept that can be clearly defined, and perhaps it does not exist at all (though even this is not necessary according to that approach). 2. There is such a connection, and it stems from a correspondence that exists between us and the world. According to this approach, simplicity means conformity to the forms of human thought (and, as noted, even those who hold the previous approach may agree to this). What accords with those forms appears simpler to us. Truth derives from a correspondence between the mode of operation of the world and our forms of thought. What seems simple to us describes the world correctly. Thus the Creator of the world and of humanity willed it.

Simplicity and Elegance in Jewish Law

As noted, we find similar approaches with respect to Jewish law as well. Like all human thought, it too apparently assumes a connection between simplicity and elegance and truth. Let us now see several examples of this.

The first example is the common rule in the study hall regarding the interpretation of disputes among Tannaim and Amoraim: we do not multiply disputes.[18] This rule states that we do not multiply disputes among the sages beyond what is necessary. In our analytic give-and-take, an explanation that leads to there being two disputes between the parties is rejected in favor of an explanation that reduces the dispute to a single principle.

According to this rule, one can derive practical Jewish law from the words of one of the disputants, even from one whose view was not accepted as the law. The assumption is that if the principle inferred is not connected to the explicit dispute between them, then his interlocutor likely agrees to this point as well. Someone who wishes to reject such an inference will have to show that the conclusion does depend on the disputed principle between the opinions (that is, that it is not independent). In halakhic discussion one cannot reject such an inference merely on the basis of the claim that it is based on the words of the sage whose opinion was not accepted in practice.

For example—one among very many—the Ran on the Rif in tractate Yoma (4b in the Rif pagination, s.v. and we read) brings a responsum of the Raavad regarding a sick person on the Sabbath for whom a physician determined that he must eat meat, but he has only the meat of an unslaughtered carcass, which is not kosher. The question is whether one should slaughter an animal for him on the Sabbath—which is a very severe prohibition—or feed him carrion, which is a less severe prohibition. The medieval authorities disagreed on this question, and some wrote that he should be fed carrion because that prohibition is lighter. The Ran argues that one should slaughter a live animal for him, for although the prohibition of slaughter on the Sabbath is more severe, it is only one prohibition. By contrast, eating carrion involves a prohibition for every olive-sized piece of meat the sick person eats.

The later authorities infer from the Ran's words the principle that eating several measure-amounts of forbidden food incurs several punishments—one for each measure. This is so even though the Shulchan Arukh did not rule the Ran's position as practical Jewish law. See, for example, Kovetz Shiurim, by R. Elchanan Wasserman, tractate Beitzah, sec. 48, and Tosefet Yom HaKippurim, by Maharam ibn Habib, tractate Yoma 83a, s.v. and we read, and many others.

As noted, the reason for this is that the dispute between the Ran and the other medieval authorities did not revolve around this question (at least as the commentators understood it), but around the question whether quantity overrides quality. If so, with respect to every point that is not connected to their fundamental dispute, one may understand that there is agreement. Therefore one can derive practical halakhic conclusions from it, even from the words of one whose opinion was not accepted as the law.[19]

As is well known, the sages do not generally occupy themselves with the rules of their own thought. Yet we do find direct engagement with the problem of simplicity and truth, attributed to R. Chaim of Brisk. R. Chaim's formulation is already closer to that of a general principle of thought. I found a written source for these remarks in the book Torat Chaim, which records teachings from the Brisk school, and there it writes as follows:[20]

Maran Rabbi Chaim, of blessed memory, used to say that when there is one question in a Talmudic passage, one gives one answer; when there are two questions, one gives two answers; but when there are already three questions, one no longer gives three answers. Rather, one must give a single answer, namely: "we are not learning the passage correctly." For the abundance of difficulties proves that there is something basic in the passage that we do not understand, and then further answers will not help; rather, one must learn the passage again from the beginning. (Our master R. Chaim, of blessed memory, used to say that when there is one question in a passage, one gives one answer; when there are two questions, one gives two answers; but when there are already three questions, one no longer gives three answers. Rather, one must say one thing: 'we are not learning the passage correctly,' for the multiplication of difficulties proves that something basic is not understood by us in the passage, and then no more answers will help; rather, one must study the passage from the beginning.)

And he brought proof for this from what is stated in the Talmud, Hagigah 3b: The Rabbis taught: Who is considered an imbecile? One who goes out alone at night, spends the night in a cemetery, and tears his clothing… ('The Sages taught: Who is an imbecile? One who goes out alone at night, sleeps in a cemetery, and tears his garment…')

Actually, he does them in a deranged manner. As for spending the night in a cemetery, one might say he does so in order that a spirit of impurity rest upon him [that is, if he only sleeps in a cemetery, perhaps he does so for that reason, and therefore he is not necessarily an imbecile]. And as for one who goes out alone at night, one might say anxiety has seized him [that is, he is worried, or he is hot, and therefore he is not necessarily an imbecile]. And as for one who tears his clothing, one might say he is deep in thought [that is, absorbed in his thoughts and distracted, he tore his clothing, and therefore he is not necessarily an imbecile]. Once he did all of them, he is like one whose ox gored an ox, a donkey, and a camel, and thereby became established as dangerous with respect to all.

