More on the Hierarchy between the Commandments and Life (Column 421)

In the previous column I concluded the discussion with a nuanced conclusion about the relationship between the commandments and life:

Commandments without life are not commandments, but life without commandments is not life either. It seems that the hierarchy between life and the commandments is not unequivocal.

There I noted that this connects to the tension between two rationales cited in the Yoma sugya regarding pikuach nefesh (saving life) overriding Shabbat. In this column I would like to touch further on the relationship between these two rationales and to examine that conclusion from a different angle.

The Yoma Sugya

The Talmud in Yoma 85a–b discusses the source for the rule that saving a life overrides Shabbat:

It once happened that Rabbi Yishmael, Rabbi Akiva, and Rabbi Elazar ben Azariah were walking along the road, with Levi the arranger, and Rabbi Yishmael the son of Rabbi Elazar ben Azariah walking behind them. This question was posed before them: From where do we know that saving a life overrides Shabbat?

Several tannaitic sources for this law are then brought, and at the end appears the source of Rabbi Shimon ben Menasya:

[…] Rabbi Shimon ben Menasya says: “And the children of Israel shall keep the Shabbat.” The Torah said: Desecrate one Shabbat on his account so that he will keep many Shabbats.

At the end of the sugya the amora Shmuel arrives and proposes his own source:

Rav Yehuda said in the name of Shmuel: Had I been there, I would have said my source, which is superior to theirs: “and live by them” (Lev. 18:5)—and not die by them. Rava said: All of them have refutations, except for Shmuel, whose source has no refutation…

Shmuel’s rationale receives Rava’s praise, for it alone is unrefuted. Seemingly, this is the rationale that remains as the conclusion.

The sugya ends with further praise for Shmuel’s rationale (in contrast to the others, which have refutations):

Ravina—and some say Rav Naḥman bar Yitzḥak—said: One sharp pepper is better than a basketful of gourds.

Shmuel’s words are like a pungent pepper seed, preferable to a heap of bland gourd seeds.

Is the Halakha Indeed like Shmuel?

At first glance there is no halakhic dispute here, only a question of the source of the law (“the import of the derivations”). Yet the poskim did find practical ramifications (e.g., whether to save a gentile’s life, someone who does not keep Shabbat, a person with only moments to live, and the like). The Talmud in Shabbat 151b explicitly brings the rationale of Rabbi Shimon ben Menasya as the halakha—indeed, it brings only that:

It was taught: Rabban Shimon ben Gamliel says: For a one-day-old infant who is alive we desecrate Shabbat; for King David of Israel who is dead, we do not desecrate Shabbat. For a one-day-old infant who is alive we desecrate Shabbat, for the Torah said: Desecrate one Shabbat on his account so that he will keep many Shabbats. For King David of Israel who is dead, we do not desecrate, for once a person dies he is exempt from the commandments; and this is what Rabbi Yoḥanan said: “Among the dead [he is] free”—once a person dies, he becomes free of the commandments.

Desecrating Shabbat for a one-day-old infant is permitted by virtue of Rabbi Shimon ben Menasya’s reasoning. This is a baraita—a tannaitic source—and there they apparently were not yet aware of Shmuel’s rationale (he being an amora). In any case, the second rationale is brought here, implying that it too (and perhaps only it) is the halakha. Moreover, some commentators explained that it is no accident that only the second rationale is brought here, since the case concerns desecration of Shabbat to save an infant who does not yet keep commandments and thus does not “live by them”; therefore, for him only Rabbi Shimon ben Menasya’s rationale is relevant (because in the future he too will keep many Shabbats).

So too we see in the words of the Meiri as cited in the Bi’ur Halakha, Orach Chayim 329:4, where the Shulchan Aruch rules that we desecrate Shabbat even for a person with only moments to live:

And even though the rationale “Desecrate one Shabbat so that he will keep many Shabbats” does not apply here—because it is not specifically about Shabbat; the same applies to other commandments—so wrote the Meiri in Yoma: “Even if it has been clarified that he cannot live even a single hour, in that hour he will repent in his heart and confess.”

The Meiri explains that we desecrate Shabbat for a person with only moments to live even though the consideration “he will keep many Shabbats” is not present (for he will not live to keep many Shabbats). He explains that this is because it is not specifically about keeping Shabbat, but about any commandment; even if he lives one minute, he can repent or perform some commandment, and therefore we desecrate Shabbat to save him.

