Maimonides’ Rules for Counting the Commandments*

Maimonides’ Works – A Jubilee Volume for Rabbi Rabinovitch – 2012

Rabbi Michael Abraham

Introduction

In this article we shall discuss Maimonides’ rules for counting the commandments (‘the roots,’ in the conventional translation). This is something of a neglected duty. For some reason, very few have dealt with this subject, despite its principled importance. Even revolutionary and puzzling determinations, such as Maimonides’ ruling in the second root (that laws derived from exegetical interpretations have the status of ‘the words of the scribes’), arouse only relatively minor discussion among his commentators, and among scholars as well.

The reason for this gap in modern research literature is not clear to me. By contrast, in traditional literature the reason is easier to understand. Traditional literature generally does not tend to deal with its own methods and assumptions. It is not reflective. Hence it is unsurprising to discover that the main discussion in rabbinic literature concerning the roots deals with specific halakhic disputes that arise between Maimonides and Nachmanides in the course of the discussion. Direct discussion of the principles of the roots as such scarcely exists.[1]

Against this background, it is important to note that Maimonides himself was exceptional among Jewish-law sages in his reflective approach, and few followed in his footsteps. Both Maimonides’ systematic-scientific cast of mind and the purpose of his halakhic work (which will be briefly described below) led him to define and formulate explicitly the methodological rules that guided him in his work, and to lay out his methodological assumptions. These unique and rare characteristics of Maimonides led, on the one hand, to the scant attention paid to the roots, and on the other hand, precisely they give added importance and significance to studying them. These rules are a rare example of a prominent halakhic sage’s reflective treatment of his own work, and it is therefore important to study them in order to be exposed directly and immediately to his assumptions about Jewish law (or at least to his conception of it).[2]

It should be noted that this lack is not unique specifically to the roots. No less puzzling a deficiency exists in the treatment of Sefer HaMitzvot generally, which has not received adequate scholarly and exegetical attention (including the roots).[3] Treatment of Sefer HaMitzvot cannot be undertaken in isolation from the roots, which stand at its foundation; therefore the present article should contribute something toward filling this gap. Owing to the breadth of the subject, we shall deal here mainly with the roots, but the connection to Sefer HaMitzvot is self-evident (for the roots define the structure and aim of Sefer HaMitzvot). Hence, naturally, during the discussion a good number of implications regarding Sefer HaMitzvot, its aim, and its structure will arise.

The honored jubilarian, in his Yad Peshutah (in the section ‘Introduction to the Enumeration of the Commandments,’ ch. 2), points to the influence of the aim of Sefer HaMitzvot upon its formulation and structure, and proposes to resolve in this way various contradictions between Sefer HaMitzvot and the Mishneh Torah. Following him, we too shall show here that in order to understand the roots, the enumeration of the commandments, and Sefer HaMitzvot itself, there is no escaping reference to the context of these compositions, in light of the fundamental aims of Maimonides’ halakhic oeuvre as a whole.[4]

The purpose of this article is to provide a general survey of the information accumulated on the subject of the roots, from the range of perspectives relevant to it. The subject itself is broad, and therefore we cannot enter here into a detailed discussion of the content of the roots themselves and their implications. The interested reader will be given references and pointers for further detail.[5] For the same reason, I shall generally not quote the relevant sources but only refer to them. The present discussion is based in part on a series of articles about Maimonides’ roots published in the framework of ‘Middah Tovah’ during the year 5768.[6] Where necessary, I shall refer the reader to those articles for elaboration and detail on particular points.[7]

 

A. The Context

Introduction

Maimonides, in a responsum (Blau ed., no. 447), writes regarding the roots:

As for what you said, my friend, that there are many matters in the enumeration of the commandments about which you were uncertain—certainly so. I have one book, which I composed before this work, which is the Mishneh Torah, and I called that book Sefer HaMitzvot. In it I explained the enumeration of all the commandments, commandment by commandment, and I brought proofs for every commandment about which there is doubt, from the Sifra, the Sifrei, the Tosafot, and from every relevant place in the Gemara. At its beginning I made fourteen chapters containing broad principles and many fundamentals, like mountains upon which everything depends, and I presented them all in advance in order to explain the methods a person should follow in counting the commandments so that they come out properly, and so that he not count in the way many have counted, without grasping the principles upon which one ought to rely…

He likewise writes in another responsum (ibid., no. 355):

I have a work in Arabic on the subject of the enumeration of the commandments, and it is in the possession of our student R. Saadia the cantor. At its beginning are fourteen chapters containing broad principles regarding the fundamentals of counting the commandments; a person must know them, and afterward the error of everyone who counted the commandments besides me—from the author of Halakhot Gedolot until now—will become clear.

If so, Sefer HaMitzvot is an introduction and preface to Maimonides’ Mishneh Torah, and that is indeed how Maimonides himself relates to it. The roots are the theoretical and methodological basis of Sefer HaMitzvot, and without them one cannot understand either its structure or its content.

The roots in the perspective of Maimonides’ overall halakhic oeuvre[8]

Maimonides’ book, the Mishneh Torah, is a unique creation in the world of Torah literature. It is the only comprehensive and classified codex written in the history of Jewish law (apart from, perhaps, Arukh HaShulchan, written many centuries later, and of course relying on earlier works, including Maimonides’ own). Other halakhic compositions are arranged according to the order of the Talmud and therefore do not have the character of a codex (such as Rif and Rosh, which also deal almost exclusively with practical law), or else they have the character of a codex but their scope is partial (such as the Tur and the Shulchan Arukh).[9]

In order to create this monumental work, Maimonides defined for himself a detailed methodological path. He lays out his aim and method in his introductions to Sefer HaMitzvot and to the Mishneh Torah. As stated, by Maimonides’ own testimony, Sefer HaMitzvot serves as an introduction and preface to the Mishneh Torah. As emerges from those introductions, he first reviewed the literature of the Sages—from the Mishnah, the midrashim, and the Toseftot, through the two Talmuds and the literature of the Geonim—a review whose fruits are primarily his Commentary on the Mishnah, as well as commentaries on some Talmudic tractates.[10] Thereafter he began composing the Mishneh Torah as a summarizing codex of the entire halakhic corpus.

In order to define a skeletal structure for this work, he decided to use the number of 613 commandments as the basic organizing categories of his book.[11] As we shall see below, adherence to this number created a severe constraint, and it is possible that Maimonides’ commitment to it led to some corner-cutting as a result of the difficulties that arose in the course of his work.[12] Once he had in hand a complete list of the commandments, he did not stop there. He further divided the commandments by subject matter, thereby producing from them (through an additional thematic-conceptual classification,[13] which was his own original classification; see Twersky, pp. 193–198) the fourteen books that make up the Mishneh Torah, and within them the sets of laws that compose each of those books.[14] Each such halakhic unit opens with a list of the commandments with which it deals and which it explicates, and in this sense the Mishneh Torah is nothing other than an elaboration of Sefer HaMitzvot. It is important to understand this aim of Maimonides while he was creating his count of the commandments, for it differed from that of the other enumerators of the commandments. We shall return to this point at length below.

In order to create a list of 613 commandments, Maimonides had to define rules of enumeration. The task of extracting 613 commandments from Scripture and the literature of the Sages is by no means simple, and here too Maimonides held, in a certain sense, a pioneering position. The enumeration that dominated the field until his time was that of Halakhot Gedolot (Bahag),[15] and all the other enumerators of the commandments followed in its wake quite sweepingly.[16] In his introduction to the Mishneh Torah, Maimonides mocks them and claims that most of them were not Torah scholars in the realm of Jewish law but liturgical poets (For their authors were poets, not rabbis).[17] After the writing of Sefer HaMitzvot, it may be said that Maimonides’ enumeration gained pride of place in this area as well, and indeed largely replaced that of Bahag.

One of the fundamental changes that Maimonides introduced into the enumeration of the commandments, and which was adopted by almost all his successors, was a change in the basic classification of the commandments.[18] In the various enumerations known to us, the commandments are generally divided into positive commandments and prohibitions, as already appears in the literature of the Sages. To be sure, in enumerations prior to Maimonides the commandments were divided into three or four categories, including a division between commandments imposed on the community and on individuals, as well as ‘sections.’ Bahag even divided the prohibitions according to the punishments attached to them.[19]

Now according to the accepted interpretation, the ‘sections’ are commandments imposed upon the community. However, see the introduction of R. Yerucham Fishel Perla (hereinafter: RYFP) to the sections,[20] where he proves that this is not their definition. He shows that each section is a cluster of commandments that gather into one whole (which generally also appears this way in the Torah, as a passage containing several commandments).[21]

From Maimonides onward these categories disappeared, at least in the context of the enumeration of the commandments. Almost all enumerators divide them into two groups (positive commandments and prohibitions). Matters reached the point that, as we saw above, it is no longer entirely clear what those ‘sections’ were that appeared with the early enumerators of the commandments. This is powerful testimony to the revolution that Maimonides wrought in this field.[22]

As stated, in order to carry out this revolution, Maimonides had to define rules for counting and classifying the commandments. These rules were meant to determine which commandment would enter the count and which would not; which commandments would be counted together and which separately, and so on. In order to confront the hegemony of Bahag, and even to ‘rebel’ against it, he had to define these rules and justify them (thereby rejecting the structure proposed by Bahag). For this purpose he wrote ‘the roots,’ which contain fourteen rules for counting commandments. In almost every one of these rules/roots he confronts Bahag, who does not accept a significant portion of them.[23]

This picture presents Maimonides’ halakhic enterprise as a work based on systematic classificatory labor of a scientific character. Maimonides addresses the raw materials before him, gathering, sorting, and classifying, deciding and arranging. He establishes methodological rules from which, and by means of which, he creates his monumental enterprise. It is difficult to see how one could produce such a codex—one that is complete (containing the main body of material before him, organized and without omissions)[24] on the one hand, and on the other hand as free of contradictions as possible—without the methodological infrastructure built by Maimonides. Systematicity was necessary for the creation of the complete structure.[25]

This is one of the reasons that Maimonides is exceptional in his reflectiveness. Jewish law is basically casuistic in character; that is, it is not based on rules but on discussion of precedents involving specific cases.[26] Maimonides almost created halakhic systematicity ex nihilo and produced a comprehensive codex. He therefore required an explicit, formulated methodology, and was even compelled to create much of it himself. It seems that the clearest expression of this approach is the composition of the roots.

Considerations of language[27]

An interesting aspect of Maimonides’ thought process may be found in his linguistic considerations. In his introduction to Sefer HaMitzvot, Maimonides explains that he decided to write the composition in the language of the Mishnah (= Hebrew, though not biblical Hebrew). The reason is that this work was intended for dissemination throughout the Jewish diaspora, and there is no single language that Jews understand throughout the world except the language of the Mishnah. Thus he writes:

I also saw that I should not compose it in the language of the prophetic books, because that language is too concise in our hands today to convey fully the legal discussions. Nor should I compose it in the language of the Talmud, because only a few among our nation today would understand it, and many of its terms are unfamiliar and difficult even for those proficient in the Talmud. Rather, I will compose it in the language of the Mishnah, so that this will be easier for most people.

The language of Scripture is not rich enough in legal-halakhic terms, and therefore Maimonides does not see it as possible to compose the work in that language. The natural remaining possibility is the language of the Mishnah, whose whole essence is the language of Jewish law.

Precisely against this background, it is very interesting to note that the roots and the whole of Sefer HaMitzvot (as well as the Commentary on the Mishnah, written at a younger age, though not directly related to our discussion here) were written by Maimonides in Arabic.[28]

R. Judah ibn Tibbon remarks on this in the introduction to his translation: I composed this work in the language of the Mishnah so that it would not be joined to my larger work, even though it serves as an introduction to it. Even so, the matter still requires further explanation.

One may perhaps suggest that the roots were not intended for future generations and the whole Jewish people. They were written for their own time and circumstance, in order to persuade the public to reject the authority and hegemony of Bahag and accept Maimonides’ enumeration. Once that goal had been achieved, and once the Mishneh Torah attained its dominant status, Maimonides no longer saw value in discussion of the roots and the enumeration of the commandments. These were no more than methodological instruments for purposes of polemic, and so they lost most of their significance after the task had succeeded. The paucity of discussion of the roots and of Sefer HaMitzvot in the generations after Maimonides would seem to testify that this indeed was also the outcome. The revolution succeeded, and now one could throw away the ladder that had helped us climb, for it was no longer needed.[29]

In this connection, it is interesting to note that Maimonides himself[30] regretted having written Sefer HaMitzvot in Arabic, and even planned to translate it into Hebrew—apparently unsuccessfully. Over the generations several different Hebrew translations of Sefer HaMitzvot and the roots were of course produced.[31]

In recent generations, however, as awareness of the importance of systematic study and examination of meta-halakhic assumptions has grown dramatically, the importance of the roots as a singular reflective meta-halakhic composition by one of the greatest sages of Israel seems beyond measure, as I noted above. It is therefore important to return and study them, notwithstanding Maimonides’ original intentions. In any event, when studying the roots and Sefer HaMitzvot generally, one must remember that the text we possess is translated from Arabic, and therefore great caution is required in drawing conclusions from fine readings of these versions.[32]

The time of composition of the roots

Another important point is the accepted view that the roots and Sefer HaMitzvot were composed after the completion of the Commentary on the Mishnah, but apparently several years before the composition of the Mishneh Torah itself.[33] That means there may be contradictions between what emerges from the roots and Sefer HaMitzvot and Maimonides’ final halakhic ruling, since it is well known that Maimonides changed his mind over the years (much ink has already been spilled on this question with respect to his Commentary on the Mishnah and the Mishneh Torah; some of it is even grounded in his own manuscripts, which contain numerous erasures). One must therefore be cautious in drawing conclusions from such contradictions.[34]

On the other hand, in responsa written after the Mishneh Torah one can find direct evidence about aspects that remained unchanged in Maimonides’ thought. For example, there is a responsum by Maimonides to R. Pinhas the Judge of Alexandria (Freimann ed., no. 166) that addresses the principle laid down in the second root (that laws derived from exegetical interpretations are rabbinic laws), and explains through it Maimonides’ words in his great code. This is evidence that the principle stood the test of time and that Maimonides still maintained it when he composed his halakhic work. There is therefore no room for the suggestion that Maimonides changed his mind, when composing the Mishneh Torah, from what he had written in the second root, in order to resolve contradictions between the second root and many passages in the Mishneh Torah. We also saw this in the quotation with which we opened the article from responsum 447 (which was likewise written after the publication of the Mishneh Torah), where Maimonides describes Sefer HaMitzvot and the roots as an important basis for understanding the Mishneh Torah.

At the most basic level, it is certainly legitimate to use the rules in the roots to illuminate the rest of Maimonides’ teaching and his words in his various works, although it is important to do so with appropriate caution.

