The Logic of *Kal Va-Chomer*: III. A Note on Maimonides’ Approach (Column 737)

With God’s help

Disclaimer: This post was translated from Hebrew using AI (ChatGPT 5 Thinking), so there may be inaccuracies or nuances lost. If something seems unclear, please refer to the Hebrew original or contact us for clarification.

In the previous two columns I discussed the “relevance assumption” lurking behind *kal va-chomer* (a fortiori) arguments. Here I continue analyzing *kal va-chomer* and wish to examine it in light of Maimonides’ approach to “Divrei Sofrim” (rabbinic law) and in light of my remarks in the two previous columns.

Pilpul of *Kal Va-Chomer* and the Conclusions

We dealt with an argument based on two data points of the following type:

The argument seeks to obligate a lintel in tzitzit or a garment in mezuzah by force of a *kal va-chomer*: if a garment, which is exempt from mezuzah, is obligated in tzitzit—then a lintel, which is obligated in mezuzah, is it not all the more so that it should be obligated in tzitzit?! And conversely: if a lintel, which is exempt from tzitzit, is obligated in mezuzah—then a garment, which is obligated in tzitzit, is it not all the more so that it should be obligated in mezuzah?! When the table contains only two data points, one can always run the *kal va-chomer* in both directions.

In the following column I brought several exceptions where the Talmud does employ a *kal va-chomer* on the basis of two data points. To explain why these cases do not contradict each other, I explained the *kal va-chomer* in a “layers (slides)” model: one layer (the textual) contains the halakhic data appearing in Scripture, and the second (the measured/derivative layer) contains those learned from the *kal va-chomer*. The claim was that even if we fill the missing cells with a 1, in truth we should write (0,1): 0 means it does not appear in Scripture, and 1 means it is true after the *kal va-chomer*. Therefore the 1 cannot be counted within the *kal va-chomer* argument itself. My assumption is that *kal va-chomer* operates only on Scriptural data.

The resulting table is the following (instead of “Torah” and “food” I left here the lintel and the garment in the table—despite the lack of real relevance—just to make it easier to follow):

However, a difficulty arises here: the result of the *kal va-chomer* undermines its assumptions. For example, we assume that mezuzah is more stringent than tzitzit (from the top row), yet on the bottom row we see the reverse: a garment is explicitly obligated in tzitzit by the Torah, whereas in mezuzah it is obligated only by force of the *kal va-chomer* (textually it is exempt).

I answered that each cell in such a table can be filled in three ways and not just two: 0 – exempt; 1 – obligated; X – not relevant (a lintel is irrelevant to tzitzit, or a garment to mezuzah). In a case with only two Scriptural data points, the starting point is that the lacunae express X, i.e., the rule is irrelevant (and not 0, a halakhic exemption from the rule). But this solves the problem only for a table with two data points. In such a *kal va-chomer* it is indeed hard to accept running it in both directions (and usually one chooses, by sevara, which direction is correct). Yet the difficulty remains in an ordinary *kal va-chomer* built on three data points:

Looking at the Scriptural data, we still see that the hierarchy in the top row (mezuzah is more stringent than tzitzit) does not hold in the second row. The same is true, of course, for the columns.

I explained there that in the “layers” picture the problem is resolved, since the assumptions of *kal va-chomer* are based on the Scriptural data alone (the first component in parentheses). My assumption was that if the result (0,1) depicted a degree of stringency between 0 and 1, it would indeed generate an internal contradiction within the *kal va-chomer* argument. But I claimed that these parentheses do not in fact express an intermediate level of stringency; rather, they are two independent components, each of which must be evaluated on its own. From the perspective of the halakhic result (after the *kal va-chomer*), the hierarchy is preserved, since in the end the learned case (*lamed*) truly is more stringent.

Maimonides’ Approach

In his Second Root, Maimonides presents a very novel position regarding laws derived from the hermeneutic rules:

Since this is so, not everything the Sages derived by analogy from the Thirteen Rules is to be said as having been stated to Moses at Sinai; nor should we say regarding everything we find in the Talmud that they base upon one of the Thirteen Rules that it is rabbinic. For at times it is an accepted interpretation. Therefore, what is proper is this: anything you do not find written in the Torah but you find in the Talmud that they derived by one of the Thirteen Rules—if they themselves explained and said that “this is Torah itself” or “this is from the Torah,” then it should be counted, for the recipients said it is from the Torah. But if they did not explain this and did not speak about it—then it is rabbinic, for there is no verse pointing to it.