Thus we see from the Talmud that the first and second time one gives an answer explaining why he did so, but by the third time there is no longer any answer. For in the case of answering with three answers, it is preferable to answer with one answer—namely, that this person is necessarily different in essence from others: he is an imbecile.[21]

For a similar argument, see also Rabbi Abraham Isaac Kook's book Orot, in the section Zer'onim, in the subsection called For the War of Opinions and Beliefs. There Rabbi Kook points to the generality of a phenomenon as an indication of its truth.[22]

All these are examples of what in philosophy is called the principle of 'Ockham's razor.' This principle states that we do not posit more assumptions than we are forced to posit. In other words: we choose the simple explanation, the one containing the smallest number of assumptions or components.

In the next chapter we will see an application of the above discussion in philosophy of science to understanding the methods of Jewish law. There we will discuss whether simplicity is a criterion for understanding Jewish law itself, or whether it is instead a fictive tool of the learner, intended to organize within himself the totality of halakhic phenomena.

F. An Application to the Problem of Dichotomies

In light of the foregoing discussion, it seems that the criterion of simplicity and elegance of explanations is one of the parameters that determines the choice among different explanations.

It is important to note that Ockham's razor, in its common formulation, states that one ought not posit assumptions beyond what is necessary. This explains the rule that emerges from what is common to the two cases?. If we have no strong constraint, we will not tend to assume that two different explanations can generate one and the same law. Every law derives from only one factor.

But as we saw above, the problem of dichotomies is different, and perhaps the reverse. Here the question is not the adoption of two explanations, but specifically the adoption of a single explanation that has two components. The number of assumptions proposed in the non-dichotomous possibility is no greater than the number of assumptions in the basic explanations. Each explanation is based on one assumption. The only question is how many components will be included within that one assumption.

To that question the razor principle gives no direct answer, for it deals only with the criterion of the number of assumptions. But the criterion of simplicity and elegance does appear applicable to this question as well. If we understand a one-component assumption to be simpler than a two-component assumption, then the criterion of simplicity will once again lead us to the dichotomous conclusion.

If so, the problem of dichotomies too can be understood in the same two ways we saw in the previous chapter: either the dichotomy is only an efficient and convenient working assumption, or simplicity is a criterion of truth.

If the dichotomy is indeed only an efficient working assumption for us, then we begin the inquiry by examining the simple explanation. If we find such an explanation, we are satisfied with it, even though it is not necessarily true. From our point of view it is preferable and more convenient than the complex alternative.

It seems that this possibility is not relevant in the context of halakhic interpretation. The reason is that in Jewish law every such explanation has practical ramifications. Perhaps one cannot decide between the two possibilities (the simple and the complex), but in such a situation we would have to be stringent in Torah-level cases of doubt and lenient in rabbinic cases of doubt, or use the other rules governing doubtful cases. The fact that one explanation is simpler—if this is not an indication of its truth—cannot justify adopting it as practical Jewish law.[23]

Therefore it seems that within the framework of Jewish law we are forced to conclude that the simpler explanation is also the truer one. As we saw above, the explanation for this is that there is a correspondence between our intellect and Jewish law. The Holy One, blessed be He, created our intellect, and He expects us to interpret Jewish law through it. We have no other tool. Therefore it is reasonable to assume that our conclusions are the correct interpretation (at least from our perspective. This is not the place to discuss the meaning of halakhic truth in general, nor the question of both these and those; see a little on this below in Chapter 8). We pointed out that the simplicity of an explanation means its conformity to the forms of our thought, and therefore a simpler explanation means one that accords with those forms of thought. And if there is correspondence between Torah and human reason, then a simpler explanation is also a more correct one.

Had we not accepted this argument, we could have asked a far broader question: what justifies the assumption that what we infer through our intellect is indeed the will of God? Such a question can be asked not only with respect to the problem of simplicity and truth, but with respect to every form of human thought. Therefore it seems clear that a necessary assumption is that the conclusions that emerge from our halakhic-Torah reasoning correspond to God's will for us. If so, the same is true with regard to the criterion of simplicity in explanations.