However, the Bi’ur Halakha there rejects his words:

In truth it appears that all this is merely a rationale, but as to the ruling it does not depend on the commandments at all. The reason is not that we set aside one commandment for the sake of many commandments; rather, we set aside all the commandments for the life of a Jew—as Shmuel derives it from “and live by them,” and as the Rambam writes (Hilkhot Shabbat ch. 2) that the laws of the Torah are not vengeance in the world but mercy, kindness, and peace in the world. Likewise, for all those Tannaim who derive it from the altar, from circumcision, and from the burglar [who tunnels in], it is compelled that the Merciful One is concerned with human life (aside from that [source] of “Desecrate one Shabbat,” etc.).

If, in the conclusion, the source for the law is Shmuel’s “and live by them,” then there is no need to speak specifically about Shabbat or other commandments (such as confession). We desecrate Shabbat for someone with only moments to live because life has immense value regardless of its length and regardless of the commandments that might be fulfilled within it.

The Bi’ur Halakha also proves his point from the law of a foundling (Ketubot 15):

So too it is necessary according to Shmuel, who says regarding a foundling in a place where most are gentiles that we [still] do [life-saving] desecration—even though by law he would be a full gentile and would not fulfill any commandment; nevertheless, since in matters of saving life we do not follow the majority, we worry that perhaps he is from Israel. Accordingly, it is clear that even for a crushed minor we also desecrate, even though he will not keep Shabbats, will not confess, and will not reach adulthood; nevertheless, we desecrate. And likewise for a deaf-mute and one of unsound mind—although they are not obligated in the commandments, nevertheless we desecrate for them, for the fact that they do not fulfill commandments is due to their circumstances.

The Gemara there says that if the majority are non-Jews, the presumption is that he is a gentile; yet we still desecrate Shabbat to save him (because of the possibility that he is Jewish). If he is presumed a gentile, then in practice he will not perform any commandment (out of doubt), and yet we desecrate Shabbat on the factual concern that he might be a Jew. We see that the permission to desecrate Shabbat is not conditioned on the performance of commandments but on the value of life—evidence against the Meiri.[1]

But of course the Meiri himself is not content to leave it at that, and contrary to the Bi’ur Halakha’s remark he does find it necessary to explain the law of someone with only moments to live even according to Rabbi Shimon ben Menasya. It seems that according to him Rabbi Shimon ben Menasya’s rationale also remains in force alongside Shmuel’s, and therefore all the laws must be explained according to his view as well.

This indeed also emerges from the progression of the Yoma sugya there. When one looks at the various refutations, one sees that the refutation that ultimately strikes all the tannaitic sources is the following:

All of them we have found [to permit desecration in a case of] certainty; for doubt—from where do we know? And Shmuel’s [source] surely has no refutation.

That is, all the tannaitic sources can teach us that certain life-threatening danger overrides Shabbat, but not doubtful danger. Only Shmuel’s rationale provides a source that even doubtful danger to life overrides Shabbat. We see that the other rationales remain standing; they just serve as sources for the rule that certain danger to life overrides Shabbat, not doubtful danger. But regarding certainty, they constitute a valid source alongside Shmuel’s rationale. Moreover, if that rationale had ramifications with respect to someone with only moments to live—namely, that there would be no permission to desecrate Shabbat to save him—the Gemara should have attacked the tannaitic rationales on that ground as well. Therefore the Meiri concluded that Rabbi Shimon ben Menasya’s rationale also remains the halakha (as we also saw in Shabbat 151), and the law must be explained according to his view as well.

A Further Look at the Two Rationales

On further reflection, however, it seems hard to hold both rationales together, because at first glance there is a direct contradiction between them. Rabbi Shimon ben Menasya’s rationale is based on the assumption that the commandments are the goal, and the value of life derives from the fact that life enables the performance of commandments. Life is a means to the performance of commandments; therefore the entire permission to desecrate Shabbat in order to save a life does not stem from the value of life but from the fact that life enables observance of the commandments. By contrast, Shmuel’s rationale from “and live by them” seems to assume the opposite: there is value to life in and of itself, as the Bi’ur Halakha proves from Ketubot. One could even say that according to Shmuel the commandments are a means to life (to proper life), and not that life is a means to observing commandments, as Rabbi Shimon ben Menasya holds.

If so, there appears to be a direct contradiction between the two rationales, since they are based on opposing value conceptions about the relationship between life and the commandments and about the very value of life. How, then, can the Meiri hold both conceptions at once? As we saw, both the Yoma and Shabbat sugyot indicate that both rationales remain in the conclusion. This would seem to be precisely the basis of the Bi’ur Halakha’s view, and thus he rejects the Meiri’s words and leaves only Shmuel’s rationale.

Is There Really a Contradiction?

I think this contradiction is only apparent. To see this, let us examine the situation. Seemingly we are to compare two alternatives:

• Desecrate Shabbat and save a life.
• Do not desecrate Shabbat and do not save a life.