That is, his view is that a rule derived from one of the hermeneutic rules that does not support/confirm an existing law but creates a novel law has the status of rabbinic (derabbanan). As is well known, Nahmanides (the Ramban) sharply attacks him on this point (see our book Yishlach Shoreshav, in the essay on the Second Root), and he presents many difficulties with this innovative view—a view that almost all Rishonim and Aharonim rejected outright. The issue is so difficult and so novel that among later authorities a position took hold that Maimonides did not mean to say these are rabbinic laws. Clearly, even according to him, they are de’oraita; their source, however, is from the Sages. He is speaking about the question of source, not the question of legal force. Yet in our book we demonstrated that this is not so, and that Maimonides truly intends to claim these are rabbinic laws. Our contention was that the question of force (status) is based on the question of source: when a law has no clear source in the plain sense of the Torah but is derived by derash, its force is like that of a rabbinic law.

This means that, according to Maimonides, the parentheses (0,1) in the tables above—where the first component expresses the question of source (whether there is a textual source) and the second component expresses the question of force (the law in practice)—can in fact be translated into an intermediate degree between 0 and 1, e.g., 0.5. This reflects Maimonides’ view that a law learned by *kal va-chomer* is less stringent than a law explicitly written in the Torah.

But now we obtain the following table:

Before the derivation, the data direct us to perform a *kal va-chomer*, with the row-argument assuming that mezuzah is more stringent than tzitzit. But the result of the *kal va-chomer* is that mezuzah is lighter than tzitzit, for mezuzah on a garment has stringency 0.5 whereas tzitzit on a garment has stringency 1. The same holds for the relationship between lintel and garment (in the column-argument, which, as we saw in previous columns, is not a different argument from the row-argument).

Thus, the difficulty we saw above regarding *kal va-chomer* intensifies according to Maimonides. Unlike all the other hermeneutic rules, *kal va-chomer* rests on the assumption (generalization) that the learned case (lamed) is more stringent than the teaching case (melamed). Yet the melamed is a rule explicitly appearing in Scripture, while the lamed is a rule emerging from the *kal va-chomer* itself—i.e., it arises from derash—and is therefore lighter. Consequently, the conclusion of the *kal va-chomer* contradicts its assumptions. Put differently: the result of the *kal va-chomer* constitutes a rebuttal (*pircha*) to the very inference: what of the learned case, seeing that the force of the rule in it is lighter (since it comes from derash) than in the teaching case (which is explicit in Scripture)?

The Textual Solution—and Its Undoing

One might consider here as well the textual solution: if we assume that *kal va-chomer* is based on Scriptural data and does not address the practical law, then we can relate only to the left component in the parentheses. On this proposal, *kal va-chomer* is a textual, not a substantive, rule and does not reflect true relations of stringency between the data. But now this already begins to look very problematic, because in practice—halakhically—the learned case comes out lighter than the teaching case, which is effectively a rebuttal to the *kal va-chomer*’s assumption.

This is seen even more clearly if we compare *kal va-chomer* to “kol she-ken,” i.e., an a fortiori based on a single datum. For example: “Behold, the children of Israel did not listen to me; how then will Pharaoh listen to me, and I am of uncircumcised lips?” Here the relation of stringency is grounded in reason, not in Scriptural data, and therefore it is clearly about actual stringency. It is implausible that in a measured/derivative *kal va-chomer* based on three data points we would ignore altogether the true relations of stringency between the learned and the teaching cases.

This becomes sharper still in light of what I explained above according to Maimonides, who links the question of source with the question of force. For him, the source of the law determines its force. If it lacks a source in the Torah, its force is lower. Accordingly, it is even harder to sever the textual question from the halakhic question and to claim that *kal va-chomer* is based on relations among data as they appear in Scripture, detached from the question of their force. On the contrary, in his view the reference to the text is only because it serves as an indication of force.