G. The Criterion of Simplicity: 'Juxtapositional' and 'Blended' Composition of Concepts

The Criterion of Simplicity

Thus far we have dealt with the question of the connection between simplicity and truth. But, as we saw, philosophy of science asks an additional question: what is the criterion of simplicity? In our context one may ask it as follows: why is a two-component theory (one explanation including two parameters) less simple than a one-component theory (an explanation including one parameter)?

The 'simple' explanation in the second inquiry is that the factor generating liability to pay is only the benefit. Another 'simple' explanation is that only the loss is the factor generating liability. Why is the explanation that benefit and loss together generate monetary liability less simple? Here too there is only one factor, except that its description in language seems more involved. Yet it is difficult to provide a sharp logical basis that would explain this criterion of simplicity. As noted, Ockham's razor treats a structure containing more assumptions as a more complicated structure. But here the number of assumptions in all three explanations is identical (one assumption). If so, why is the explanation that contains a complex assumption not as simple as the other two?

For example, is the explanation that a certain act was carried out by a person who is wise and tall less simple than the explanation that the same act was carried out by a tall person? The fact that the explanation requires more components in order to define it does not necessarily mean that it is less simple. Seemingly, this is a problem of language rather than of essence.[24]

Blended and Juxtapositional Composition

Here a distinction is called for between two types of combination of different components, and for this purpose I will use terminology drawn from the teachings of the Rogatchover (author of the Tzafnat Pa'neach).[25] If the conjunction is truly 'blended'—that is, it creates a third essence compounded from both, so that in principle it would have been justified to attach a single linguistic term to it—then the fact that in order to describe it we use a combination of two components (both linguistic and essential) neither adds nor detracts. That is a merely linguistic phenomenon. But if the conjunction is 'juxtapositional'—that is, the two components merely stand alongside one another and do not fuse into one new essence—then the new explanation is less simple, not only on the linguistic plane but on the plane of essence as well.

Presenting the relationship between the two components as a structure consisting of one dominant component and another that serves as a secondary condition is itself the construction of a structure that binds them into a new, third essence that includes them both. Combining them at equal status is usually interpreted as juxtapositional combination, and therefore it appears to us less simple. But if the combination of two components at equal status can also be interpreted as the creation of one new structure—a result of a blended rather than juxtapositional combination—then a complex explanation containing such a combination will also be possible.[26]

The same applies to the construction of concepts out of atomic concepts (as was done in my article in Misharim 2). If the resulting concept is one third concept, and not a juxtapositional combination of its two predecessors, then it can be created. If not, we must seek the common denominator shared by the two concepts, and that will be the factor generating the law under discussion. This is the difference we saw there between what is common to the two cases? and 'conceptual construction.'

Application to the Two Inquiries Above

We will now try to apply what we have proposed thus far to the two earlier inquiries.

With respect to the inquiry into what generates liability to pay in cases of damage caused by property, we indeed saw that the Pnei Yehoshua himself proposes a model that combines the two components—ownership and negligence—at equal status. But we also saw that both are required in order to create the liability: one in order to create the liability on the damaging animal itself, and the other in order to transfer the liability to its owner. In such a case the combination creates a single track, and therefore it is no less simple than the basic proposals raised in the original inquiry. Had the two components been required together in order to perform the same role, and had we been unable to find a distinct role for each of them, then this would have been a juxtapositional combination, less simple, and we would have had to reject it.[27]

The Pnei Yehoshua's innovation is not methodological. He is not trying to argue that complex explanations may be adopted and simple ones need not be preferred. His claim is that sometimes the complex explanation is no less simple, and therefore it too must be taken into account when the basic alternatives of the inquiry are formulated. In the inquiry into what generates liability in cases of damage caused by property, there are three simple alternatives, not two simple ones and one complex one, as one might have thought (and as may perhaps be the case in other inquiries). Therefore there the Pnei Yehoshua is prepared to adopt the complex option.

With respect to the inquiry into what generates liability for benefit, there the Pnei Yehoshua himself refuses to accept that a combination of benefit and loss can generate what neither of them alone can generate. Apparently, there the Pnei Yehoshua does not view this combination as creating one third concept. That is, in his view this is a juxtapositional rather than a blended combination, and therefore it is not a simple explanation, so he rejects it as implausible.

One should note that he rejects this explanation even though he has no proof for doing so, and in fact he himself sees that such an explanation follows quite strongly from Tosafot. He prefers to offer a novel alternative, and perhaps a less simple one, involving the principle of the measure of Sodom in the laws discussed by the passage.[28] This position runs against the view of the other commentators, who understand the passage as dealing with ordinary legal law, not with a meta-legal principle such as the measure of Sodom. In any event, it is clear from here that there is no principled contradiction in these remarks of the Pnei Yehoshua. Even in the passage about what generates liability in damage caused by property, he is not prepared to adopt a juxtapositional combination of explanatory elements; rather, his claim is that there the combination is blended. Here, where the combination is juxtapositional, he rejects it out of hand.