A bibliography of commentaries on the roots

The roots are printed in most editions as part of the introduction to Sefer HaMitzvot.[35] Around them are usually printed also the glosses of Nachmanides, whose main purpose is to defend Bahag and save him from Maimonides’ criticisms. It should be noted that Nachmanides himself in most cases tends to agree with Maimonides; however, he presents a way to understand Bahag’s method and rescue it from the difficulties.[36] In many cases there are disputes between Maimonides and Nachmanides regarding the application of the various roots. For example, even if one accepts that commandments said only for their own time are not to be counted (third root), it is not always agreed regarding every commandment whether it was said for its own time or for future generations (see Nachmanides’ glosses there).

Most editions also print several additional supercommentaries, such as Megillat Esther, Lev Sameah, Kin’at Soferim, and Marganita Tava, which generally explain and defend Maimonides’ own position. Another important source for the interpretation of the roots is the book Ma’aseh Nissim, which records a correspondence between R. Daniel the Babylonian, who criticized Maimonides’ words in Sefer HaMitzvot and in the roots, and R. Abraham, Maimonides’ son, who replies and explains his father’s intention.[37]

Another source (from the beginning of the twentieth century) is RYFP’s introductions to Saadia Gaon’s Sefer HaMitzvot, where he discusses each of the roots in detail, clarifies it, and also notes what Saadia Gaon and other medieval authorities thought about it.[38] There are of course also several other important modern sources for the interpretation of the roots.[39]

B. The Roots: Sources and Character

Introduction: the enumeration of the commandments and the roots

There are several sources in the literature of the Sages for the number 613 commandments, primarily the statement of R. Simlai in the Babylonian Talmud, Makkot 23b.[40] But this obscure statement conceals more than it reveals. First, it is not clear what those commandments are of which R. Simlai speaks. Which commandments are included in the list, and how are they divided? From the literature of the Sages it is very difficult to extract a closed list of 613 commandments, and they can be divided in many different ways. This is evident from the differences among the various enumerators of the commandments, who disagree among themselves, sometimes very radically.[41]

Moreover, even given a closed list of 613 commandments, it is still extremely difficult to extract from it a closed and unequivocal list of rules that underlie that enumeration. The very same body of material can be explained on the basis of several different principles. It is no accident that there are disputes over many of Maimonides’ roots.[42] A good illustration of this difficulty may be found in the disputes among Saadia Gaon’s commentators regarding his methodological assumptions. In Sefer HaMitzvah VehaMikra, p. 29, two interpretations are presented of Saadia Gaon’s position with respect to Maimonides’ roots. There we find that, in Rabbi Shimon Halperin’s view, Saadia Gaon’s teaching accords with seven of Maimonides’ roots, whereas in RYFP’s view there is agreement with only one root (!). It should be noted that these two commentators on Saadia Gaon are dealing with the same list of commandments, since it is the list set out in Saadia Gaon’s own composition; and nevertheless they agree regarding his position in only five of the fourteen roots. This is in addition to the simpler fact that Saadia Gaon himself disagrees with Maimonides on most of the roots, and on very many individual commandments as well.

This interpretive problem stems from the fact that these methodological principles must be inferred from Saadia Gaon’s enumeration of the commandments, since in his case, unlike Maimonides, there is no introduction setting out explicitly the principles that guided his count. If so, the movement from the list of commandments to the rules is by no means univocal. Consider, then, how difficult it is to extract a list of principles from the entirety of the literature of the Sages, and to build a closed list of commandments on the basis of those principles.

Two kinds of sources for Maimonides’ roots

As stated, despite the difficulties, Maimonides succeeds in laying down fourteen rules that guide him in constructing the enumeration of the commandments. What is the source of these rules? How did he determine them? As we have seen, this is a complicated question, and Maimonides does not provide a direct answer. A reading of the roots themselves shows that there are two principal sources for Maimonides’ roots: 1. His own a priori reasoning. 2. Sources (usually indirect) from the literature of the Sages. We shall now make several remarks about these two sources.

1. It is important to note that Maimonides, as throughout the rest of his writings, possessed a very high degree of self-confidence, and regarded his own a priori reasoning as an almost absolute source; at times he sees anyone who does not accept it as suffering from a defect in thought.[43] Sometimes it even seems as though Maimonides ‘bends’ what emerges from the literature of the Sages in order to adapt it to logical necessity. It should be noted that Nachmanides criticizes him for this more than once in his glosses.

A clear example may be found in the second root, where Maimonides determines that laws derived through exegetical interpretation are laws ‘of the words of the scribes,’ contrary to what is accepted by all the other medieval authorities, and contrary to what emerges from the plain sense of Talmudic passages in dozens and hundreds of places. The reason is that Maimonides’ theory of interpretation assumes that every text has one and only one meaning, and therefore only the plain meaning is the meaning of the text. Maimonides sees exegesis as an expansion of the text, not as its interpretation. Here we have an example of an a priori conception imposing itself on rabbinic material. And indeed, Nachmanides attacks him on this point in extremely sharp language and claims that the whole Talmud cries out against this conclusion. In other words: the man of science must not force his a priori logic upon the facts.[44]

Another example is found in their dispute in the second root. There Maimonides relies on his well-known view that the source of the sages’ authority to enact ordinances and decree decrees lies in the biblical command do not deviate. This claim raises substantial difficulties, for it apparently implies that the halakhic status of these laws is like that of Torah law. That is precisely what Nachmanides argues there against Maimonides’ method. Here too, the consideration that led Maimonides to this conclusion is probably an a priori one: if there is no command in the Torah, it is impossible to see what the source of the sages’ authority is. Nachmanides, by contrast, prefers what better fits the Talmudic material (= the facts) to the more systematic, a priori methodological reasoning.[45]

This fundamental dispute finds other expressions in the various roots, as can be seen throughout the series of articles. In philosophical terms, one could say that Maimonides is a rationalist—that is, he upholds conclusions derived from a priori logical considerations—whereas Nachmanides is an empiricist, who prefers the facts to a priori considerations. This issue is discussed at greater length in the article on the seventh root.

In any case, Maimonides’ own reasonings are a principal source of his roots. There are many expressions in his criticisms of Bahag in which he claims that any person of sense can see that he is right, and sometimes he does not even bother to bring a proof for his words.[46]

2. We mentioned above that another source of the roots is various indications from the words of the Sages. Now with regard to sources from the Sages, it is commonly thought that there is a fundamental dispute between Maimonides and Nachmanides. Nachmanides, in his glosses, consistently in several places, holds that the Sages were not engaged in the enumeration of the commandments at all.[47] For example, in his glosses to root 11 he writes:

Did this Tanna really come now to enumerate the 248 positive commandments and teach us that we should count the commandment of tzitzit as only one commandment? That is not a mistake one could reasonably make.

As noted, some assume that Maimonides himself disagrees with this assumption, and that indeed seems to emerge from several places in the roots, where he learns directly from formulations of the Sages to the enumeration of the commandments.[48] It is precisely against this that Nachmanides objects. However, in my article on the eleventh root I argued that this interpretation is not necessary, and it is entirely possible that Maimonides too agrees with Nachmanides’ claim that the Sages did not engage in the enumeration of the commandments. According to this proposal, one can show that all the proofs Maimonides brings from the language of the Sages concern the content of their words, not the wording as such. The content of their statements leads Maimonides to his conclusions about the enumeration of the commandments discussed there, but nowhere, it seems, is it proven that Maimonides learned this from the wording of the Sages itself.[49] This accords with Urbach’s conclusion,[50] that we find in the literature of the Sages no attempt to enumerate commandments. Because a full discussion would require more detail, I shall not expand on it here.

Recourse to the tradition of the Sages

In several places Maimonides notes that statements of the Sages led him to count the commandments as he did, even where, according to his own rules, the conclusion should have been otherwise. It seems that he departs from his own logic in light of a rabbinic source; and from this it is certainly not correct to see him as a rationalist who entirely ignores facts and sources. For example, in the second root he determines that laws derived through exegetical interpretations are not counted in the enumeration of the commandments unless the Sages determined that they are Torah-level laws, in which case they are to be counted. Likewise, see responsum 447 in Blau’s edition, where he explains in this way his attitude to an inclusive prohibition.

Also in the eleventh root Maimonides cites the words of the Sages as the basis for a distinction between two kinds of parts of commandments that do not invalidate one another. The blue thread and the white thread of fringes are counted as one commandment, whereas the head-tefillin and hand-tefillin are counted as two separate commandments. Nachmanides attacks him there in his glosses, but in any event Maimonides derives this from precision in the wording of the Sages (whereas Nachmanides, consistently with his view, maintains that the Sages were not engaged in the enumeration of the commandments). There are other examples as well.

At first glance, this trait points to an empiricist aspect of Maimonides’ method, in contradiction to his basic position, which is fundamentally rationalist. It is important to emphasize that many commentators understood these exceptions as lacking any systematic explanation. That is, in their view, only because of submission to the words of the Sages did Maimonides deviate from the rules he himself laid down.[51]

The great question is why Maimonides insists on these rules if there are sources in the Sages that contradict them (as Nachmanides asks in several places, such as in his glosses to root 11 and elsewhere). Seemingly, such contradictory sources constitute evidence against the rule, and therefore the rule should be dropped from the list of roots. But that is precisely the point. Maimonides does not remove these rules from his list because he is a rationalist. The rationalist assumes that his a priori beliefs must be true and bends the facts to those assumptions. The empiricist, by contrast, bends himself to the facts and derives his beliefs from them. Therefore Maimonides, as a rationalist, interprets the facts in a way that preserves the fit with a priori logic.

Maimonides’ rationalist stance teaches him that facts cannot contradict common sense (and a priori reason),[52] and therefore when he encounters a source in the Sages that contradicts a rule he has laid down, he interprets it in accordance with that logic. As I show in the articles on the various roots (especially the second and eleventh), in all the cases where Maimonides cites a source from the Sages that ostensibly contradicts his rule, there are several different ways to understand the rule, and the words of the Sages merely decide among those possibilities.[53] The a priori principle is not rejected; rather, it is qualified and interpreted in a way that fits the facts. Thus Maimonides remains a rationalist, and his enumeration is ultimately always based on a substantive principle. The words of the Sages do not override Maimonides’ rules, as one might perhaps sometimes infer from his language; rather, they direct and define them.

For example, in the eleventh root Maimonides determines that parts of a commandment that invalidate one another are counted as one commandment. At first glance, parts that do not invalidate one another should be counted as two separate commandments. Yet Maimonides cites a source from the Sages showing that the blue and white elements of fringes are counted as one commandment even though they do not invalidate one another. In the article on that root I explain that the difference lies in the question of how these parts relate to the overall whole. The fundamental principle of the root remains intact—that parts of a commandment that do not invalidate one another are counted as two commandments—but one must be careful in applying it. Where the parts combine into a single whole, then even if they do not invalidate one another they are counted as one commandment, and that is the situation with fringes (I also adduced there an explicit source from the language of Scripture itself).

On the rules Maimonides presented and those he omitted

At first glance, a work such as that of Maimonides should rest upon very many methodological and substantive rules. Yet Maimonides chooses to present and discuss only fourteen of them.[54] How did he choose them, and why did he omit the others? It seems that we can learn this from his own words. Already at the beginning of the first root (as in other places) Maimonides writes:

Know that this matter (= the first principle) should not have required explanation at all, for since the language of the Talmud is that 613 commandments were said to Moses at Sinai, how could one say of something rabbinic that it is included in the count? But we addressed it because many erred in it and counted Hanukkah lights and the reading of the Megillah among the positive commandments…

From the very opening of Maimonides’ language we see that he does not at all regard it as necessary to present self-evident rules.[55] From his words here it emerges that the reason he presents a rule as obvious as the first root is solely that, despite its simplicity, it is not accepted by all enumerators of the commandments (that is, by Bahag and his followers). We thus learn two things about Maimonides’ approach in the roots:

1. He presents only rules that are not trivial, and does not present rules that are simple and obvious (in his view). Presumably, he presents innovative rules even if they are agreed upon by the other enumerators of the commandments.
2. He also presents trivial rules if they are disputed (even if, in Maimonides’ eyes, for incomprehensible reasons). This point is self-evident in light of what we said above about the polemical aim of the roots (against the hegemonic position of Bahag and his followers).

From this we may infer that Maimonides’ fourteen rules do not constitute a complete set of his meta-halakhic principles regarding the enumeration of the commandments. There are many more rules—simple and undisputed ones—that he did not mention.

A further important conclusion follows. Even if we find someone who disagrees with Maimonides on a large majority of the roots (see the interpretation of Saadia Gaon cited above, according to which he disagrees with almost all of them), this does not mean that the difference between his enumeration and Maimonides’ will be dramatic. The reason is that a significant part of the rules is agreed and obvious, and therefore is not discussed here at all. Those will be shared rules, and they underlie the main body of the enumeration of the commandments. The composition of the roots presents only the innovative rules or those in dispute, and it is therefore quite natural that not a few of them should be contested. On the other hand, for precisely that reason, the influence of these disputes on the enumeration of the commandments is rather limited, since they concern only a small part of the principles on which it rests.

C. A Brief Survey of the Various Roots

Introduction

The roots enumerate fourteen principles that govern and guide Maimonides’ enumeration of the commandments. I shall present them briefly, make a few short remarks on each one, and afterward discuss them more generally and classify them.[56] Maimonides’ own formulations are highlighted (see the list of roots at the end of his introduction to Sefer HaMitzvot), and my remarks appear immediately afterward in regular type.

The first principle among them is that one should not count within this total commandments that are rabbinic: Here Maimonides determines that rabbinic commandments are not counted within the 613. It should be noted that Bahag includes several such commandments, and therefore Maimonides found it necessary to discuss this rule despite its simplicity.

The principled reason there was room to include rabbinic commandments in the count is Maimonides’ own view that the obligation to observe the sages’ enactments and decrees is itself learned from the verse Do not deviate from all that they instruct you (see also the opening of the Laws of Rebels). For this reason, the disputes that arise around this root do not concern its underlying principle (which is probably agreed upon by almost all enumerators of the commandments), but rather the question of the source of the sages’ authority and the force of their enactments.

This aspect, however, raises a question that accompanies several of the roots: is the principle laid down here classificatory in character (these commandments are not to be counted because they are already included within the commandment do not deviate), or because these are commandments of a different kind and force (rabbinic, not Torah-level), and therefore they should not be included in the count for substantive reasons? There is also the possibility that these commandments are not counted because they are not explicitly commanded in the Torah, and in Maimonides’ enumeration only commandments explicitly commanded in the Torah appear (see more on this condition below). It should be noted that Maimonides’ arguments throughout this root do not at all touch the question of the halakhic force of rabbinic commandments, and this is more than a hint regarding his intention on the previous point. See further in the article on this root.

The second principle is that one should not count everything derived through one of the thirteen hermeneutical principles by which the Torah is expounded, or through amplification: This root is Maimonides’ most revolutionary innovation (perhaps in his halakhic teaching in general). On the plain sense of his language, he argues that every law or commandment derived through exegesis cannot be included in the enumeration of the commandments. In the course of his remarks he elaborates further, and it seems that in his view these are rabbinic laws.