Let the value of a human life be X, and the value of keeping Shabbat be Y. If I choose option A I gain the (axiological) benefit X; choosing option B yields the benefit Y.[2] To determine which is preferable I must decide which value is greater, X or Y. In other words, it would seem that I must decide which is higher in the hierarchy—life or Shabbat/the commandments—and construct a scale of values. As we saw, this is ostensibly the substance of the dispute between Shmuel and Rabbi Shimon ben Menasya. I will now try to show that even if I cannot determine a hierarchy between the two values (whether X is greater than Y or vice versa), Rabbi Shimon ben Menasya’s reasoning can still be invoked, and the conflict can be decided according to his consideration.

In fact, I claim more than that: Rabbi Shimon ben Menasya seeks precisely to bypass the need to decide which value stands higher on the halakhic scale of values. His contention is that even without ranking X against Y, the comparison yields a clear result in favor of desecrating Shabbat. His claim is based on the fact that if we save the person’s life, he will be able to keep many additional commandments that will offset the Shabbat that was desecrated to save him. We can present this arithmetically as follows. Suppose saving his life will lead him to keep many additional Shabbats (in fact, we saw that on his view it is about many commandments, not specifically Shabbat, but for simplicity I will speak of Shabbats), say their number is n. Then choosing option A gives us, in addition to the value X, also Y(n−1), for in addition to life we gained many Shabbats (and lost one that was desecrated).[3] In contrast, if we do not desecrate Shabbat, we have gained one Shabbat that we kept and that is all; that is, the total benefit is Y.

To decide the conflict we must compare the benefits of these two options. But of course, regardless of the relationship between the values of X and Y—indeed, regardless of the value of X (so long as it is positive)—we always have (as long as n>2):[4]

X + (n−1)Y > Y

If so, the benefit of desecrating Shabbat outweighs, irrespective of our value scale, because—as Rabbi Shimon ben Menasya observed—in saving we gain not only life but also many Shabbats.

In other words, the conclusion that Rabbi Shimon ben Menasya assumes a hierarchy between Shabbat and life (that life is a means to the keeping of Shabbat) is incorrect. He may accept or reject Shmuel’s view that the commandments are a means to life and are of lesser importance than life; either way he can still say what he said. The inequality above does not depend on Y<X (indeed, it does not depend on X at all). Thus the Meiri is correct that both rationales/sources can be held together; there is no contradiction between them. On the contrary, from the fact that the halakha is like Shmuel—that is, we adopt his conception that X>Y—we can infer that Rabbi Shimon ben Menasya agrees to that as well; but unlike Shmuel he contends that there is no need to rely on it, because even without it the decision is to desecrate Shabbat and save a life.

A Note on Doubtful Danger to Life

What about a doubtful danger to life? This is indeed unclear. According to Shmuel it seems the Gemara assumes that X/2 > Y. But according to Rabbi Shimon ben Menasya, it would seem that even here we should desecrate Shabbat, for even if we erase X entirely, the comparison between the two sides of the inequality still favors the left-hand side (at least when n>2). It seems there is no refutation to Rabbi Shimon ben Menasya.

Perhaps here the difference needs to enter between desecrating Shabbat and not keeping it; that is, it is not (n−1)Y but rather nY − Z, where Z is the “cost” of desecration. Still, it remains unclear why in a case of certain danger to life this inequality would decide according to him. He needs to assume something about the relation between X and Z. This brings us to the next note.

Incommensurability of Values

In analytic moral philosophy there is discussion of the incommensurability (the absence of a common measure) of values. The claim is that we have no way to compare values and determine which is more important. Consider an example Sartre gives. During World War II, one of his students in occupied Paris came to consult him about a dilemma. His elderly mother lived in Paris alone and needed help. He himself wanted to flee Paris and join the Free French to fight the Nazis. What should he do? Which outweighs which—the value of aiding one’s elderly mother, or the value of fighting evil? It seems there is no practical way to establish such a hierarchy, that is, a scale of values. Beyond the practical difficulty, the claim is that there is a principled difficulty.

Leibowitz wrote and said in several places (see, for example, the last essay in his collected writings, Faith, History, and Values, where he discusses disconnecting a patient from machines) that a value, in its essence, cannot be reduced to or grounded in another principle (reduction or rationalization). That is essentially the definition of a value. Every ethical directive is based on certain values, but the values themselves are not based on anything outside them. The first step in any ethical justification is the adoption of a value; therefore the value itself cannot be justified (otherwise we fall into an infinite regress). Leibowitz describes this as an arbitrary act, but in my view he means an act that cannot be grounded in a logical argument (which is not the same thing. See, for example, here). But if there is no way to justify or rationalize a value, then there is also no way to measure it against another value.