We mentioned above R. Shimon Shkop, who explains Tosafot as holding that benefit is indeed what generates liability, while loss is a condition for the emergence of the liability. If his intention is that the absence of loss is an exculpatory principle, then we have come very close to the Pnei Yehoshua. It seems more likely that R. Shimon Shkop means that loss too is a positive condition and not a negative one—that is, it is a condition for the emergence of the liability, not that the absence of loss exempts from liability. If so, a combination has indeed been created here that is seemingly juxtapositional. Why should two factors be required in order to create monetary liability?

It seems that R. Shimon Shkop argues that here there is a structure in which two factors are combined at equal status, and neither has a defined role (unlike the Pnei Yehoshua's proposal), and nevertheless a third concept is created here—that is, this is a blended combination. According to R. Shimon Shkop, apparently the conversion of loss into benefit is the criterion that obligates payment. What obligates payment is the process, not either of the two states at its endpoints.[29]

H. General Implications: These and those are both the words of the living God

The Maharal and R. Tzadok on both these and those

The approach of the Maharal and R. Tzadok to the issue of These and those are both the words of the living God holds that all human truths—and halakhic truths in particular—are partial. The all-encompassing divine truth is a composition of all of them into one complex structure.[30] It seems that the source of this idea is the only place in the Talmud where the meaning of the claim These and those are both the words of the living God is directly explained. In Gittin 6b, R. Yonatan and R. Evyatar disagree about the concubine at Gibeah. R. Evyatar says he found a fly in her ('he found a fly in her'), and R. Yonatan says he found a hair in her ('he found a hair in her'). The conclusion of the passage is that before the Holy One, blessed be He, These and those are both the words of the living God: he found a fly and was not particular, he found a hair and was particular.. From here, apparently, the Maharal and R. Tzadok derive their view that the combination of the two opinions is what creates the full truth.[31]

Application to the Analytic Inquiries

The question that arises in light of what we have said until here is whether one can build, out of these analytic inquiries that are presented as dichotomous, a complete system that includes all of their sides. Seemingly, just as with any dispute among opinions in Jewish law, here too one can generalize the two polar opinions and create from them a more true and more complete complex position.

But matters are not so simple. The complex structure formed by combining the views should apparently be a blended combination of them. But if there is indeed a possibility of blended combination between the two sides of the inquiry, then the inquiry itself has no justification. As we saw, in a situation where a blended combination exists between the two sides of the inquiry, we must in fact consider three possibilities: A, B, and their combination. If so, the generalization is not a combination of the two fundamental sides, but simply adoption of the third side.

Let us illustrate this with the inquiry about 'one benefits and the other does not lose.' One side holds that only benefit generates liability to pay. A second side holds that only loss generates liability. If it is indeed possible to combine the two, as R. Shimon Shkop holds (and unlike the Pnei Yehoshua), then their combination simply yields R. Shimon Shkop's view. If so, where does the Pnei Yehoshua fit into this combination? Is there a combination between the Pnei Yehoshua and R. Shimon Shkop?

From the conception of the Maharal and R. Tzadok, we must reach the conclusion that there is indeed a higher synthesis here.

As we saw, there is a side according to which only benefit generates liability to pay. There is a side according to which only loss generates liability to pay. And there is a side that combines them and holds that both together are required in order to obligate payment (R. Shimon Shkop). The generalization that makes peace among all the positions does not, in this context, mean the stance that both together obligate payment. That is merely one of the positions within the inquiry itself. The meaning of the generalization is that there is a side according to which only benefit obligates payment, and a side according to which only loss obligates payment, and a side according to which both obligate—and all of these together are true sides. The structure of the overall truth contains the blended conjunction of all the explanations, not the blended conjunction of all the elements.

The same can be said regarding the inquiry into what generates liability in cases of damage caused by property. There we saw that the inquiry operates on the plane of the elements, not on the plane of the explanations. The Pnei Yehoshua adopts a third explanation that includes a combination of the two elements. But as we saw, this is not a generalization (= a simple logical conjunction) of the two explanations. Such a generalization exists only on the plane of divine truth. It cannot appear as one side of the inquiry.

Jewish Law and Thought

The conclusion is that the generalization of both these and those exists on a plane that is not halakhic. Jewish law asks: what generates liability to pay? To this question there are three possible answers. Practically speaking, we as human beings must rule for only one of them as halakhically correct. There are practical ramifications in Jewish law and in actual practice between the various possibilities, and therefore the generalization is not implementable on the halakhic plane. But on the plane of thought, we too, as human beings, can understand that there is a generalization that includes all these possibilities together, as aspects of one inclusive structure. This structure is a blended combination of all these possibilities (and not of all the elements).