It should be noted that Maimonides’ commentators disagreed over how to understand his intention in this root (see the article on this root): Nachmanides and those who follow him understood his words literally, namely, that these are indeed rabbinic laws. Rashbatz and those who follow him interpreted Maimonides as speaking not on the halakhic plane but on the interpretive plane—that is, that these laws are rooted in the ‘scribes’ (= the sages), and not in the Torah itself; but clearly the status of these laws is that of Torah-level laws.

The question still arises why these commandments are not counted. Neubauer, in his aforementioned book, sees this root as an application of the previous root, which determines that rabbinic commandments are not counted. But from Maimonides’ own words, and from the fact that he devotes a distinct root to this principle, that does not seem to be his intention. One can view the principle in this root as a classificatory principle—namely, these commandments are not counted because they are included in the ‘parent’ commandment (= that learned from the plain meaning of the verse from which they are derived). Such an understanding does arise in Nachmanides’ glosses (regarding the obligation to honor step-parents and an older brother, where he explains that these are included in the commandment of honoring father and mother), but Maimonides apparently does not mean this.

It can be shown that Maimonides’ intention here is that these commandments are not to be counted because there is no explicit command concerning them in the Torah. This is based on Maimonides’ interpretive conception, according to which exegesis is not another interpretation of the verse but an expansion of it (for Maimonides, each verse has only one meaning, and that is the plain meaning).[57] For this reason, Maimonides sees the Talmudic rule Punishments are not derived by legal inference as the source for the principle that one does not punish for violations derived from exegetical interpretation (and not only from an a fortiori argument, as appears from the Talmud and as all other medieval authorities understood it; and Nachmanides attacks him sharply on this in his glosses to the second root). See also Maimonides’ introduction to Sefer HaMitzvot, in the section after the fourteenth root (where he states this explicitly).[58]

In the article on this root I argued that Maimonides means an intermediate model (that is, he does not at all accept the conventional dichotomy between Torah-level and rabbinic law in its simple form; in his view there is a continuum of several halakhic levels between those two poles). All the rules dealing with Torah-level and rabbinic laws must be interpreted non-dichotomously; they are classified according to the types of laws and their character, not according to their halakhic rank. Maimonides does not recognize a hierarchical axis of legal force and status in the sense familiar to us.[59] Nachmanides, who understood the matter literally, vigorously protests against this innovation, arguing that it runs against the tradition in our hands and against dozens and hundreds of sources in the literature of the Sages. This is precisely the reason for Rashbatz’s innovation, which was indeed accepted by most later commentators on Maimonides, especially the later authorities. As noted above, this root received the greatest amount of attention among Maimonides’ commentators and modern scholars.

The third principle is that commandments that do not apply for all generations should not be counted: Here Maimonides excludes from the count commandments that were stated only for their own time, such as the commandment to make a fiery serpent or the jar of manna. The enumeration includes only commandments given to us for future generations.

Here too one must ask whether Maimonides regards these directives as commandments that simply are not eternal, or whether they are not commandments at all in the accepted halakhic sense. See the article on this root.[60]

The fourth principle is that one should not count the general commands that encompass the entire Torah: In this root Maimonides excludes general commandments that lack concrete content, such as and you shall keep all My commandments. This is a command that includes the whole Torah.

Here too it is unclear whether the meaning is that such directives are not commandments at all, or that they are not counted because of redundancy. See the article on this root, where I argued that Maimonides and Nachmanides likely disagreed about this (both here and with respect to positive commandment 5).

The fifth principle is that one should not count the reason for a commandment as a separate commandment: In this root Maimonides deals with verses that have the appearance of a command but whose function is to provide the reason for a commandment. For example, regarding a king it says: He shall not have too many wives, lest his heart turn astray. The second part is not a command about his heart’s turning aside, but the reason for the first part.

In the course of the discussion between Nachmanides and Maimonides on this root, the question of deriving law from the reason of a verse arises, and it becomes clear that Maimonides opposes such derivation even where the reason is written explicitly in the Torah. In my article on this root I showed that from here it emerges clearly that Maimonides does not base the rule that one does not derive law from the reason of the verse on fear of error, but on a different foundation. Nachmanides, in his glosses here, tends to think that there are no mere ‘reasons’ in the Torah if they have no halakhic significance at all.

The sixth principle is that when a commandment contains both a positive commandment and a prohibition, its positive element should be counted among the positive commandments and its prohibitive element among the prohibitions: This root is a reservation to the rule appearing in the ninth root (which excludes redundancies). It determines that when there is duplication between a positive commandment and a prohibition we include both in the enumeration of the commandments.

This determination raises the question what the difference between a prohibition and a positive commandment really is, since there are cases in which the positive commandment and the prohibition have the very same content (such as refraining from labor on the Sabbath). Indeed, precisely for this reason, RYFP rejects Maimonides’ position and supports Saadia Gaon’s method, which does not count two commandments even in the case of duplication between a prohibition and a positive commandment.

In my article on this root I explained the basis for the substantive distinction between a prohibition and a positive commandment, and showed that according to Maimonides and his school it is not directly related to the distinction between positive action and passive omission, except in a theoretical sense. Later in his remarks on this root Maimonides deals with a prohibition derived from a positive commandment, and there too the question of the distinction between prohibition and positive commandment naturally arises. See also Nachmanides’ glosses there.

The seventh principle is that one should not count the laws of a commandment: This root excludes from the count occurrences of different applications of the same commandment. For example, the three possibilities in a variable offering are counted as different applications of one commandment (each concerns a person of a different economic standing—wealthy, poor, and very poor), not as three commandments.

In my article there I show that underlying Maimonides’ words in this root is the question of generalization: what is the relation between particulars and the general category composed of them? Maimonides holds a rationalist conception, according to which the logical operations of the intellect are the source of our understanding of the facts (= the commandments, which are the objective facts in the halakhic context). In my article I showed that Maimonides uses here two classic rationalist tools: generalization and the positing of causal connection, both of which were attacked by empiricists (the first by the seventeenth-century British philosopher David Hume), and empiricist conceptions underlie the glosses of those who disagree with Maimonides (R. Daniel the Babylonian and Nachmanides).

The eighth principle is that one should not count the negation of an obligation together with a prohibition: This root determines that sometimes the Torah uses language that looks like a prohibition, but its intention is only to establish a fact or its negation (that is, to negate a law in a certain situation), and not to forbid something. Maimonides determines that such verses are not included in the enumeration of the commandments.

In my article on this root I showed that the basis of this determination is the distinction familiar in analytic philosophy between two different conceptions: evaluative/normative statements as opposed to descriptive or factual statements (prescription versus description).

I then showed there that there are several types of negation of obligation, and that one can in fact prove that, at least according to Nachmanides, there are also positive verses (not only negations) that are not counted for the same reason.[61]

The ninth principle is that one should not count the prohibitions and the positive commandments, but rather the matters that are prohibited and the matters that are commanded: This root excludes counting duplicate commandments, whether prohibitions or positive commandments. For example, the Torah commands observance of the Sabbath twelve times, and we count this as only one commandment.

The question that arises here is why the Torah repeats the same commandment several times, and what the meaning of that repetition is. Maimonides’ claim is that the Torah repeats only in order to emphasize the importance of the commandment in question. Nachmanides sharply objects to this, and in his view there is no such superfluity in the Torah. See more in my article on this root.

At the beginning of his remarks on this root, Maimonides presents a puzzling and detailed introduction that divides the commandments by type: commandments of thought, speech, character traits, and action. I explain that his intention was not only to sharpen what we also saw in the previous root, namely that ‘commandment’ always means a normative instruction of one of these types. His intention was to say that in every commandment or transgression there are two aspects: obedience or rebellion against the command, and the repair or damage done in reality, in the soul, or in society. Any command that lacks such an instruction is not counted in the enumeration of the commandments. Hence what we saw in the second root: where there is no command, we do not count; and here we complete the picture and determine that where there is no unique substantive content, we do not count.

Below we shall see that this emerges almost explicitly from Maimonides’ words in this root, and once this is understood, commentators’ difficulties disappear (they objected to an apparent contradiction between the two parts of the root, since in the second part Maimonides determines that an inclusive prohibition is counted as one commandment, even though it contains several different halakhic commands).

The tenth principle is that one should not count preliminary steps that serve a single end: Here Maimonides determines that a preparatory act for a commandment is not to be counted in the enumeration of the commandments.

One must ask whether in his view such preparation is not a commandment at all, and for that reason is not counted (in which case this principle belongs to the fourth category to be presented below), or whether it is included in the commandment for which it serves as preparation (in which case it belongs to the fifth category). In my article on this root I show that Maimonides has a unique method on this matter and that a preparatory act occupies an intermediate status.[62]

The eleventh principle is that one should not count the components of a commandment individually, part by part, when all of them together constitute a single commandment: Here Maimonides determines that parts of a commandment are to be counted together. For example, the four species are counted together as one commandment, not as four commandments.[63] In the course of his remarks here Maimonides discusses the relation between parts of a commandment that do or do not invalidate one another. He argues that parts that invalidate one another are certainly parts of one commandment, but parts that do not invalidate one another may sometimes be counted separately and sometimes together.

In my article I explain these different kinds of relation in terms of the question of the relation between particulars and the whole (which also arises in the seventh root), and in terms of the question of how to understand the essence of concepts (whether a concept is merely a linguistic convention, or whether it has an essence of its own that precedes its parts and characteristics).[64]

The twelfth principle is that one should not count the component acts of a labor for which there is a command to perform it, each component as a separate commandment: Here Maimonides deals with halakhic procedures, such as the offering of a sacrifice, and determines that the whole procedure, with all its parts, is counted as one commandment. On the other hand, he limits this and determines that not every obligation involving a given sacrifice is counted as an independent commandment; rather, each kind of sacrifice is counted as an independent commandment.

The distinction between this root and its predecessor is not clear (and some of Maimonides’ commentators even interchange them because of the similarity). The previous root speaks of parts of a commandment, whereas this one speaks of parts of an act/procedure (which is itself also a commandment). In my article I explain the differences between these roots. The point is that in this root the discussion concerns explanatory sub-sections of existing commandments (such as the explanation of what ‘sprinkling’ is, or what the procedure of offering a sin-offering is), while the command itself appears separately (in the commandments imposing the obligation to bring a sin-offering for certain violations). By contrast, in root 11 each of the parts is itself a halakhic obligation (to take a lulav, an etrog, myrtle, and willow), but those obligations are performed together.

For this reason, the question whether Maimonides’ enumeration is systematic or substantive arises here with great sharpness. We shall deal with this question below.

The thirteenth principle is that commandments should not be multiplied according to the number of days on which that commandment is obligatory: Here Maimonides deals with commandments that recur on several different days (such as the additional offerings of the New Moon, or the additional offerings of festivals that continue for several days—Sukkot and Passover), and determines that every such commandment is counted only once.

One can find in Maimonides’ remarks here a continuation of his conception of time, as it appears in the third root. In principle, this is of course a root dealing with redundancy.

The fourteenth principle is how the establishment of boundaries is to be counted among the positive commandments: Here Maimonides determines that punishments for transgressions are not each counted separately; rather, each type of punishment (lashes, stoning, beheading, strangulation, and so on) is counted as a commandment.

It is not clear whether he means that punishment is not a commandment at all, but a legal consequence of the transgression, and therefore should not be counted; or whether he means that punishment is not counted because it is included as a detail in the transgression for which it is imposed (but if so, this root is superfluous, since roots 11–12 have already established this). I discuss this in the article on this root.

Here too the question of the character of the enumeration of the commandments (systematic or substantive) arises, as we shall see below.

D. Different Classifications of the Principles in the Roots and Their Significance

Introduction

We have already noted that most of the treatments we find among commentators on Maimonides’ roots concern halakhic determinations, or halakhic disputes, that arise in the course of the discussion. Both in traditional literature and in academic literature, very few deal with the principled determinations of the roots themselves (such as the questions raised in the survey in the previous chapter). In this chapter I shall discuss the nature of the roots by presenting two different classifications of these rules, based on their different characteristics as presented in the previous chapter.[65] The classifications presented here are based on two attempts made in the literature to characterize and define the kinds of rules found in the roots. The considerable similarity between the two attempts indicates that these are probably reasonable and natural characterizations of the roots.

It should be noted that in the book Mitzvah Berah (part 1, beginning on p. 11) a completely different division of the roots is proposed. Several of them are based there on the principle that each counted commandment should in some way include a complete act by which one merits the world to come. I shall not discuss that division here, since it belongs more to theology than to methodology.

It is important to note that the different divisions do not have direct consequences, and it is difficult to define a ‘correct’ or ‘incorrect’ division of the roots. Clearly the same collection of objects, or principles, can be divided in several ways.[66] Indirectly, however, the form of division can certainly have many consequences, at least for understanding the aim of the various roots, but this is not the place to elaborate.

The rules for counting the commandments from a general perspective

We can see that most of the roots are rules of exclusion; that is, they determine which commandments are not to be counted. There is only one exception, namely the sixth root, which presents a consideration of what is to be counted (when there is a parallel prohibition and positive commandment, both are to be counted, unlike two positive commandments or two prohibitions—as detailed in the ninth root). In the fourteenth root as well there is a combined rule of counting (each type of punishment separately) and non-counting (not each individual punishment-obligation separately).

The non-counting is based on different kinds of reasons. Sometimes it is based on substantive considerations, such as halakhic status (rabbinic commandments are not to be counted—the first root), or on the ground that a verse is not a commandment at all (as in the fifth, eighth, and other roots). At other times it rests on considerations of grouping and classification (the seventh and twelfth roots and others), or on other technical considerations, such as redundancy (the ninth root).

In general, when a given commandment is not counted in Maimonides’ enumeration, this does not mean that it is not a Torah commandment. There are many reasons why a commandment might not be counted (if it is included in another commandment, or constitutes a preparatory act for another commandment, and so on), even if halakhically it is defined as a Torah-level commandment.[67]

In the introductory article to the roots I proposed dividing the roots into five principal groups, or categories:[68]

1. Rules concerning the force of the commandments. These rules determine principles of counting or non-counting because of halakhic force. In the first root Maimonides determines that rabbinic commandments are not to be counted (though see my remarks on this in the previous chapter). In the second root too he determines that commandments derived by exegetical interpretation are not to be counted, and these too he calls ‘rabbinic commandments.’ If so, these two roots are principles of non-counting because of halakhic force (see also the remarks in the previous chapter on the second root). It seems that only the first and second roots are included in this category.
2. Rules concerning time. These rules determine principles of non-counting that relate to the axis of time. In the third root Maimonides determines that commandments not applicable for future generations are not to be counted. In the thirteenth root Maimonides determines that the number of commandments is not to be counted according to the number of days on which the commandment applies; this too belongs to this category.

It is important to note that this category has no methodological independence of its own. From that perspective there is not much common ground between these two roots, and the category was defined only for convenience in discussing them together. The third root deals with verses that are not commandments at all in the halakhic sense (even though they are imperative verses in the grammatical and substantive sense), and as such it ought to be included in the fourth category (defined below). The thirteenth root, by contrast, does not concern halakhic force at all, since no one disputes that the festival additional offerings are Torah commandments. It concerns classification and categorization, and as such it ought to be included in the fifth category, or perhaps the third (of redundancy).