We can put it this way: to establish a hierarchy between two values we must ground them on a common basis, measure each of them in those shared units, and then see which one scores higher. For example, if I hesitate whether to drive fast in an ambulance—even though it endangers lives—because speed might save the patient’s life, this is a consideration entirely within the framework of saving life and the value of life; therefore, in principle, one can weigh and determine where the value of life is better served—by driving fast or slowly; the question is what will save more life. But desecrating Shabbat versus saving life, or fighting evil versus helping an elderly mother—these are conflicts between different values that have no common measure; therefore there is no way to measure them on a shared scale to build a hierarchy and resolve the conflict.

It must be understood that if indeed a conflict between values is incommensurable, then even if we find an explicit verse telling us that saving a life overrides Shabbat—or the reverse—that will not help. The problem of incommensurability is not that we are not smart enough to know the answer, but that there is no answer. If I ask you what there is more of in the world: water in the Indian Ocean or human kindness—this is not a question that is difficult to answer; it is a question that has no answer. In such a case, even an explicit verse cannot teach us that there is more water or more kindness. A verse can help when there is an answer and we lack the tools to reach it; but a verse cannot answer a question that has no answer.

Back to Rabbi Shimon ben Menasya

I will not enter here into solving this problem,[5] but will only say that perhaps Rabbi Shimon ben Menasya seeks to bypass it. He rejects all the sources his colleagues bring, because a source cannot answer a question that has no answer. Instead of a source he proposes an answer based on reasoning—but a reasoning that bypasses the need to build a scale of values (for reasoning too cannot build such a scale when there is no relation between its rungs). As we saw, his answer resolves the conflict without needing to rank X against Y, even if there is in principle no way to do so.[6]

However, as we saw, if we try to explain the refutation to him from doubtful danger to life, this forces us to assume that there is a common measure—that is, room for comparison—between X and Z. That does not accord with the assumption of incommensurability, and this requires further thought.

Back to the Previous Column

The conclusion we reached here indeed dovetails with the conclusion of the previous column. The hierarchy between life and the commandments is not unequivocal—neither according to Shmuel nor according to Rabbi Shimon ben Menasya. We saw there that without the possibility of keeping commandments there is no point to life (and therefore the commandment of procreation lapses), and at the same time Shmuel says that without life there is no point to the commandments.

[1] The Chatam Sofer in his novellae to Ketubot there also makes this point against Rabbi Shimon ben Menasya. However, the Meiri in the Yoma sugya cited above writes that the reason we do not save a gentile’s life on Shabbat is that they are not bound by “the norms of the nations” (they do not keep the seven Noachide commandments to which they are obligated, which are essentially moral and human norms). It is clear from his words that regarding people who belong to nations that are indeed bound by norms (such as the Christians in the Meiri’s time—about whom he writes that their status is different; see my essay here)—there is an obligation to save their lives even at the cost of Shabbat desecration by Torah law, just like for a Jew. If so, the Bi’ur Halakha’s argument against him collapses of itself, for according to the Meiri, even with respect to a gentile who is obligated to keep his own commandments, one can apply Rabbi Shimon ben Menasya’s reasoning—that is, to desecrate Shabbat for him so that he will keep (other) commandments, from among his seven, many times.

[2] In principle there is an asymmetry between the two sides of the equation: keeping Shabbat contains a prohibition and a positive commandment; therefore there is value to keeping Shabbat and a negative value to desecrating it. By contrast, saving life has a positive value, but not saving it is not a negative (for I did not murder him; there is no transgression here). Of course there is a loss of the life itself, but it seems hard to say that this has a negative value and not merely the absence of a positive value. See on this in columns 415–416.

[3] Rabbi Shimon ben Menasya ignores the fact that desecrating Shabbat is not merely the loss of keeping one Shabbat; it also has a negative value. Perhaps because of the high value of n he thinks this is insignificant. Even if we include the loss involved in desecrating Shabbat, it would not change the relationship between the two sides of the equation.

[4] Recall that n is the number of commandments, not the number of Shabbats.

[5] See my video and audio lectures on Torah and morality (there are several series on the site), and also the beginning of the third book in my trilogy.

[6] Indeed, from the very fact that X and also Y both appear in the same inequality, it would seem they share units; otherwise there would be no way to compare them. But we can read the plus sign in the inequality as a symbol expressing that beyond the advantage on the left-hand side in terms of the number of commandments/Shabbats (which suffices to decide), one must also add the value of life (measured in entirely different units). Even without assuming a common measure, one can see that this inequality is decided in favor of desecrating Shabbat. The problem arises in the case of doubtful danger to life, for there we see that the value of X does play a role in the decision; that is, there is no assumption of incommensurability. I noted this just above.