The conclusion is that the realm of thought allows higher integrations than Jewish law does. But it seems that one cannot arrive at them without the halakhic inquiry conducted in simple logic. Only after the dichotomous clarification can one perform a full, higher generalization that will include all the possibilities exhausted in analytic study. The inquiry exhausts the two possibilities, and that is a necessary condition for integrating them in a dialectical movement.

The Possibility of Generalization

The question now arises whether human beings can perform such a generalization at all. At first glance there seems to be a logical contradiction here. If we say that both elements are required in order to create monetary liability, that is a straightforward logical statement. The generalization of the statements 'benefit generates monetary liability' and 'loss generates monetary liability' is not problematic at all. One can easily say that both together are required in order to create the liability. In logical terms, this is simply the compound statement 'benefit generates monetary liability and loss generates monetary liability,' which does not appear problematic (it is only less simple, as explained). But the generalization of the two explanations (and not the two elements)—that is, the logical conjunction of the two polar claims, 'only benefit generates monetary liability' and 'only loss generates monetary liability'—yields the compound statement: 'It is true that only benefit (and not loss) generates monetary liability, and also that only loss (and not benefit) generates monetary liability.' That is an outright contradiction!

'The Unity of Opposites'

At this point the ordinary student tends to pull from his hat the winning card of the 'unity of opposites.' With the Holy One, blessed be He, opposites can dwell together, since He is beyond logic, and so forth.

In my first book (see Gate Twelve)[32] I already pointed out that the approach of the unity of opposites is borrowed from medieval Christian thought (Nicholas of Cusa). Beyond that, I showed there that this provides no real solution to any problem whatsoever, since we are speaking as human beings. Even the claim 'with the Holy One, blessed be He, the above statement is the correct one' cannot be made within human discourse. At least we, if not the Holy One, blessed be He, are subject to the laws of logic, and therefore in our language the above statement says nothing at all. As I cited there (in the name of Rudolf Otto), such a statement is usually an expression of intellectual laziness.

Back to Logic: Two Types of Opposition

Therefore, if we wish to understand, and certainly if we wish to adopt, the position of the Maharal and R. Tzadok, we must examine this combination according to the rules of ordinary human logic.

It turns out that this problem does not exist at all once we notice that contradictory statements do not always contain a logical contradiction. In moral philosophy a distinction is drawn between conflicts of beliefs (contradictory beliefs) and conflicts of duties and desires.[33] When two beliefs are incompatible with one another, then when we adopt one we necessarily reject the second. But this is not so with desires. When I want to eat something, and on the other hand I do not want to eat it (because it is unhealthy), both desires exist in full force, even if in practical behavior I must choose one of them.

The same applies to duties (where the discussion is about moral duties). When there are two conflicting moral duties, choosing one of them does not necessarily negate the obligation to the second. True, here too we must decide—that is what happens in every moral dilemma—but this does not nullify the validity of the second obligation.

Saving life overrides the Sabbath, but this does not mean that there is no value to Sabbath observance. Both values remain in force. Jewish law imposes obligations upon us. When there are two contradictory obligations, we must decide between them in practice—that is, issue a halakhic ruling. But both still remain true on the plane of ideals.

Similarly, halakhic decision between different explanations does not mean denying the truth contained in the rejected opinion. In halakhic language one might say that issuing the ruling turns the rejected opinion into one that is 'overridden,' not one that is 'permitted.'[34]

From another angle, one may say that there is one perspective from which it follows that monetary liability can arise only from benefit, and another perspective from which it follows that monetary liability can arise only from loss. For example, the desire to eat food from the standpoint of its taste leads one to prefer certain foods. By contrast, the desire to eat it from the standpoint of its contribution to health leads (necessarily, alas) to not wanting to eat those very same foods, or to preferring completely different foods. From each perspective there is one truth, which apparently contradicts the other; yet combining these two perspectives does not amount to a logical contradiction.

When one views the conflict in this way, it is entirely possible that human beings too can perform this generalization, and not only the Holy One, blessed be He. There is certainly no need to arrive at the unity of opposites.[35]

This is not the place to elaborate further. Let the wise hear and grow still wiser.

[1] Rabbi Abraham Blidstein's article, Mimidbar Mattanah, Naso, 5758, issue 78, already discussed this matter. It was what prompted me to continue engaging with it.

[2] Rabbi Zevin, in his book Ishim VeShitot, notes that the Rogatchover often speaks of three sides to a law, not two (as has been customary in the yeshivas since R. Chaim). However, the discussion there concerns the term 'two laws,' whose meaning is the conjunction of both sides of the inquiry together, and not deciding and choosing one of them. Very often the move is dialectical: after presenting two (or three) sides that are the subject of inquiry, the conclusion is that all of them are together present in the context under discussion. Below we will see that even when this occurs, the basic intuition is that one must first consider the possibility that only one of them is correct. See more on this below in Chapter 7.