The definition of these two roots as an independent category was made only in order to present the conception of time that finds expression in them. For our purposes here, therefore, this category is not relevant. As stated, only the third and thirteenth roots are included in it.

3. Rules concerning redundancy. These rules determine principles of counting and non-counting on the basis of the principle that there is no redundancy among the 613 commandments. Here too we are not dealing with questions of halakhic force, since according to everyone these are Torah commandments. The rules of non-counting in this type stem from the fact that the commandments under discussion already appear as other counted commandments. It seems that this type includes the fourth root (though with respect to it the classification is not unequivocal; see the article on that root, where I suggested that, at least according to Maimonides, it should specifically be seen as belonging to the fourth category. Most discussions, however, see it as one of the roots of the third category, and Nachmanides certainly understood it that way), the sixth, and the ninth.
4. Rules concerning the character of the commandment as a command. For example, the fifth root determines that the reason for a commandment is not to be counted as a commandment in its own right. In the plain sense (though see my article’s challenge to this), the reason is that the reason for a commandment is not itself a command, and therefore should not be counted. So too with the eighth root, which determines that the negation of an obligation is not to be counted because it is not a prohibition. So too with the tenth root (not to count preparatory acts for commandments),^[69] and perhaps also the fourteenth root (whose classification too is unclear; see my remarks on it in the previous chapter). It seems that this category includes the fifth, eighth, tenth, and perhaps also the fourteenth root.
5. Rules concerning classification and categorization of commandments. For example, the seventh root determines that one should not count the laws/details of a commandment, because these are already included in the counted commandment. This type includes the seventh, eleventh, and twelfth roots, and perhaps also the fourteenth.^[70]

A general note on the division

In the division I proposed, the classification is not unequivocal with respect to at least some of the roots. For example, the fourth root, which excludes counting commandments that encompass the whole Torah, can be interpreted as part of the roots of redundancy (for it has no content not already included in other commandments), and this is indeed how Nachmanides saw it (in his glosses to the root and to positive commandment 5). But in my article I showed that Maimonides also sees these verses as lacking the content of a command (and not merely as redundant with other commands). If so, it may be included both in the third category and in the fourth. I already noted several other examples above.

It should also be noted that the roots of redundancy, by their very nature, are roots of classification and categorization, and therefore it is far from clear that the third category should stand on its own as a distinct category rather than be included in the fifth.[71] The same is true, as we saw, of the second category, which is not at all distinct from the others at the substantive level. The fourth category too could be included in (or could include) the first category, since all of these roots speak about verses that are not commandments in the full halakhic sense.

A further look at the counting rules: the classification of Hanina Ben-Menahem

Ben-Menahem deals with the question of individuation of commandments among the Sages, in Maimonides, and in Ibn Ezra.[72] Out of that discussion he extracts a principled division of the principles in the roots. To this end he introduces a concept drawn from jurisprudential literature, following Jeremy Bentham,[73] who defined principles of individuation of laws.

An ‘individuation theory’ in the general sense is a theory that instructs us how to divide a given field into distinct entities.[74] In the legal context, individuation deals with dividing legal material into distinct units. As Ben-Menahem notes,[75] these principles of division may differ in character. Some principles concern the philosophy of law and have no legal consequence whatsoever, certainly not from the standpoint of the addressee of the law.[76] These he calls principles of meta-legal individuation. By contrast, there are principles that have consequences for the implementation of law, such as the fixing of punishment for an offense, which depends inter alia on how many distinct offenses a given person committed (for example, if he drove unlawfully or embezzled funds). These he calls principles of legal individuation.

There are several ways to carry out meta-legal individuation, and a fine illustration appears in Makkot 23b, following the statement of R. Simlai:

David came and reduced them to eleven, as it is written: “A psalm of David: Lord, who may sojourn in Your tent, who may dwell on Your holy mountain? One who walks blamelessly, acts righteously, and speaks truth in his heart; who does not slander with his tongue, does no evil to his fellow, and casts no disgrace upon his relative; in whose eyes a vile person is despised, but who honors those who fear the Lord; who swears to his own hurt and does not change; who does not lend his money at interest, and takes no bribe against the innocent. One who does these shall never be moved”…
Isaiah came and reduced them to six, as it is written: “One who walks righteously and speaks uprightly, who despises the gain of oppressions, who shakes his hands free of taking a bribe, who stops his ears from hearing of bloodshed, and shuts his eyes from looking upon evil”…
Micah came and reduced them to three, as it is written: “He has told you, O man, what is good and what the Lord requires of you: only to do justice, love kindness, and walk modestly with your God”…
Isaiah returned and reduced them to two, as it is said: “Thus says the Lord: Keep justice and do righteousness.”
Amos came and reduced them to one, as it is said: “Thus says the Lord to the house of Israel: Seek Me and live”…
Rather, Habakkuk came and reduced them to one, as it is said: “But the righteous shall live by his faith.”

Although the Riban on the spot interprets this statement literally (that the various prophets exempted the people from some of the commandments), the simpler interpretation is that what we have here are different ways of counting the commandments and arranging them at a meta-legal level.[77]

Ben-Menahem, in Individuation, ch. 7, distinguishes between filtering rules and individuation rules, and explains that not all of Maimonides’ roots are rules of individuation. He defines two kinds of filtering rules within Maimonides’ roots:

• Identification rules: rules dealing with the identity of the material that is to be broken down. He explains that this category includes Maimonides’ first three roots, which exclude from the counted material rabbinic commandments, commandments derived through exegesis, and commandments not applicable for future generations. In another article,[78] he explains that these are normative propositions that, for various reasons, are not included in the enumeration.
• Interpretive rules: rules determining the nature of the material (whether it belongs to the normative sphere: what is a command and what is not). He includes in this category roots 5–8[79] and root 10.

It should be noted that the second category overlaps exactly with my fourth category.[80] The first category, however, has no parallel in my classification, and indeed I am not clear on the logic by which Ben-Menahem includes the first three roots in one category. Had I not created a special category for roots dealing with the conception of time, then the third root too should have been included in the fourth category, since commandments not applicable for future generations are not really commands in the accepted sense.[81] Alternatively, he should have included in this category also the roots of the fourth category in my division.

After the filtering, says Ben-Menahem, we are left with halakhic (= normative) material, whose classification and categorization are carried out by means of individuation rules. He argues that Maimonides’ principles of individuation are likewise divided into two kinds:[82]

• Rules that clarify the concept of individuation. These rules do not provide us with instructions for dividing the whole, but only establish the principles. For example, in root 11 Maimonides determines that ‘it is not appropriate to count the parts of a commandment separately, part by part, when their combination is one commandment.’ Maimonides does not determine how we know what is part of a commandment and what is not; he establishes only the principled determination that if there are parts of commandments, they are not counted in their own right but only as part of the whole.

The seventh root, which deals with different applications of a commandment, is also a clarificatory rule of this type. The same applies to the twelfth root. We can now see that this category overlaps entirely with the fifth category in my division presented above (with the same reservation regarding the fourteenth root).[83]

• Rules that provide practical guidelines for dividing the whole. For example, the ninth root determines that if a commandment recurs in Scripture several times, we are to count all those commands as one commandment.^[84]

Additional roots belonging to this category are the sixth and thirteenth roots. We can now see that, at least in practice, this is a category that overlaps with my third category, which dealt with redundancies.[85] The reason is that a practical criterion (that is, one including concrete and practical instructions) for individuation is always based on redundancy.

A concluding note to the chapter

From the classifications presented here it emerges that the order in which Maimonides chose to present his fourteen roots is not clear. It does not seem that there is a meaningful arrangement here, and that is uncharacteristic of Maimonides’ systematic oeuvre. However, in a composition of the kind represented by the roots, containing only fourteen principles—that is, very brief in relation to Sefer HaMitzvot, and certainly in relation to the Mishneh Torah—perhaps Maimonides did not think it worthwhile to dwell on the order and invest effort in achieving a more complete and systematic arrangement.

Moreover, as I already noted above, his intention in writing the roots was mainly to undermine the dominance of Bahag’s enumeration and ‘conquer’ the stage from it; afterward he did not see in them genuine importance (although he apparently later changed his mind, as noted above). This, I suggested, explains the fact that he composed them in Arabic. If so, it is no wonder that he did not invest great thought in the arrangement of these roots.

E. What Is a Commandment?

Introduction

In this chapter I shall try to propose a general picture of the concept ‘commandment’ as it emerges from the roots, and on that basis also to ground the map of the roots as presented above and their classification into the various types.[86] In the course of this I shall point to a certain problem in the classification proposed by Ben-Menahem.

Positive commandments and prohibitions

In the sixth root Maimonides notes that duplication between a prohibition and a positive commandment is not like duplication between two positive commandments or two prohibitions. That is, he sees a categorical distinction between positive commandments and prohibitions, irrespective of contents that may sometimes overlap (as with the commandment to observe the Sabbath and the prohibition against desecrating it, which at the practical level have exactly the same content).

In my article on the sixth root I discussed at length the distinction between positive commandments and prohibitions. In any case, it is clear that this distinction does not depend only on the language of Scripture; it also has a conceptual root. My conclusion there is that a prohibition is the Torah’s indication of an undesirable state (Do not profane the Sabbath), while a positive commandment is the Torah’s indication of a desirable state (You shall keep the Sabbath). As I showed there, according to Maimonides there is no necessity that a positive commandment require an action and a prohibition command abstention from inaction.[87] We can now understand that there is a difference between a positive commandment to refrain from labor on the Sabbath and a prohibitory commandment forbidding labor on the Sabbath. For example, with regard to a person who did no labor because of some coercive circumstance, one might say that he did not fulfill the positive commandment,[88] but also did not transgress the prohibition.

As already mentioned, Maimonides does not treat additional kinds of commandments (such as sections, or commandments imposed on the community) as separate categories. In the period after him, almost no one does so either. As we shall see below, however, Maimonides too has another category, namely procedural commandments, or halakhic definitions.

Content and command[89]

In the ninth root Maimonides deals with duplicate commands in Scripture. This root is divided into two parts: the first excludes counting duplicated commands separately (for example, commands regarding Sabbath observance appear twelve times in Scripture, yet we should count only one commandment of Sabbath observance). In the second part Maimonides deals with the inclusive prohibition and determines that all the branches derived from it are to be counted as one commandment.[90] For example,[91] the verse Do not eat over the blood commands us not to eat from an animal before its soul has departed, and likewise not to eat sacrificial meat before the sprinkling of the blood, and not to eat on the day when someone is led out to execution, and not to eat before prayer,[92] and in addition it also serves as the warning for the rebellious son. Here we have several commandments derived from one verse, and apparently there is no substantive connection among them; yet Maimonides determines that all of them are to be counted as only one commandment (prohibition 195, regarding the rebellious son).

RYFP remarks in his comments to this root that, at first glance, the two parts of the root contradict one another. The first part instructs us that, for purposes of the enumeration of the commandments, a ‘commandment’ is only a command with unique content of its own. Commands that appear explicitly in Scripture are not counted because they lack unique and independent content (they repeat the content of another command already counted). From this it would seem that what determines the enumeration of the commandments is content, not the existence of a command in Scripture. By contrast, the second part yields the exact opposite conclusion: what determines the count is the existence of a command, and therefore commandments that differ in content, but are derived from one command verse (in the case of the inclusive prohibition), are not counted except as one prohibition. From this it appears that what determines the count is the existence of a command, not unique content.

The simple resolution of this contradiction is of course that, according to Maimonides, both requirements are necessary. We learn from here that for a given commandment to be counted separately in the enumeration of the commandments, two conditions are required: 1. There must be an explicit command for it in Scripture.[93] 2. It must have content unique to itself (that is, it must not repeat the content of another commandment already counted). The first part of the ninth root deals with the second condition, and the second part deals with the first.

The basis of the matter is apparently the understanding that in every commandment or transgression there are two aspects: the command and the content. Every commandment a person performs has two aims or two benefits: fulfilling God’s command, and the specific benefit for which the commandment is intended (= the reason for the commandment). The same is true of transgressions. Every transgression has two aspects: rebellion against the command, and the damage/harm that the transgression is intended to prevent.[94]

From here we can understand Maimonides’ remarks in several roots where he prefaces his discussion by noting that the commandments under discussion appear explicitly in the Torah.[95] Without the appearance of the command in the Torah, there is no place at all to discuss counting the commandment.[96] Once the commandment appears in the Torah, the discussion begins about the manner of its appearance in the enumeration of the commandments (both for reasons of individuation and for filtering reasons. As we shall see below, it is difficult to divide these kinds of roots and rules in the way Ben-Menahem proposed).

Application to the first two roots: what is a Torah-level commandment?

In my articles on the first and second roots I argue that Maimonides does not accept the division between Torah-level and rabbinic law as a sharp dichotomy. He has several categories of rabbinic commandments (or ‘words of the scribes’), but their halakhic status and definitions differ. For example, doubt regarding a law given to Moses at Sinai is treated leniently,[97] whereas doubt regarding laws derived through exegesis is treated stringently.

In those articles I explained this by saying that a law given to Moses at Sinai is a commandment for which we have a command, but which lacks objective Torah-level content. There is here no defect or repair for whose sake we were commanded in this commandment (rather, it was stated for the purpose of a safeguard, or to define the boundary of an explicitly written commandment, such as the measures that are a law given to Moses at Sinai). By contrast, laws derived through exegesis have objective Torah-level content, but we have no explicit command in the Torah regarding them. In this case, the rule is that doubt is treated stringently (see a detailed explanation, and the reasoning for it, in my article on the second root).

Thus, a Torah-level commandment is only a commandment with respect to which we have a command, and which is also intended to achieve some repair or prevent some corruption. Only such commandments are counted in Maimonides’ enumeration of the commandments. As I already noted, other commandments that are not counted are not punishable, because according to Maimonides that would be ‘punishment derived by inference’ (see his introduction to Sefer HaMitzvot, immediately after the fourteenth root).

In this way Nachmanides’ powerful objection in his glosses to the second root is also resolved. Maimonides determines (both in his introduction to the Commentary on the Mishnah and in the second root) that a commandment that comes to us by tradition but has no anchor in the Torah (whether plain meaning or exegesis) is not Torah-level. He likewise determines that a commandment that it has only an exegetical anchor (and not one in the plain meaning of Scripture) is also not Torah-level. But if both conditions are met—namely, we have a tradition concerning it and it is also alluded to in exegetical reading of the verses (that is, the exegesis is supportive rather than creative; see my article on the second root)—then it is a Torah-level commandment. Nachmanides asks: what is there in the conjunction of the two conditions that is not present in either one alone? If exegesis is not a sufficient source, and tradition is not a sufficient source, how does the combination of the two suffice to give the law in question Torah-level status and even include it in the enumeration of the commandments?