On Force and Stringency

I think we must distinguish between a law’s force and its stringency. The force of laws may be de’oraita or derabbanan. But their stringency does not necessarily mirror their force. In the end I will argue here that, according to Maimonides, the source of a law indeed determines its force, but not its stringency. The force of a law derived by derash is indeed like that of a rabbinic law, but its stringency can be full—like that of a de’oraita. Therefore, the relations of stringency are preserved even after the *kal va-chomer*, and there is no contradiction to its assumptions. In other words,

To understand this better, I will preface a principle I have discussed elsewhere (see at length in my essay on sevarot). In the essay “Teshuvah,” in his Collected Essays, Ra”u notes that there are two planes in performing a mitzvah or transgressing a prohibition: the command and the essence. The assumption is that each mitzvah is meant to achieve some benefit, and each prohibition prevents some harm that the prohibition comes to avert. Accordingly, one who performs a mitzvah accomplishes two positive things: he obeys the command, and he brings about some positive outcome (which is the reason we were commanded). Similarly regarding one who transgresses a prohibition: he rebels against the command and causes harm. Thus, for example, Tosafot ha-Rosh (Kiddushin 31a) explains the principle that “one who is commanded and does [is] greater than one who is not commanded and does” as follows:

“Greater is one who is commanded and does.” Although we say in Hagigah chap. 2, by way of parable, that if a king told his servants to rise early at his door—to whom is there more reason to be grateful? To the women, whose way is not to rise early as men do…—that is not like our case. There both are commanded, and one who does what is not his way is better when he does. And the reason here is that one who is commanded and does is preferable because he constantly worries and is pained, lest he transgress, and he must subdue his inclination more than one who is not commanded, who may desist if he wishes. Moreover, the Holy One, blessed be He, needs nothing from the commandments; rather, He says and His will is done. Therefore, the one who is commanded and does fulfills the will of his Creator; but one who is not commanded and does—regarding him it is not apt to say that he does his Creator’s will, for He did not command him anything. Nevertheless, there is reward.”

In his second explanation he writes that one who is commanded and acts simply achieves both benefits, and therefore is greater (the reward reflects that he performed a positive act).

The “essence” says that a given mitzvah brings some benefit (and a prohibition prevents harm). This is a good reason to fulfill it, but it does not create an obligation. Our obligation to fulfill the commandments stems from the command. From this we may conclude that the stringency of a mitzvah is not necessarily tied to the command, but to the essence. The greater the harm a prohibition causes, the more stringent it is. And the greater the benefit, the more important (stringent) the mitzvah. By contrast, the force of the mitzvah is determined by the command: if there is a command in the Torah, it is de’oraita; if not—then it is rabbinic.

In the essay cited I argued that there is no direct linkage between command and essence. For example, among de’oraita commandments there are lighter and more severe ones, even though all are commanded in the same way. Their legal force and status are identical, since all are commanded and are de’oraita; but their stringency certainly differs (in accordance with the benefit or harm involved). Thus, “You shall not murder” is the most severe prohibition, yet its force is like that of any de’oraita prohibition (for it is written in the Torah as an ordinary prohibition). One may discuss whether the punishment reflects a prohibition’s stringency or not (see, for example, here and in this work), but command is only a necessary, not sufficient, condition for liability to punishment. Moreover, there may be commandments severe in essence for which, for various reasons, there is no command (see the end of the essay on sevarot).

We saw above that, according to Maimonides, laws derived from derash are “Divrei Sofrim.” I have explained in several places that in such laws there is essence but no command (for according to Maimonides, derash does not uncover something latent in the Torah itself but expands what is written there; see my essay on the Second Root). Thus, for example, fear/awe of Torah scholars is a de’oraita obligation derived from an inclusion: “You shall fear the Lord your God”—to include Torah scholars (Pesachim 22). This means there is no command regarding awe of Torah scholars, but in essence there is such an obligation and its stringency can be like that of an ordinary de’oraita (which does carry a command). There is a reason we were instructed to revere Torah scholars; i.e., there is benefit in such awe. Therefore there is essence here. This is in contrast to rabbinic laws, which are laws without essence (there is an obligation to fulfill them, but from the Torah’s perspective there is no harm or benefit in their performance or transgression—otherwise the Torah itself would have commanded them). See on this in columns 582 – 583.

In my essay on the Second Root I inferred from here that even according to Maimonides, doubt in a law derived by derash is treated stringently, despite its being rabbinic. I explained that the halakhot of doubt are set according to stringency, not command; and in a derash-derived law, the stringency is like that of a de’oraita, only its force is lower (for there is no command). One implication: according to Maimonides there is no punishment for violating it (since the threshold condition—a command—is lacking, and there is no punishment without warning).