[3] See my article, 'On Liability to Pay for One's Property That Caused Damage,' in Mishpetei Yisrael, Shlomo Grintz (ed.), Institute for Mishpetei Yisrael, Petah Tikva, 5763. In note 2 there I briefly addressed the problem of the dichotomies of the inquiry, which is the subject of the present article.

[4] And see there further in subsection 18, where he discusses this and brought proof from the Talmud, 99b. See Rabbi Chaim Vahev's article in Mishpetei Yisrael mentioned above, which discusses this and rejects his proof.

[5] See, for example, the novellae of R. Shmuel (Rozovsky) on Bava Kamma §1, and others.

[6] The Rosh Yeshiva pointed out to me, however, that the language of the Chazon Ish does not imply this. From his wording there, it appears that the burden of proof rests upon the damager because of the assumption that ordinary damage occurs through the owner's negligence. Therefore, if the owner claims that it was an unavoidable accident, that is not the default state, and therefore the burden of proof rests upon him. And this indeed clearly emerges from his wording there, both in subsection 7 and in subsection 18; see there carefully. For our purposes below, what matters chiefly is the Pnei Yehoshua's view, and the Chazon Ish's view was brought only as an opposing possibility in order to sharpen the understanding that lies at the basis of the Pnei Yehoshua's position. One may relate to such a position as one side that arises according to R. Shmuel Rozovsky, and not as the Chazon Ish's own position.

Let us note that according to this explanation of the Chazon Ish, the difficulty raised by Keren Orah in Yevamot 64b is very easily resolved. He objected to the medieval authorities there, who wrote that with regard to a warned ox, as practical Jewish law we follow three occurrences and not two, since there is a doubt, and in cases of doubt one does not extract from one in possession. Keren Orah objected that according to this, in order to remove it from the presumption of being a warned ox and restore it to its former status of non-warnedness, two times should suffice, since here one is not extracting money from one in possession but leaving it in his possession. Yet we do not find this among the medieval authorities (and see Or Gadol on the Mishnayot there). According to our explanation, this is very well resolved, for the intention of the medieval authorities is not extraction of money from one in possession, but departure from the default explanation. That is, when the ox has the status of a warned ox, the default explanation is that it gores as a warned ox, and the one who wishes to restore it to a state of non-warnedness bears the burden of proof. Since this concerns money, he is called one who extracts from one in possession. The default explanation determines possession.

One should note, however, both regarding the Chazon Ish and regarding this explanation in those medieval authorities, that as practical Jewish law we hold that a presumption does not extract from one in possession, and even a majority—which is stronger than a presumption—does not extract money from one in possession. If so, it is seemingly difficult why estimative presumptions, such as the one according to which damage does not occur without negligence on the part of the owner, or the one establishing that an ox which gored three times is a warned ox, do extract from one in possession.

One may say that in these cases the presumptions define possession rather than extract from known possession. A presumption does not extract from known possession, but when possession itself depends on that very estimate, it may be that the presumption does extract from one in possession. Yet it still requires explanation: in which cases do we say that the estimate determines possession, and in which cases do we follow the ordinary rules of possession?

Therefore it seems preferable to say more than this. In truth, everywhere we find that such presumptions extract from one in possession (or determine possession), this is only because even without that estimate there is no established possessor. Only in situations where there is no established possessor does the presumption determine possession. Of course, we must now ask why in these cases, absent those estimates, there is no established possessor.

Let us return to the above Chazon Ish and see this in his case. According to what we have said, it follows that according to the Chazon Ish, the damager is not in fact the established possessor, for otherwise the estimate would not suffice to extract from him. The reason is that in any event the Chazon Ish apparently does not understand, like the Pnei Yehoshua, that negligence is what generates liability to pay, for if that were so, the damager would be the possessor and the estimate would not suffice to extract from him (or to turn the injured party into the possessor). If so, the reason there is no established possessor in cases of torts is apparently that according to the Chazon Ish there is at least a side according to which the fact that his property caused damage is what generates liability to pay. Only against that background can the estimate determine that the injured party is the possessor, as explained above.

Accordingly, those later authorities mentioned above were not entirely mistaken in their understanding of the Chazon Ish, but this is not the place to elaborate further.

[7] To some extent, even the attribution of the Pnei Yehoshua's view to the side according to which negligence in guarding generates liability to pay can be challenged. Perhaps one can say that negligence is indeed only a condition, yet in the final analysis it too is required in order to obligate him to pay. Therefore, so long as it is not known that he failed to guard properly, the injured party is still the extractor and the burden of proof rests upon him. The later authorities mentioned above, who understood the Pnei Yehoshua as holding that negligence in guarding is what generates liability, apparently assume that the condition of negligence in guarding is an exculpatory condition rather than a condition for liability.