According to my approach, the explanation is simple: if there is a tradition regarding this law, then that proves that the exegesis reveals what is present in the verse, and therefore there is here not only repair/content but also a command (unlike creative exegesis, that is, without tradition concerning the law derived). Therefore this law is Torah-level. But if one of the two components is lacking, then we are dealing with a law for which we have no command in Scripture, since, as already noted above, according to Maimonides exegesis expands what is written in the plain sense of the verse rather than uncovering what is already present there. Such a law is not a Torah-level law. For Maimonides, a Torah-level law is a law explicitly written in Scripture,[98] without direct relation to its halakhic status (which is merely a derivative of the fact that it is explicitly written in the Torah). In effect, Maimonides identifies the term ‘Torah-level law’ with the term ‘Written Torah.’

This identification is not accepted by the other medieval authorities, because usually the division between Written Torah and Oral Torah expresses source rather than halakhic force, whereas the division between Torah-level and rabbinic pertains to halakhic force without direct connection to the question of source. As noted, Maimonides identifies these two divisions, and there is abundant evidence for this in his writings.[99] Likewise, he does not regard the term ‘rabbinic’ as unitary, and therefore he divides it into several distinct sub-shades (also in their halakhic implications, as noted).

Application to additional roots

We have seen that this picture, which sees in every commandment two faces—repair/corruption and obedience/rebellion—is what underlies the roots of redundancy. But it is important to understand that here too there is usually a fundamental organizing principle, as in Maimonides’ other roots. For example, in the first root, which excludes counting rabbinic commandments, Maimonides mentions that the obligation to obey the sages is based on the verse Do not deviate from all that they instruct you. Nachmanides (and many others) already objected that according to this, rabbinic commandments have Torah-level force, and therefore Maimonides should seemingly also have counted rabbinic commandments. But in light of what we have said here, we can understand that indeed there is a command to listen to the sages, but the content of those commandments does not exist at the Torah level. We do not refrain from eating poultry with milk because of the obligation to obey the command do not deviate, but the act of eating itself does not produce a defect; rather, the rule comes to prevent eating meat with milk, which is forbidden by the Torah. If so, we have here a command without content, and therefore it is not counted.[100]

All the roots of classification and categorization (my fifth category) are also based on the absence of unique content. In all these cases we are dealing with commandments for which we have an explicit command, and nevertheless they are not counted because of the absence of unique content. For example, in the seventh root it becomes clear that we do not count different applications of a commandment, even in cases where we have an explicit command (as with the variable offering), since there is one content applied in different ways under different circumstances. According to the eleventh root we do not count parts of a commandment (such as the four species), even though we have an explicit command for them, because the contents of all the parts combine into one general content. This is another mode of absence of unique content that prevents counting. The same applies to the twelfth root (which is very similar to the eleventh).[101]

The roots of the fourth category

The exceptions are the roots of the fourth category. These roots deal with characterizing the term ‘commandment’ in its accepted normative sense. For example, the fifth root deals with reasons for the commandments. There too the verses are explicit, for otherwise there would be no room even to imagine counting them as commandments. For example, the second part of the verse He shall not have too many wives, lest his heart turn astray teaches us the reason for the prohibition on a king’s multiplying wives. This is a reason, not a command, and therefore it is not to be counted. Here of course there is no command, nor halakhic content. The same applies to the eighth root (the negation of obligation) and the tenth (preparatory acts), for these too deal with verses that command nothing. All of these are descriptive verses, not imperative ones (for this definition and its significance with regard to the enumeration of the commandments, see the article on the eighth root).

In Ben-Menahem’s terminology, one can say that these points apply only to the filtering rules (the rules that sort the propositions belonging to the normative sphere), and especially to interpretive rules (not identification rules). Thus, the two requirements from every commandment—that it have a command and also unique content—are requirements that underlie the rules of individuation, that is, the rules of classification and categorization, since these operate among normative propositions themselves. The first and second roots belong to the identification rules, and as we have seen, although they are a kind of filtering rule, they certainly belong to the normative sphere. Their non-counting stems from the absence of one of the two requirements (unique content or command). From this angle too one can see that Ben-Menahem’s division is not methodologically efficient. The interpretive rules (which in his view are part of the filtering rules) are in fact more similar to the rules of individuation, since they too belong to classification within the normative sphere.

Commandments that are not commands[102]

Against the background of what I have said in this chapter, it is important to attend to another point. Maimonides, in several places, notes that his enumeration of the commandments includes commandments that are not commands.[103] For example, in positive commandment 95, which deals with the annulment of vows, he writes:

Positive commandment 95: that we were commanded to adjudicate regarding the annulment of vows. That is, the Torah instructed us to judge according to those laws. The point is not that we are necessarily obligated to annul vows in every case. This same principle should be understood whenever you hear me count one of the legal rules, for there is not necessarily a command to perform an action. Rather, the commandment is that we are commanded to rule according to this law in this matter.

The ‘commandment’ of annulment of vows commands us nothing. It merely details the laws of annulment of vows, but there is no obligation imposed on us here: neither a positive commandment to annul, nor a prohibitory commandment to annul or not to annul. It merely defines a state of a vow as annulled or valid; there is no normative instruction here at all. It is important to emphasize that the laws of vows themselves appear as independent commandments: the obligation to fulfill a vow (positive commandment 94) and the prohibition on violating it (prohibitions 157 and 158). Thus in commandment 95 there is only a technical definition of what is called a ‘vow’ (a vow that has been annulled is not a vow, and therefore the above prohibitions and commandments do not apply to it), but no command at all.

And in commandment 96 (which, whether accidentally or not, appears immediately afterward), whose subject is the impurity of a carcass, Maimonides writes similar things:

And I will now mention an introduction that you should keep in mind throughout all that we mention concerning the various types of impurity. Namely, when we count each and every type of impurity as a positive commandment, this does not mean that we are obligated to become impure through that impurity, nor does it mean that we are warned against becoming impure through it, such that it would be a prohibition. Rather, the Torah’s statement that whoever touches this type becomes impure, or that this thing imparts impurity in such-and-such a manner to one who touches it—that is the positive commandment. That is, this legal rule with which we were commanded is the positive commandment: that one who touches such a thing in such a manner becomes impure, while one who is in such-and-such a circumstance does not become impure. Becoming impure itself is optional: if one wishes, one may become impure, and if one wishes, one may refrain from becoming impure…

And the commandment is what we were told in this law: that one who touches this becomes impure and will be impure, and will be subject to what impure persons are subject to—namely, to go outside the camp of the Divine Presence, not to eat sacred food, not to touch it, and the like. That is the command: that one be impure through this type when he touches it or when he is near it under such-and-such circumstances. Remember this principle for every type of impurity.

Maimonides is speaking about all the commandments that deal with the different kinds of impurity. In all these commandments there is no normative instruction at all. They are nothing but definitions of states of impurity. Here too we should emphasize that all the consequences of the different kinds of impurity receive commandments of their own, such as the prohibition of entering the Temple in impurity (prohibitions 77–78), the prohibition of eating terumah in impurity (prohibition 136), the prohibition of eating sacrificial food in impurity (prohibitions 129–130), performing Temple service in impurity (prohibitions 75–76), and so on. Yet Maimonides still finds it necessary to devote several commandments to the mere definition of the various states of impurity. Again, these ‘commandments’ define, but do not command. They inform us of some halakhic status or legal condition.

At first glance, such commandments have no place in the enumeration of the commandments, for two reasons: 1. They command us nothing (the element of command is lacking here). This reason belongs fundamentally to the fourth category (which excludes descriptive verses that are not commands). 2. They are merely details within other commandments. This reason belongs to the fifth category (classification and categorization, in which one does not count separately parts or details of a commandment that has already been counted).

It seems that Ben-Menahem erred in understanding Maimonides’ method on this point.[104] He is troubled by the presence of such commandments in Maimonides’ enumeration of the commandments, and resolves it by saying that all such definitional commandments are addressed to the religious court, and for the court they indeed impose a normative obligation (to adjudicate the laws of annulment of vows, and so on). He even infers this from Maimonides’ wording in positive commandment 95, where he wrote: ‘that we were commanded to adjudicate in this law.’ But, as stated, this is a mistake. Maimonides does not mean that the court must adjudicate this law; rather, every Jew must conduct himself in accordance with this law.

Aside from the fact that this is clearly his intention, it can be seen plainly from his words in commandment 96. There the subject is the laws of impurity, and these are not entrusted to the court (annulment of vows too is not entrusted to the court, but to the one who violated the vow—and that already belongs to the commandments of vows, not to the definition of the annulment procedure). Moreover, there the wording is not even of the type ‘that we were commanded to adjudicate this law’; rather, Maimonides writes, ‘this law with which we were commanded.’ That is, he determines that these definitions are included in the enumeration of the commandments even though they impose neither duties nor prohibitions, neither on the individual nor on the court.

In the next chapter I shall try to discuss this obscure issue further.

F. The Enumeration of the Commandments and the Meaning of Engaging in It: On Systematicity and Essentialism

Introduction

In this chapter I return to issues discussed at the beginning of our remarks, namely the place of Sefer HaMitzvot and the enumeration of the commandments within the whole of Maimonides’ halakhic oeuvre. We can now see that this issue may clarify the appearance of the definitional commandments in Maimonides’ enumeration, and perhaps other phenomena as well.

The aim of engaging in the enumeration of the commandments[105]

The occupation with enumerating the commandments was not very widespread among halakhic sages, and in the Talmud it is not discussed explicitly at all (aside from R. Simlai’s statement, which itself does not practically enumerate the commandments). Many of the enumerations up to that of Maimonides were written in the form of liturgical poems and Azharot (mainly for Shavuot and around the Ten Commandments), in ways that, at least in Maimonides’ view, convey lack of substantive precision and a preference for considerations of style and rhyme over halakhic and meta-halakhic considerations. If the considerations are stylistic and aesthetic, halakhic precision is liable to suffer (Maimonides’ mockery of these enumerations was mentioned above, and Ibn Ezra joins him in Yesod Mora at the end of the second gate).

Beyond the liturgical aims, among the enumerators of the commandments (usually the later ones) we find several additional reasons for the importance of engaging in the enumeration. The aim of Sefer HaChinukh, whose enumeration is built according to the order of the weekly Torah portions, seems more pedagogical: to extract a halakhic essence from each portion. For that reason the order of the commandments there is different. The same applies to the various liturgical poets mentioned above, who, beyond their poetic and aesthetic aims, some of them also wanted in this way to present some kind of digest of Jewish law as it emerges from Scripture (and sometimes everything is classified according to the Ten Commandments).

In the introduction to the short enumeration of the commandments by the Chafetz Chaim, which includes 77 positive commandments and 194 prohibitions, all of them among those applicable in our time, the issue seems to be preparation for fulfilling the commandment and you shall remember all My commandments (perhaps as part of the commandment of Torah study), but also a means for fulfilling those commandments (and therefore he brings only commandments applicable in our time).[106] Several sources for both of these directions are brought in the introduction to Mitzvot HaMelekh, ch. 7, and also in the introduction to Mitzvah Berah.

By contrast, Ibn Ezra sees the enumeration of the commandments as a theoretical-intellectual subject. He writes in Yesod Mora:[107]

Therefore, in every field of wisdom and every craft, a person does not seek to know the particulars, for a human being has no power to know their full extent or their number, since they never remain in one fixed form even for a moment, but are constantly passing away. General principles, however, endure forever.

Standing against all these is an interesting phenomenon connected with the Vilna Gaon. It is well known that the Vilna Gaon did not leave a single stone in Torah and Jewish law unturned, without engaging with it and offering innovations and emendations. Against that background, the lacuna in his treatment of the enumeration of the commandments is all the more striking. As far as we know, there is no written discussion by him of the enumeration of the commandments.[108] This phenomenon becomes clearer if we attend to what his brother, R. Avraham, wrote in his name at the beginning of his book Ma’alot HaTorah: many were perplexed by the enumeration of the commandments, and no one succeeded in creating a consistent method. It appears from his words that he means to say that the 613 commandments are merely general categories, and there is no importance to this, since it has no halakhic or other consequence (see more below).[109] The same was transmitted by his disciple in the introduction to Pe’at HaShulchan.

This approach leads us to the question of the consequences and halakhic significance of the enumeration of the commandments.[110]

The halakhic significance of the enumeration of the commandments

The great question raised by the Vilna Gaon is the question of the halakhic significance of the enumeration. Does the fact that a given instruction is counted in the enumeration of the commandments or not have any halakhic consequence? At first glance it seems not.[111] As we have already seen, most of the counting rules are not connected to the force and halakhic status of the commandments (apart from the first two roots). The considerations of the count concern redundancies, details included in a general commandment, and so forth. For this reason, in most cases the fact that a commandment is not included in the enumeration says nothing about the character or status of that commandment. Many Torah-level commandments are not included in the count for various technical reasons, and therefore the fact that a certain commandment is not included does not mean that it is not a Torah commandment. If so, why engage in this subject at all?[112]

In rabbinic literature over the generations several proposals were raised that try to find halakhic significance in the enumeration of the commandments. I shall briefly review a few of them:

1. Lashes – In a responsum of Maimonides[113] it appears that there is a consequence with respect to lashes; namely, one receives lashes for a counted prohibition, whereas if it is not counted one does not. So too Mabit suggested, in his introduction to Kiryat Sefer, ch. 6, as did Pri Megadim at the beginning of the ‘general introduction.’

However, it seems that Maimonides’ words there should not be understood in this way. His intention there concerns only the roots dealing with separating commandments into their parts,[114] where the indication for counting is lashes—that is, if one receives lashes for a certain part, then it is counted as an independent prohibition; if not, it should not be counted independently.[115] But one need not infer from there a general rule that one never receives lashes for an uncounted prohibition. One can even bring contrary examples.[116]

2. Halakhic status – Pri Megadim, at the beginning of his ‘general introduction,’ raises a number of possible consequences of inclusion in the enumeration of the commandments, the main one being that commandments not included in the enumeration have the status of rabbinic commandments. He claims that a half-measure would be forbidden in them only rabbinically, that cases of doubt should be treated leniently, and so on. But these claims are very puzzling. As we saw, only the first two roots introduce principles that depend on the force of the commandment and its halakhic status; all the rest are principles of classification, categorization, and filtering.[117]
3. Consequences of the total number – Standing against all these considerations is the fact that, in practice, very many halakhic consequences emerge from discussions of the enumeration of the commandments. Dozens and hundreds of halakhic conclusions and disputes arise from the disagreements of Maimonides and Nachmanides regarding the enumeration of the commandments, and likewise among their commentators. How can this be, in light of what we have seen—that the count as such has no halakhic consequences?