Thus we learn that a law’s force is not identical with its stringency. Its force is measured on the de’oraita/derabbanan plane according to whether there is a command; but its stringency is a continuous matter with varying degrees that are not necessarily dependent on the command. They depend on the act’s consequences—the harm or benefit it causes.

A similar idea I proposed in column 679, which dealt with the relationship between a law’s halakhic force and the strength of the obligation to fulfill it. I showed there that the strength of the obligation to fulfill may differ from its halakhic force. In the terminology presented here, it is more accurate to say that a law’s stringency (its essence) does not depend on whether there is a command. The strength of the obligation to fulfill is determined by the command (for only a command creates obligation), but the stringency depends on the essence—i.e., on the act’s consequences. Therefore, with all de’oraita laws there is the same strength of duty to obey, since all are commanded in the Torah and there is a duty to obey the Torah’s commands. But the stringency varies according to the essence. Moreover, it is possible that the duty to fulfill rabbinic laws is identical to the duty to fulfill de’oraita laws (certainly according to Maimonides, who grounds the duty to obey rabbinic laws in the verse “You shall not deviate”; see columns 652 and 443).

Back to *Kal Va-Chomer* According to Maimonides

We can now understand that even according to Maimonides, it is not the case that the result of the *kal va-chomer* undermines or contradicts its assumptions. We saw that a *kal va-chomer* table according to Maimonides looks like this:

The difficulty was that, according to Maimonides, the rule in the lower-right cell is lighter than 1 (we marked it 0.5), creating a reversed hierarchy in the bottom row compared to the top row, and in the left column compared to the right.

We can now understand that this is not the case. Maimonides does not say that the stringency of a law learned by *kal va-chomer* is lighter than that of a law explicit in the Torah. What he says is only that the strength of the obligation to fulfill it is lower (hence he treats it as rabbinic).

This can take us to two different formulations:

• Following the distinction I drew between the textual plane (whether the law is written in the Torah) and the halakhic plane: one could say that *kal va-chomer* is made on the basis of a law’s force (command and Scriptural appearance) and not on the basis of its stringency. Therefore, even if the product of the *kal va-chomer* is a law less stringent, this does not contradict the *kal va-chomer*’s assumptions. It is less stringent, but its force is identical.

But this formulation is implausible, since, as we saw, according to Maimonides the opposite is true: the stringency of laws derived by derash can be like that of ordinary de’oraita laws, while it is their force that is lower.

It therefore seems better to explain the reverse:

• Indeed, a law learned by *kal va-chomer* has lower force than ordinary de’oraita laws. But its stringency is not necessarily less than theirs, for the *kal va-chomer* assumes that it is actually more stringent (its consequences are no less grave than those of the de’oraita law that teaches about it). We saw that, for Maimonides, derash expands what is written in the Torah rather than uncovering what is within it; therefore, according to Maimonides, the learned law of the *kal va-chomer* does not carry an actual command, whereas the teaching law, of course, does. Nevertheless, the stringency of the learned law is naturally greater than that of the teaching law—that is the very basis of the *kal va-chomer*. Hence the result of the *kal va-chomer* does not contradict the generalization on which it rests, for even in the halakhic plane the law learned by *kal va-chomer* is more stringent than the teaching law.

Of course, according to this proposal Maimonides would hold that the *kal va-chomer* inference is based on the stringency of the given laws, not on their force. In other words, the *kal va-chomer* is made in the halakhic (substantive) plane and not in the textual plane (as I suggested in previous columns to explain certain kinds of two-datum *kal va-chomer*).

One implication of this proposal is that there will be situations in which we have a *kal va-chomer* data table (i.e., a table with three data points as we saw), where the upper-right cell is 0 because there is no Scriptural command—and nonetheless we will not perform a *kal va-chomer*. This could occur if, for some reason, we understand that that cell concerns a law that is very stringent (and the reason the Torah did not command it is circumstantial, as I noted above). In such a case we cannot perform a *kal va-chomer*, since on my proposal the *kal va-chomer* is based on relations of stringency and not on relations of force. Note that under the previous proposal, which assumes that the *kal va-chomer* inference is based on force and not on stringency (i.e., that it operates in the textual plane), we would perform a *kal va-chomer* even in such a case.