However, such a formulation already brings us close to the position that both parameters are required equally, and it is not clear why negligence should be considered secondary.

It is highly plausible, however, that a negative condition cannot serve as the primary condition for the application of a law, but there may perhaps be room to view a positive condition as a secondary condition for the law's application (at least if it is not the only positive condition). We will return to this below in greater detail. See also below in note 13.

[8] See the explanation of this developed at length in my aforementioned article.

[9] It is clear that even after this explanation, the Pnei Yehoshua's words still contain an enormous innovation, and one that is by no means necessary. Even if one understands that the fact that my property caused damage is what generates liability to pay, it still seems more reasonable to obligate the owner of the dog who set it upon someone else. If so, this is clearly a forced ruling even according to the understanding that what generates liability is the fact that his property caused damage. But it is equally clear that without this understanding—if what generated liability to pay were negligence in guarding—there would have been no room at all to exempt the owner who set the dog on. Therefore it is indeed proved from the Pnei Yehoshua that what generates liability is the very fact that my property caused damage, and not negligence in guarding.

[10] However, see the Rosh Yeshiva's comment cited above.

[11] See my aforementioned article, where I detailed this two-stage process in the emergence of liability and proved its existence.

[12] According to most medieval authorities this is an incorrect reading of the Talmud, but it is undoubtedly an inquiry that the commentators make on each of the sides in the Talmud itself (with respect to each of the two sides of the doubt). Therefore, for our purposes, it is a useful and sufficient abstraction.

[13] In terms of the practical ramifications mentioned above, this would mean that if there were a doubt whether the occupant benefited, the burden of proof would rest upon the owner of the courtyard. But if there were a doubt whether the owner of the courtyard suffered a loss, then there too the burden of proof would rest upon the occupant (since the absence of loss is exculpatory; it is not that loss is what obligates. See above in note 7 on the distinction between a positive and a negative condition with respect to liability for damage caused by one's property).

[14] See my book Shtei Agalot VeKadur Porach – On Judaism and Postmodernism, Beit El, 5762, p. 404, as well as further aspects in the second part, which will, God willing, be published in the near future. See also my articles in Misharim 2 and in Tzohar 15.

[15] See the discussion in my book mentioned in the previous note.

[16] Regarding that, if we were to find a common basis (a common denominator) for loss and benefit, then indeed we could infer that each of them can generate monetary liability. In fact, it is the common basis found in both that generates it.

[17] The very intuition that we choose the simplest among the possibilities is almost not disputed at all in the philosophical world, but even it itself is not simple (!) at all. There are theories that seem complicated, and it is difficult to argue that they are the simplest. It seems to me that in all such cases we are dealing with some intuitive sense of simplicity, though the criteria for it are not always uniform.

[18] I have not found explicit formulations of this rule (at least in this wording) in medieval literature, but it is clear that it already forms the basis of many halakhic discussions in the Talmud and among the medieval authorities. For an explicit formulation, see the responsa Torat Emet §155, s.v. 'All this'; Mikhtam LeDavid, Yoreh De'ah §16, s.v. 'However, again'; and Responsa Maharshdam, Choshen Mishpat §304, s.v. 'And as for the question…'

[19] One should note that one could have explained the dispute between the Ran and the medieval authorities precisely on this very principle itself—whether one incurs a punishment for each measure consumed. But that explanation requires additional assumptions (regarding the interpretation of the Mishnah in Makkot 21a), and therefore it is rejected because it is more cumbersome. See the novellae of Rabbi Yosef ibn Migash on tractate Shevuot 22a (and also the Rashba's novellae there, where he discusses his words), where it appears that he raises the possibility that even when one eats several measures, he transgresses only one prohibition. Even if that is indeed his intent, it is apparently a solitary opinion in Jewish law.

[20] Torat Chaim, edited by Y. Hershkovitz, Land of Israel (the publisher and year of printing are not indicated). The remarks quoted above appear in the booklet Limmud HaTorah, printed at the end of the book, on p. 52.

[21] In fact, we have here the rule of presumption that is accepted in several halakhic contexts. An ox that gores three times is considered warned to gore (this is the example brought at the end of the quotation). And more generally: if something happens three times, then even if we could have offered a good independent explanation for each occurrence, we attribute the three events to one underlying factor.