The answer emerges from RYFP’s introduction, in section 10 of his introduction (p. 56), which is entirely devoted to clarifying the question of the halakhic consequences of the enumeration of the commandments. His main claim there is that the total number itself is the most important consequence:

And all this is because necessity compelled them to complete the number 613 exactly, with no addition or subtraction, each according to his own approach. It follows, then, that this tradition of the number 613 is the source for many of the commandments and prohibitions, which otherwise have no other source at all. Had we not needed to preserve this exact number, it would never have occurred to us to make them into commandments and prohibitions without any source for doing so. Likewise, there are many commandments that they excluded from the total solely because otherwise we would end up with more than 613 commandments. As you can see in the words of Rashbatz (Zohar HaRakia, positive commandment 11), who wrote: “The conclusion from all this is that the words of the Gaon, of blessed memory (BaHaG), are closer to the plain meaning of the verse and to the straightforward sense of the statements of the Sages, unless the count compels us, when the full total of 248 positive commandments cannot be reached except by including the statement ‘I am’ in the count”…

The reason is that anyone who removes one commandment from the count must point to another commandment to enter in its place, in order to complete the overall total of 613, as tradition requires. See there several examples of such consequences.

It should be noted that Maimonides himself—see, for example, root 3 and elsewhere—uses such a consideration against Bahag: Others besides us also erred in this principle and counted [such items] because their resources were too limited…, by which he means that Bahag counted certain commandments because he was lacking several commandments needed to complete the overall total of 613.

So too Tashbetz writes in his book Zohar HaRakia (positive commandments 11):

The conclusion from all this is that the words of the Gaon, of blessed memory (= BaHaG), are closer to the plain meaning of the verse and to the straightforward sense of the statements of the Sages, unless the count compels us, when the full total of 248 positive commandments cannot be reached except by including the statement “I am” in the count.

It should be noted that this is not really a direct halakhic consequence of inclusion in the enumeration itself. For example, the fact that the commandment to honor parents is included in the count has no direct halakhic consequence. But the overall count may force us to exclude the commandment to conquer the Land from the enumeration, and in order to justify that we shall need various discussions and innovations about the character and halakhic significance of this commandment. Those discussions will sometimes (though not always) involve conclusions with halakhic implications.

Maimonides’ aim in the enumeration of the commandments[118]

As we saw above, Maimonides’ introduction to Sefer HaMitzvot indicates that the aim of the count is to create a theoretical skeleton for his comprehensive halakhic composition, the Mishneh Torah.[119] In Halbertal’s terms,[120] for Maimonides a commandment is an ‘organizing category,’ not necessarily a command. The commandments are no more than subheadings that serve as the classificatory skeleton for dividing the Mishneh Torah.[121]

From Maimonides’ words in several places it emerges that this count indeed has no halakhic consequences, and were it not for this aim he would not have entered into this question at all. From here we may infer that the considerations in this subject are not necessarily halakhic considerations (except perhaps in the first and second roots)[122]; rather, considerations of method and system (= logic of division) are also involved. We shall see examples of this phenomenon below.

It is therefore not surprising that we sometimes find that the considerations used for the enumeration of the commandments are not the considerations that determine halakhic status. There may be a principle of individuation that serves us in Jewish law, but it says nothing with respect to the enumeration of the commandments. For example, the prohibitions of neshekh and tarbit, in the beginning of ch. 4 of the Laws of Loans and Borrowing, appear as two separate prohibitions. That is the halakhic statement. By contrast, in Sefer HaMitzvot there appears only one prohibition on interest (prohibition 235; see the footnote above). Conversely, the commandment of tefillin is counted in the enumeration of the commandments as two separate commandments (12 and 13), whereas at the halakhic level the status of these two is not clear (this finds expression in questions concerning the relation between the blessings of tefillin and more; see the article on the eleventh root). Usually, if there is individuation in the enumeration of the commandments, then it also exists on the halakhic plane (every commandment counted separately is indeed a different and separate commandment halakhically as well), but the reverse is not necessarily true.

A comparison between Maimonides and Ibn Ezra

All these signs point to a unique tendency in Maimonides’ enumeration of the commandments. According to Ibn Ezra, who sees the enumeration of the commandments as a tool for theoretical meta-halakhic understanding of Jewish law, it stands to reason that the halakhic and methodological considerations regarding the enumeration of the commandments will largely overlap. In his view, the count is determined by the halakhic qualities of the commandments under discussion and by the halakhic and meta-halakhic relations among them. By contrast, according to Maimonides, who builds the enumeration of the commandments as a skeleton for his total halakhic edifice, the considerations may be entirely different. Here we would expect also considerations of system and method (determined by the question how a law code ought to be divided), and not only considerations of essence (concerning the halakhic character of the commandments under discussion).

The conclusion is that if for Ibn Ezra the roots sketch the skeleton of Jewish law and the commandments, for Maimonides they sketch the skeleton of the Jewish law book. Therefore, for Ibn Ezra the roots express theoretical halakhic and meta-halakhic conceptions, whereas for Maimonides there are also roots of classification and categorization dealing with the systematic presentation of the halakhic structure.[123]

Let us now see several consequences of this difference.

• Ben-Menahem[124] notes that Maimonides adheres in his enumeration to the conception of the Sages, whereas Ibn Ezra sees the enumeration of the commandments as a theoretical conceptual matter free of the constraints of tradition. Each will set up this structure according to his own understanding and on the skeleton he sees in it. This may be a consequence of the different aims. Maimonides wants to summarize binding Jewish law, and therefore he adheres to the way of the Sages, whereas Ibn Ezra sees the matter as purely theoretical or intellectual.
• Ben-Menahem[125] notes that Ibn Ezra has many filtering roots, whereas in Maimonides the emphasis is on roots of individuation. This difference too reflects the different aims, since the roots of individuation concern method and classification, not necessarily halakhic content (as noted above, individuation for the enumeration of the commandments does not overlap with halakhic individuation). Ibn Ezra is concerned primarily with the question what is and is not a commandment, and that is what the filtering roots address.
• In the paragraph concluding his article (‘The Secret’), Ben-Menahem writes as follows:

‘Maimonides’ model for enumerating the commandments relates only to commandments by Torah law. In this model every normative proposition will appear either as a commandment in its own right or as part of a commandment. Such a model can be called “reductionist,” because every Torah normative proposition can be reduced to its source.

‘By contrast, Ibn Ezra’s model relates to the entirety of the normative propositions of Jewish law that belong to the first plane [the obligatory commandments, whose concern is objective and universal for every person and every time]. Propositions belonging to the second plane [commandments left to the decision of the individual, or representing the law’s accommodation to human weakness] will not appear. This model can be called “essentialist,” because it identifies the significant propositions of Jewish law.’

Here again we have an expression of the different aims of the enumeration of the commandments. Maimonides relates to all Torah-level commandments, because his purpose is to ensure that in his Mishneh Torah not even one of them, from any level of Jewish law, will be omitted. Hence he establishes a reductionist model, since his aim is classification and generalization, and the construction of a complete and systematic structure. Ibn Ezra, by contrast, chooses the principal ones among them (only those on the first plane), because his interest is purely intellectual and theoretical. He therefore establishes an essentialist model, not necessarily a systematic one. This is precisely what I said: Maimonides acts in order to set up a systematic structure, whereas Ibn Ezra sets up a substantive one.

Back to the definitional commandments

From this we can also understand the appearance in Maimonides’ count of commandments that are not commands (such as impurities and annulment of vows). As noted, in his enumeration of the commandments Maimonides is building the skeleton of the Mishneh Torah. In a comprehensive law code there is no doubt that one must separately include a body of laws dealing with annulment of vows or with the definition of the various states of impurity. True, these are not ‘commandments’ in the accepted normative sense, but there is no doubt that they are important segments of Jewish law, and in the skeleton of any halakhic law book they must be included. Certainly all of these are included among the ‘organizing categories’ of the law code, even if they are not commandments in the accepted sense of the word.

One may assume that Ibn Ezra would not have included such commandments in his count, for they are not commandments at all but definitions. This is a prominent consequence of the different aims of these two sages in their enumerations of the commandments.

Additional expressions of the systematic tendency in Maimonides’ enumeration

Traces of Maimonides’ systematic tendency can be seen in a number of places in his enumeration of the commandments and in his words in the roots. I shall note here a few further examples.

Examination of the roots of classification and categorization yields a fairly consistent picture of different methods among the early enumerators of the commandments concerning the ways to classify broad groups of commandments.[126] For example, the medieval authorities dispute how the commandments of sacrificial offering are to be counted. Maimonides counts the offering procedure and the obligation to bring a sacrifice, for each general type of sacrifice (there are five general types), as two commandments (for a total of ten such commandments). By contrast, Bahag does not count the obligation to offer at all, and instead details the obligation to offer into its parts, which he counts as many commandments. Nachmanides, unlike both of them, argues that there is only one general obligation upon the priests to perform the sacrificial service (and he does not address the owner’s obligation to bring the offering, because it is absorbed into the transgression for which the sacrifice is brought). Finally, Saadia Gaon counts every type of sacrifice appearing in the Torah (not only five general types, as Maimonides does).

We find a very similar pattern in other halakhic contexts as well. For example, in the fourteenth root a similar picture appears regarding the counting of punishments. Maimonides counts each general type of punishment as a commandment in its own right (lashes, stoning, burning, beheading, strangulation, and so on). By contrast, Nachmanides argues that there is a general obligation imposed on the court to punish. Bahag holds that punishments should not be counted at all, since they are included in the transgressions for which they are imposed. Saadia Gaon argues that every death penalty for every offense should be counted as a commandment in its own right (seventy-one capital punishments).

The same is true of impurities. Maimonides counts thirteen commandments of different kinds of impurity (positive commandments 96–108). Nachmanides counts none of them (because he sees them as definitional commandments and therefore includes them within the commandments defined by states of impurity, such as the prohibition of entering the Temple in impurity or eating sacrificial food in impurity, and so on). Saadia Gaon counts forty-four such commandments (positive commandments 157–200).

A final example is the counting of types of leprosy. Maimonides counts three such commandments (human beings, garments, and houses; see positive commandments 101–103). Saadia Gaon counts twelve commandments (positive commandments 186–197). Generally speaking, Nachmanides tends toward generalization and does not provide such detail (his method regarding leprosy is not entirely clear).

To summarize: Saadia Gaon specifies every commandment that appears in the Torah; Nachmanides generalizes them all into one inclusive commandment; and Maimonides takes the middle path—he counts the principal types and includes the details within them.[127]

What is the basis for these three methods? It is hard to believe that such a consistent picture is accidental. Nachmanides indeed remarks against Maimonides that he cannot find the logic in his method: if he deals with every halakhic commandment, he should specify them all like Saadia Gaon; and if he deals only with the general idea, then he should generalize like Nachmanides himself. Nachmanides assumes that the classification of the commandments should be done on a halakhic or meta-halakhic basis.

But in light of what I have said here, we can understand that Maimonides regards the enumeration of the commandments as a skeleton for his halakhic book, and therefore he counts the commandments also according to systematic considerations and not only substantive-halakhic ones. The commandments are the types, or headings, that must be elaborated in the Mishneh Torah. Therefore Maimonides’ considerations of individuation are systematic and methodological (what is the most efficient way to present these bodies of law in the complete halakhic structure), not necessarily halakhic. He asks what can serve as a heading for the sets of laws in the Mishneh Torah. Nachmanides and Saadia Gaon, by contrast, work from essentialist considerations. They examine the matters from halakhic (= Saadia Gaon) or meta-halakhic (= Nachmanides) perspectives. For them the enumeration of the commandments is merely a theoretical-intellectual expression of their conception of Jewish law, as we saw with Ibn Ezra.

Thus we see that these methods of classification also express the different approaches and aims of these sages with regard to the enumeration of the commandments.

In my article on the twelfth root I pointed out that Maimonides’ classification is systematic in character, not merely substantive. I showed there that he uses criteria of efficiency, not only of essence. Hence, for example, Maimonides does not count the parts of the sacrificial procedure, but only the kinds of sacrifices. The reason is that in a law code there is no place for ‘laws of taking the handful’ or ‘laws of burning on the altar.’ These will appear within the laws concerning the bringing of the sacrifice. Therefore, if these are not headings in the Mishneh Torah, they also will not appear as independent commandments in the enumeration of the commandments.

There are, however, exceptions to this conception of Maimonides. In his remarks on root 12 he himself presents several of them:

This sequence of acts constitutes one commandment, and it is not proper to count each and every component of the labor as a separate commandment—unless there are general commands applying to all types of offerings and not restricted to one particular type, in which case each such command should indeed be counted as a separate commandment, since it is not then merely a component of the sacrificial procedure of a particular offering.

For example, His warning, exalted be He, against offering a blemished animal (prohibitions 91–95), or His command that it be unblemished, and His command that it not be lacking the required age, as it says, “From the eighth day onward,” and His command that every offering be salted, as it says, “With all your offerings you shall offer salt” (positive commandments 60–62), and His warning not to omit it, as it says, “You shall not let the salt cease” (there, prohibition 99), and His command to eat that which is to be eaten from it (positive commandments 88–89). Each one of these commands is a separate commandment, for none of them is merely one of the component parts of the commandment involved in the sacrificial procedure of any particular offering; rather, they are general commands applying to every offering, as we will explain when we count them.

Thus he does count the commandment to salt the sacrifices, or the prohibition against offering a blemished animal. Why are these not included as details within the sacrificial procedure itself? Maimonides explains that they relate to several different commandments, not only one, and therefore must be established as separate commandments.[128]

It is clear that this is not a substantive consideration. Maimonides does not explain that salting a sacrifice is indeed an independent commandment, unlike sprinkling the blood or taking the handful in meal-offerings. The explanation he offers is a distinctly systematic consideration. An efficient law code does not define concepts anew every time it needs them; rather, it devotes a special place to defining the concepts and procedures it discusses. For example, the definition of the various states of impurity will not be given separately in the laws of entering the Temple and again in the laws of eating sacrificial food. Instead, we devote a special place to defining the states of impurity. Then, whenever we need one of the commandments or halakhic contexts relating to impurity, we will consult the defining sections and learn from them whether the person or object is impure or not.

Now slaughtering, sprinkling, and the like, which also appear in several procedures, might at first glance also have deserved separate counting. But closer inspection shows that these do not require independent elaboration, because they are simple stages that can be defined and detailed on the spot, within the laws of offering the sacrifices. Therefore they do not merit a separate section in the law code, and consequently they are not counted separately in the enumeration of the commandments either.

Thus the consideration that appears here is one of efficient presentation of these laws in the law code. This requires that they not be re-elaborated anew in every context to which they pertain, but rather defined once, in one place. This is a clear expression of Maimonides’ systematic tendency, which departs from purely substantive considerations (halakhic or meta-halakhic). For him, the enumeration of the commandments is constructed from within an aim of creating the skeleton of a law code, not only out of substantive halakhic or meta-halakhic considerations.

The commandments of belief[129]

Another controversial example that can be adduced to illustrate the systematic principle is the commandment of belief (positive commandment 1). Many commentators have objected to the place of such a ‘commandment’ in the enumeration of the commandments. How can one command this at all? At first glance this is a basic condition for a person being a fit addressee of commands (without belief one cannot command him; compare Maimonides’ formulation at the opening of his code: The foundation of foundations and the pillar of wisdoms is to know that there is a First Being…), but it is not itself a commandment. Moreover, what good does the command do? If a person does not believe, he will not believe after the command either.