A well-known discussion of this issue appears at the beginning of the third chapter of tractate Bava Batra, which deals with the three-year presumption (a person who possesses land for three years is considered its owner and need not bring proof), where the question is raised whether this presumption too belongs to the same family (see, for example, Kehillot Yaakov, by R. Yaakov Yisrael Kanievsky, Taharot §47). The discussion is conducted in light of the Talmud's comparison of this presumption to that of the warned ox (mentioned above), regarding which the medieval authorities disagreed (Maharam of Rothenburg and Rabbeinu Peretz) whether it is an evidentiary presumption (about the nature of the ox) or a constitutive one (giving the ox a nature to gore). On this matter, see Tur, Orach Chaim, end of §114.

One may ask what led R. Chaim to bring proof specifically from the above Talmudic passage in Hagigah, rather than from all these more direct sources that deal with presumption, but this is not the place to elaborate.

[22] For further detail on this issue, see Rabbi Yosef Kelner's book, Pluralism, Pantheism, and Generality (The Criterion of Truth, the Criterion of Morality), Midreshet Netzarim, Elul 5761. See also the booklet Alei Asor, Hesder Yeshiva of Yeruham, Elul 5763, p. 36 and onward.

[23] For the conception that analytic study does not reveal inner dimensions of Jewish law but rather creates such dimensions, see Chaim Navon's article in Akdamot 8. By contrast, see my article in Akdamot 9.

[24] See on this matter note 10 in my first book, where I point out that simple concepts can be described in a complex way. The simplicity of a term is not necessarily proportional to the logical simplicity of the concept that the linguistic term describes (in the language of analytic philosophy: the term's reference).

[25] See Mefa'ane'ach Tzefunot, by R. M. M. Kasher, chapter 8, for a discussion and illustration of this distinction.

[26] In his aforementioned article, Rabbi Blidstein arrives at a similar conclusion. He calls this a living (= organic) structure that contains these two parameters. According to his conclusion, these two are its characteristics, and the explanation does not require both characteristics, but rather the existence of the structure itself of which they are characteristics.

[27] There is a situation in which combining two concepts creates a third concept, and it is clear that only it can generate the law under discussion. Here too there may be a combination of two components at equal status. See note 10 in my aforementioned book, and at greater length in the second book.

[28] Seemingly, according to the razor principle this is less simple. But apparently the Pnei Yehoshua understands that here too there is no problem of simplicity, since each component has a distinct role in generating the liability. The benefit generates the monetary-legal debt, and the loss removes a secondary problem of the measure of Sodom. Thus, each result has a one-component cause, except that the result is itself complex, and therefore the causes contain two parameters.

This is a structure very similar to the one he himself proposes with respect to damage caused by property. There too we saw that ownership defines liability, while negligence allows the liability to be transferred to the owner. Thus, not only is there no contradiction in the Pnei Yehoshua's words, but in fact in both places he says something very similar.

[29] The distinction between a process and the endpoint states between which it passes is discussed at length in my article 'Zeno's Arrow and Modern Physics,' in Iyyun 46, Jerusalem, 5758. For an elaboration of the halakhic relevance of this distinction (with respect to the giving of a bill of divorce), see my article in the weekly page of the Department of Basic Studies at Bar-Ilan University, Ki-Tetze, 5755. See also notes 7 and 31 in my aforementioned book.

[30] For R. Tzadok's position, see Resisei Layla, Bnei Brak, 5727, p. 18. For the Maharal's words, see Derekh Chaim, pp. 258–259, and Be'er HaGolah, pp. 20–21.

[31] The Rosh Yeshiva pointed this out in the general lecture on the dispute between Beit Shammai and Beit Hillel regarding How does one dance ('how does one dance before the bride?'). His claim was, contrary to the common view, that there is no contradiction between truth and peace. Peace is the highest and most correct truth. Peace among the opinions creates a structure that is a combination of them all, and thus truth is created. 'Peace' derives from the root of wholeness.

[32] See also my article 'What Is Ḥalut (Legal Effect)?' in Tzohar 2.

[33] See on this topic Moral Dilemmas, Daniel Statman, Magnes, Jerusalem, 5751, pp. 36–38.

[34] The editor pointed out to me those places in Jewish law where we find a relation of 'entirely permitted' between two values. But even there it is clear that one of them is not nullified. For example, regarding impurity in the community, the views disagree whether it is entirely permitted or merely overridden (see Yoma 6b and Sanhedrin 12b and parallels), yet it is still clear that the negative value of service performed in impurity is not nullified. Therefore it seems that there is no contradiction from there to our present discussion.

[35] See my article 'What Is Ḥalut (Legal Effect)?' in Tzohar 2, where I rejected Daniel Weil's claim (in a note to his article in Higayon 1). Weil argued that every determination of 'two laws' involves a logical contradiction. The formulation there is somewhat different, since the problem there is different. In any case, I showed that the concept 'two laws,' which is itself a seemingly contradictory generalization, precisely of the kind we are discussing here, is one that human beings can in fact make without any difficulty (see there for the logical explanation of this).