In light of all that has been said so far, we may perhaps suggest that this commandment too appears in the enumeration of the commandments for systematic reasons, not substantive ones. True, it is not a commandment in the ordinary sense, but belief certainly must have a place in the law code (the Mishneh Torah) and be elaborated there, and therefore it must be treated as an organizing category. Moreover, it is fitting that a law code open with a declaration of its essential nature and aim, and therefore belief should stand at the head of the enumeration of the commandments. Not because it is a commandment in the accepted sense, for it commands us nothing, but because the enumeration of the commandments is nothing other than the skeleton of the law code.

Difficulties with the model I have presented[130]

There are several further examples expressing Maimonides’ systematic tendency in his enumeration of the commandments. As we have seen, the model presented here does indeed receive support from several sources in Maimonides’ own writings and resolves several difficulties that commentators found puzzling. For the sake of honesty, however, it must be added that this picture is not free of difficulties. I shall mention several of them:

1. If the aim is really merely organizational and systematic, why argue over the inclusion or exclusion of this or that commandment from the count? At first glance, the whole polemic with Bahag is unnecessary. Let each person choose the method of organization that seems right to him and determine accordingly the counting rules he uses. Seemingly, these disputes indicate that Maimonides does not see his enumeration of the commandments differently from his interlocutors. It seems, rather, that he too sees it as an expression of the character of Jewish law, not merely as a systematic skeleton for a law code.

One might perhaps explain this by saying that Maimonides’ main polemic is conducted against Bahag, and, as we saw, Bahag’s aim was similar to Maimonides’ own, and therefore there is room for polemic and dispute between them. But this is strained, because Maimonides directs his words toward all enumerators of the commandments; his focus on Bahag stemmed from Bahag’s status among the enumerators who preceded Maimonides. Moreover, I already noted that Gutman pointed out that even with Bahag the enumeration of the commandments was probably not created out of a methodological aim, despite its location at the opening of his book.

2. It is not clear why Maimonides needs R. Simlai’s statement at the end of tractate Makkot. In the simple sense, R. Simlai counts the commandments on the substantive plane, for he was certainly not engaged in constructing any law code. If so, when R. Simlai says there are 613 commandments, he does not mean that the skeleton of the halakhic whole contains 613 parts and headings, but that from a halakhic-substantive standpoint there are 613 distinct commandments.

In other words: if Maimonides counts the commandments for methodological-systematic reasons, there is no reason in the world that his count should coincide with the substantive count (= 613 commandments). Why assume that the number of headings in the law book coincides with the number of commandments counted for substantive reasons? In other words: why assume that halakhic individuation coincides with systematic individuation?

Moreover, Maimonides attacks Bahag through precise readings of R. Simlai’s statement and of other midrashim. For example, at the beginning of the first root he attacks Bahag, who counts rabbinic commandments, from the language of R. Simlai’s midrash, which speaks of commandments that were given to Moses at Sinai. But if indeed the issue is a systematic consideration, what relevance does R. Simlai’s statement have here? R. Simlai is concerned with the number of commandments given at Sinai, whereas Maimonides is looking for the number of headings for his book. The same is true at the beginning of the third root, where he proves from the language of R. Simlai’s statement that the count concerns only commandments given to us for future generations (an inheritance, in the sense of inheritance). He adds there a proof from the language of the midrash that temporary commandments are not to be counted:

Their statement as well, that each and every limb is as though it commands a person to perform a commandment, and each and every day is as though it warns him against transgression, is proof that this number can never be lacking. For if the count included commandments that do not apply for all generations, then this principle would be lacking at the time when the obligation of that commandment ceased, and that statement would be complete only for a limited time.

This is seemingly a clearly substantive midrash. It explains that there are as many commandments as there are limbs of the body, and each commandment corresponds to one of them. According to Maimonides, that may perhaps be the substantive number (that is, the number of commandments that differ from one another at the substantive-halakhic level), but the number produced by systematic considerations (the division into headings and subheadings of the halakhic corpus) could be completely different.
To sharpen this, let us ask it this way: Maimonides counts the commandment of the law of impurity from a corpse, impurity from a creeping thing, and so on. We have seen that someone who counts commandments for substantive reasons would not include these in his count. If so, presumably he would include other commandments in their place. The same applies to sacrificial procedures and the other definitional commandments discussed above.

3. As we saw, many understand Maimonides as drawing some of his rules from the words of the Sages, on the assumption that the Sages themselves were engaged in enumerating the commandments (contrary to Nachmanides). But if his enumeration of the commandments is nothing other than a collection of headings for his halakhic code, there is no reason to assume that proofs from the language of the Sages are relevant. The Sages certainly were not engaged in codifying Jewish law, and therefore their count certainly was not intended to be a skeleton for any halakhic codex. This too suggests that Maimonides intended a substantive enumeration of commandments, not a systematic one.^[131]
4. As we have seen, some of Maimonides’ roots do indeed deal with essence. This is true above all of the first category, which excludes from the discussion rabbinic commandments and commandments derived by exegesis. Dividing halakhic material into headings ought to have led to the inclusion of these laws in the counted material. The very fact that lashes serve as a criterion for the enumeration of the commandments (see the note above) points to substantive-halakhic considerations, not only systematic ones.^[132]

If we generalize these difficulties, we may put it thus: if Maimonides’ purpose was indeed only systematic, that is, if he was merely seeking subheadings for his comprehensive halakhic work, then all his considerations should have been systematic. There would have been no room at all for substantive-halakhic considerations in his rules for counting the commandments, nor for the polemics he conducted with his predecessors. This would seem to cast doubt on the systematic thesis I have proposed. On the other hand, as we have seen, there is substantial evidence for that thesis.

A combined model: a substantive-systematic structure

The totality of these considerations indicates that Maimonides did not mean to base his whole system of commandments on systematic considerations alone. Considerations of essence must also occupy an important place in constructing his system of commandments. His structure is therefore composed of both kinds of considerations together. As I suggested above, Maimonides sees the 613 Torah-level commandments as headings or branches in the halakhic skeleton.

From this it appears that we must soften the dichotomy between Maimonides and Ibn Ezra. Above I presented a picture of two different, even opposed, tendencies in the enumeration of the commandments. We now see that Maimonides’ aim was indeed to build a skeleton for his halakhic work, but he did so on the basis of substantive considerations, as was customary regarding other enumerations of the commandments. He counted the commandments on the substantive plane, and only afterward used that count as a systematic skeleton for his work. It is possible that the reason is that he wanted to begin his enterprise with an enumeration of commandments in accordance with accepted tradition (for all the reasons mentioned above in the methods of the other medieval authorities), and in addition wanted to build his book upon it. Thus the motivation was indeed systematic—to create a skeleton for his great halakhic work—but the path was also substantive.

At least three reasons may be suggested for why Maimonides chose specifically the substantive enumeration as the systematic skeleton for his great work:

• He wanted to rely on R. Simlai’s count and use it as a methodological tool to set a framework for his own count, and through the count for his halakhic creation as a whole. As we have seen, the total number of commandments is a hint that it is very hard to build an enumeration without it.
• He wanted to create an enumeration of commandments for the same substantive reasons that other sages desired one (as part of the commandment of Torah study, or as a pedagogical tool, or for purely theoretical reasons). Moreover, he wanted to make use of the enumerations of his predecessors, at least where he agreed with them.
• He wanted to rely on the words of the Sages in the process of creating his skeleton, and they were certainly dealing with the substantive count, not the systematic one.

In any event, it is now understandable why Maimonides assumes that his count ought to coincide with that of R. Simlai, while also being compared with the counts of Bahag and the other substantive enumerators of the commandments, despite his systematic aim.

The picture we arrive at is the following: a commandment that meets the criteria of halakhic-substantive individuation will also be included separately in the enumeration of the commandments. But the enumeration of the commandments has additional criteria, systematic in nature, that serve in constructing it. This can also be seen through Ben-Menahem’s division between filtering rules and individuation rules. Whoever does not pass the halakhic filter certainly cannot be counted as a commandment in the enumeration. After the filtering, one must perform systematic individuation, and this does not necessarily overlap with halakhic individuation.[133]

A possible implication: the commandment of repentance[134]

One implication of this combined model is the following. There are several contradictions between Sefer HaMitzvot and the lists of commandments that open the halakhic sections in the Mishneh Torah.[135] Some of them may be attributed to Maimonides’ having changed his mind, since the Mishneh Torah was composed several years later (and it is well known that Maimonides often changed his mind in the course of his vast codificatory labor). But the picture presented here opens before us another possibility as well.

Take as an example the contradiction noted by the author of Minhat Chinukh in commandment 364. On the one hand, in Sefer HaMitzvot, positive commandment 73, Maimonides defines the commandment to confess as follows:

The 73rd commandment is that we are commanded to confess the sins and iniquities we have committed before God, exalted be He, and to state them together with repentance.

From his words here it appears that there is no commandment to repent. The commandment is that when one repents, one must confess. This commandment defines the process of repentance, but does not command us to repent. By contrast, in the list of commandments that appears before the Laws of Repentance, Maimonides writes:

One positive commandment: that the sinner return from his sin before the Lord and confess.

Here it appears that there is a commandment both to repent and to confess. True, these two components are counted as one commandment, but that is only because they are two parts of one commandment (according to root 11 or 12). From this formulation it clearly emerges that each of these two parts is a full halakhic obligation, unlike what emerges from Sefer HaMitzvot.

If Sefer HaMitzvot indeed served only as a skeleton for the Mishneh Torah, then contradictions between them would not be expected. But according to the combined model I have proposed, a more complex picture arises. On the one hand, Maimonides attempts to enumerate commandments, like all the other enumerators of the commandments. This count is substantive, not merely systematic. On the other hand, he wants to base his great halakhic work upon this enumeration. Therefore only commandments explicitly commanded in the Torah will appear in the enumeration of the commandments, for that is the basic rule of the count (as we saw from Maimonides’ words in both parts of the ninth root and in the second root). In his view, the commandment of repentance does not appear as a command in the Torah, and therefore it is not counted in the enumeration of the commandments.[136] But in the Mishneh Torah all our halakhic obligations must appear, both those explicitly written and those that are not (such as those learned by exegesis). Repentance is certainly a full halakhic obligation (its basis is reason rather than command),[137] and therefore it appears in the Mishneh Torah as a commandment.[138]

This is one possible mechanism for resolving contradictions between Sefer HaMitzvot and the Mishneh Torah without resorting to the claim that Maimonides changed his mind.[139] A similar mechanism can be found in several works of Rabbi Prof. B. Z. Benedict, who showed that in Sefer HaMitzvot Maimonides tended to follow tannaitic midrashim, whereas in his legal rulings he followed the methods of the Talmuds. Usually this concerns only the source of the law in question, but in several places it is also reflected in the practical halakhic ruling.[140]

This seems to fit well with the mixed tendency identified here, since this phenomenon indicates that Sefer HaMitzvot is not constructed only as a skeleton for the Mishneh Torah, and therefore meta-halakhic considerations relating to the enumeration of the commandments for other reasons also enter into it.

Against this background one may perhaps better understand the words of R. Judah ibn Tibbon in his introduction, who wrote concerning the composition of Sefer HaMitzvot in Arabic: I composed this work in the language of the Mishnah so that it would not be joined to my larger work, even though it serves as an introduction to it. We can now see that Maimonides truly did not want us to draw contradictory conclusions between the Mishneh Torah and Sefer HaMitzvot, since at least some of those contradictions stem from the different aims of the two works.[141]

A note and a puzzle: the definitional commandments

To conclude, I note that the subject of the definitional commandments still requires clarification in light of the combined model proposed here. At first glance these are not commandments—that is, they do not meet the halakhic criteria for the definition of the concept ‘commandment.’ If so, they should seemingly have had no separate place even in the enumeration of the commandments, since that is built upon the normative-halakhic plane.

From this it emerges that Maimonides sees a connection (almost a ‘mystical’ one) between the systematic plane and the substantive plane, and therefore systematic considerations that lead him to define and single out the definitional commandments indicate that these are also commandments in some halakhic sense.[142] This issue still requires further study.

Summary and a note in honor of the honored jubilarian

Throughout this discussion we have seen several aspects of Maimonides’ fourteen roots. We examined their principled significance, classified them into five types, and pointed to the assumptions embedded in them—or at least referred to sources dealing with those assumptions. In the course of the discussion questions arose concerning the definition of the term ‘commandment,’ the kinds of commandments, the relation between the enumeration of the commandments and halakhic classification as such, and more. In the end we are left with a question mark concerning the relation between systematicity and essence in Maimonides’ classification of the commandments, a relation that runs like a thread through all the roots. I shall conclude with a few words and a note in honor of the honored jubilarian.

Rabbi N. A. Rabinovitch, the honored jubilarian, has vast expertise in the study of Maimonides’ teaching in all its branches and nuances, and I, the least of scholars, am not worthy to recount his praise. In the introduction to his book Iyyunim BeMishnato HaHalakhit Shel HaRambam, he extends a very justified critique of the accepted ways of studying Maimonides’ writings. He writes that many raise contradictions and weigh Maimonides’ words, yet do not attend to the fundamental rules according to which those words are to be interpreted and reconciled—some laid down by Maimonides himself, and some that can be established through systematic study of his writings. Many local resolutions become superfluous once we understand that the answer to contradictions must be sought in Maimonides’ aims, style of writing, and governing rules.

Now the clearest example of such rules is the system of ‘the roots,’ which are the fundamental rules concerning the enumeration of the commandments, and from which much of Maimonides’ method in these matters proceeds and can be understood, as he himself wrote (see the quotations above at the beginning of section A). Astonishingly, in the monumental commentary of the honored jubilarian, Yad Peshutah, which deals with all the books and parts of the Mishneh Torah, including Sefer HaMitzvot, we do not find direct treatment of the work of the roots themselves. True, here and there in the rabbi’s writings there are uses of rules appearing in the roots,[143] but there is no direct treatment or commentary on the roots as such.

The reason for this lack is not clear to me (see also the introduction to the present article), and I hope that in this article I have filled the gap somewhat, in honor of him and of his comprehensive and upright Torah. I wish the honored jubilarian long life and years over his realm, to magnify Torah and glorify it, as has been his way from of old, with bodily vigor and radiant illumination. May it be God’s will that all of us continue to benefit from his light for many good days and years.

*       My thanks to Dr. Yuval Sinai and to the members of the editorial board, whose important comments greatly helped me improve this article.

[1]       It should also be noted that even where there are discussions (and we shall see some below), they concern mainly the first two roots. With respect to the second root, see the remarks of Neubauer in his book Maimonides on Divrei Sofrim, Mossad Harav Kook, Jerusalem 5717, which gathers the references in traditional literature as one gathers sheaves to the threshing floor. One gets the impression that they are not numerous (nor do they purport to be comprehensive and systematic as the subject deserves), and a significant portion of them does not belong to the main literature of commentary on Maimonides, certainly not to the central body of halakhic literature and responsa. This is particularly striking in light of the hundreds of questions and difficulties that arise with respect to his words in that root.

In any event, the first two roots are exceptional, and that is probably why discussion exists concerning them, both in research literature and in traditional literature (though its scope is far from expressing their importance). The reason is that in these two roots the principled methodological discussion is itself also a halakhic discussion. Below I shall say more about this unique feature of the first two roots and about the difference between them and the other roots.

[2]       Another prominent example of such reflective engagement is R. Abraham ibn Ezra in his book Yesod Mora, in the second gate (see also his long commentary to Exodus 20:1). See Hanina Ben-Menahem, ‘The Secret of Yesod Mora and the Secret of the Torah,’ Dinei Yisrael 22 (5763), p. 177 (hereinafter: Ben-Menahem, The Secret), especially ch. 5.

Ibn Ezra’s halakhic importance, however, is not comparable to that of Maimonides, and his work is not part of a systematic effort to create a halakhic architecture, as it is in Maimonides’ case (he does not even present his own list and enumeration of the commandments, although his intention was to engage polemically with earlier enumerators of the commandments). In addition, the rules do not appear with him as an explicit list, but can only be learned out of his words. Still, it is important to note that scholars have already pointed to the influence of Ibn Ezra’s Yesod Mora upon Maimonides’ work (see the sources for this in section A of Ben-Menahem’s aforementioned article). The basic concepts and analytical principles that Ben-Menahem uses (see below) are described in his article ‘Individuation of Laws and Maimonides’ Sefer HaMitzvot,’ Shnaton HaMishpat HaIvri 14–15 (5748–5749), p. 95 (hereinafter: Ben-Menahem, Individuation).

[3]       Henshke has already remarked on this in the concluding paragraph of his article ‘Remnants of Sefer HaMitzvot in the Mishneh Torah,’ Proceedings of the Tenth World Congress of Jewish Studies, C1 (5750), pp. 180–186.

I shall mention several major sources that deal with Sefer HaMitzvot (beyond the commentators printed around it):

For the role of Sefer HaMitzvot as a whole within Maimonides’ overall oeuvre, see a short popular survey in Rabbi Y. L. HaKohen Maimon’s book Rabbi Moshe ben Maimon, Jerusalem 5720, ch. 4. See also the excellent work of Yehiel Michel Gutman, Behinat HaMitzvot Lefi Minyanan, Siduran VeHitḥalkutan, Breslau 5688 (he also discusses enumerations of the commandments and their significance generally, and the sources for the various counts). See also Zvi Karl, ‘Maimonides’ Sefer HaMitzvot,’ Moznayim, vol. 3, issues 1–6 (5695), pp. 462–471 (he also discusses the roots there, but it is a rather superficial article, without major innovations). Also Benzion Bokser, ‘Sefer HaMitzvot and the Mishneh Torah,’ Batzaron, Heshvan 5709, issue 2, pp. 85–96. See also the important (though little known) book by Azriel Ziment, Mitzvot HaMelekh, Gross Brothers, USA 5759 (especially the introduction). More recently there appeared David Henshke’s article, ‘On Maimonides’ Halakhic Thinking: Between Internal Dynamism and Institutional Conservatism—On the Nature of the Law Embedded in Sefer HaMitzvot,’ in: HaRambam—Shamranut, Mekoriyut, Mahapekhanut (ed. A. Ravitzky), Jerusalem 5769, pp. 119–154. See also the entry ‘Taryag Mitzvot’ on Wikipedia.

Regarding the internal arrangement of Sefer HaMitzvot, which it is agreed is based in one way or another on the reasons for the commandments, see Rav Tza’ir, ‘Maimonides’ Sefer HaMitzvot,’ Batzaron 7 (5703), pp. 346–354; and also Bokser cited above. See also Rabbi Menachem Schlanger, Amudei Shelomo, Jerusalem 5749 (he also draws halakhic conclusions from the order in which the commandments appear in the book). See also a note in the introduction to Mitzvot HaMelekh, end of ch. 5, sec. 4. See also a comprehensive survey in Rabbi Alter Hilvitz’s article, ‘The Order of the Commandments in Maimonides’ Enumeration,’ Sinai 19 (Nisan–Elul 5706), pp. 258–267. See also Yad Peshutah (by the honored jubilarian), introduction and enumeration of the commandments, pp. 57–68.

Comparisons among the various enumerations of the commandments may be found in Sefer HaMitzvah VeHaMikra by Rabbi A. Z. HaLevi Rabinovitz, Jerusalem 5748; and in Yaakov Levinger’s book HaRambam KePosek VeKefilosof, Jerusalem 5750 (where he also discusses at length Maimonides’ reasons for the commandments).

Over the generations several commentaries on Sefer HaMitzvot have been written. I shall mention several by contemporaries who are aware of methodological questions and also address the roots and Maimonides’ other writings, and who wrote notes and commentaries for various editions of Sefer HaMitzvot: notes in Rabbi Kapach’s edition (2nd printing, Jerusalem 5754); notes and introductions by Rabbi Nahum Eliezer Rabinovitch (the honored jubilarian), in his commentary Yad Peshutah to the introduction and enumeration of the commandments (Jerusalem 5757); and local notes by R. Hayyim Heller to his edition (3rd printing, Jerusalem 5755). See also Abraham Alter Faintuch, Peirush Pekudei Yesharim, Jerusalem 5760; Rabbi Yehoshua Weitzman, Mitzvah Berah, Ramat HaGolan 5748. For commentary and notes on Nachmanides’ glosses, see the notes of Rabbi H. D. Chavel in his edition (8th printing, Jerusalem 5754; it should be noted that there he does not address Maimonides’ own words).

Regarding the various editions and translations of Sefer HaMitzvot, see the detailed note below.

[4] In this study I do not deal with questions of textual variants. The quotations are taken from the edition that appears in the Responsa Project, based on the Shabtai Frankel edition. As far as I have checked, textual differences in other editions do not alter the general picture or the conclusions presented here.

   [5]    A reading of the article shows that there are not a few references to my own writings on the various subjects, which may seem troubling. This stems both from the significant lack of work on the roots generally, from my own intensive concentration on this subject, and from the approach I wished to present here, which naturally reflects my own point of view.

[6]       These articles are slated, God willing, to appear as a book in the Middah Tovah series.

[7]       Whenever I refer, without further qualification, to the article on a given root, I always mean an article from that series. The series includes sixteen articles: the first is an introductory article, followed by fourteen articles on the various roots (in the order defined below), and finally an appendix article on the enumeration of the commandments of the Noahides (following Nachmanides’ remark at the end of his glosses to the roots).

[8]       See on this issue Moses Halbertal, ‘Maimonides’ Sefer HaMitzvot—The Architecture of Jewish Law and Its Theory of Interpretation,’ Tarbiz 59 (5750), pp. 457–480. One can also consult any of the introductory books to Maimonides’ teaching, especially Isadore Twersky, Introduction to the Code of Maimonides, Jerusalem 5751. A good description and analysis may also be found in Gutman’s aforementioned book. See also the introductory article, the first in the Middah Tovah series on the roots, and Bokser’s aforementioned article (see n. 4), as well as Henshke’s article cited above, ‘On Maimonides’ Halakhic Thinking.’

[9]       Maimonides, and after him also the Shulchan Arukh, came under sharp criticism for this innovation—presenting the halakhic whole as a codex—from various angles. See Menachem Elon’s HaMishpat HaIvri, vol. 2, in the chapters dealing with the codification controversies, and the sources cited there.

On Maimonides’ aims and tendencies in his great work, see the article of the honored jubilarian, ‘Maimonides’ Way as Decisor and Commentator,’ in: Sefer HaZikaron LeRav Yitzhak Nissim, Ḥikrei Talmud HaRambam (ed. Meir Benayahu), Jerusalem 5745, pp. 163–173. The article also appears in the collection of his articles, Iyyunim BeMishnato Shel HaRambam, Ma’aliyot, Jerusalem 5759, pp. 9–66.

[10]     In the introduction to his Commentary on the Mishnah, Maimonides writes: I composed commentaries on three orders: Mo’ed, Nashim, and Nezikin, aside from four tractates on which I am presently trying to write something, though I have not yet found the time for it. I also explained Hullin because of the great need for it…. He means a commentary on the Talmud, since his Commentary on the Mishnah is mentioned immediately afterward and was written on all six orders. On Maimonides’ Talmud commentary, see Israel Ta-Shma, ‘Maimonides’ Commentary on the Talmud—A Riddle and the Way to Solve It,’ Shnaton HaMishpat HaIvri 14–15 (5748–5749), p. 299.

[11]     See Twersky on this at the beginning of chapter 4. Further there he discusses the transition from the enumeration of the commandments to the Mishneh Torah, and the implications for the structure of Maimonides’ great composition.

[12]     The most striking difficulty is the one I shall present at the end of the article concerning the relation between substantive considerations and systematic considerations in the enumeration of the commandments.

[13]     For the definition of this classification, and the distinction between it and philosophical classification, see Twersky, pp. 225–229.

[14]     The categorical division of the commandments into the various books is described at length in part III of the Guide of the Perplexed. It is discussed in detail by Twersky, pp. 198–240. On differences between the classification in the Guide and that in the Mishneh Torah, see Twersky, pp. 239–240. See also Yad Peshutah, introduction and enumeration of the commandments, p. 187 and onward.

[15]     With regard to the identity of the author of Halakhot Gedolot, opinions already differ among the medieval authorities. See briefly the introduction to Mitzvot HaMelekh, ch. 3 n. 5, and the introductions to Halakhot Gedolot in the Machon Yerushalayim edition.

[16]     With minor deviations. See the introduction to Mitzvot HaMelekh, ch. 3 n. 6. Gutman, however, in the first chapter of his book, challenges the chronological relation that Maimonides posits between the various enumerators of the commandments. His claim is that some of the other enumerations preceded Bahag and even served as his source. He further argues there (later in the chapter, and in the following chapters) that there probably existed an ancient tradition concerning the total number of commandments, which also included some specific enumeration of them, from which both the authors of the Azharot and Bahag drew, and which was the source of their counts (see also there, ch. 3, for rabbinic sources pointing in this direction). If this conjecture is correct, then Maimonides was consciously or unconsciously also setting himself against that earlier tradition.

[17]     It should be noted that Saadia Gaon also composed Azharot, and R. Yerucham Fishel Perla’s famous work was composed around them. Beyond that, he also wrote his own Sefer HaMitzvot, which at the time of this writing was due to appear by Rabbi Ḥayyim Sabato. See this in R. Yerucham Perla’s introduction to Saadia Gaon’s Sefer HaMitzvot in his edition (see also Gutman, p. 13 n. 2 and thereabouts). This point strengthens the conclusion that Maimonides did not know these writings of Saadia Gaon, for otherwise it is unlikely that he would have included him among those liturgical poets whose learning Maimonides doubted (see R. Yerucham Perla’s introduction on Maimonides’ attitude toward Saadia Gaon, and Rabbi Sabato’s article in this volume on that very question).

Another indication may be found in the sixth root, where Maimonides writes that he knows of no enumerator of the commandments who disagreed with him that one should count duplicated positive commandments and prohibitions (such as the positive commandment and prohibition regarding labor on the Sabbath) as two separate commandments. As RYFP proves, Saadia Gaon indeed disagrees with this, from which it follows that Maimonides apparently did not know Saadia Gaon’s count or his Azharot.

It should also be noted that Ibn Ezra in Yesod Mora, similarly to Maimonides, compares these authors to enumerators of medicinal plants who do not know their properties.

Maimonides nevertheless testifies about himself (sometimes with regret) in several places that he was in fact influenced by them, especially by the book of R. Hefetz ben Yatzliaḥ (see sources in Twersky, p. 188 n. 20 and thereabouts). It should be noted that R. Hefetz was one of the most important Geonim, and it is difficult to assume that Maimonides would have spoken of him dismissively as though he were not a Torah scholar but merely a poet. The same holds for the Sefer HaMitzvot of R. Samuel ben Hofni Gaon (see A. A. Harkavy, ‘Part of Sefer HaMitzvot by R. Samuel ben Hofni Gaon,’ HaKedem 3 [5672], pp. 107–110).

Even regarding these figures, however, it appears that Maimonides criticizes them (to the extent that he knew them) at least for lack of care in the enumeration of the commandments and for preferring stylistic-artistic considerations over substantive-analytical ones (see Twersky, pp. 186–193).

On the earlier enumerations see also B. Z. Halper, in the introduction to his edition of the Sefer HaMitzvot of R. Hefetz ben Yatzliaḥ, Philadelphia 5675; and Gutman, cited above, in chapter 1; and A. Neubauer (in German), JQR(OS) 6 (1894), pp. 698–709.

[18]     See the introduction to Mitzvot HaMelekh, ch. 3, sub-sections 2–3, and Gutman in the second half of chapter 1, where they discuss the different divisions of the commandments. It should be noted that in chapter 3 of his book Gutman suggests that with Bahag the commandments are divided according to punishments and according to the persons obligated, and that, generally speaking, the idea of reward and punishment underlies Bahag’s enumeration of the commandments (he claims that some editions tried to obscure this; see p. 38).

[19]     See detailed discussion in Gutman, second half of chapter 1 (he also cites earlier sources for Bahag’s divisions).

It should be noted that Maimonides in the fourteenth root claims that Bahag counts punishments as independent commandments, and disputes him. Nachmanides, already in his glosses there, points out that Maimonides apparently erred in understanding Bahag’s opinion (or else had before him a different text of Bahag). Bahag does not count the punishments themselves as commandments; rather, he uses them as a classificatory tool for the various commandments (those punishable by karet, those by stoning, and so on).

[20]     In part 3 of Saadia Gaon’s Sefer HaMitzvot, and also in his introduction to part 1.

[21]     See also the appendix to the article on the tenth root.

[22]     For a list of the various categories of commandments, both in the literature of the Sages and in the literature of the commandments prior to Maimonides and after him, see HaMitzvah VeHaMikra, p. 31.

On Maimonides’ division between positive commandments and prohibitions as the basis for the enumeration of the commandments, see Gutman, chapter 4. He wonders whence it derives and suggests that it may be rooted in an Azharot tradition unknown to us that was based on the Ten Commandments (see there at the end of the book).

[23]     Apart from one place (and apart from the introduction), throughout his remarks in the roots Maimonides does not mention Bahag by name (perhaps out of respect, since his remarks are sometimes presented dismissively as foolish). The exception is the fourteenth root (and in the seventh root he mentions him in a side context). Perhaps the exception exists because this is the concluding root of the work, and Maimonides did not want to leave any doubt about whom his remarks were directed against.

[24]     Twersky, pp. 179–181, however, points out several omissions in the Mishneh Torah. Some are because the matters are self-evident, and for other reasons as well.

[25]     Gutman discusses whether systematicity indeed helped Maimonides’ project. See also the article in German (Mishorun 5675) by Maharash Katz, mentioned there on p. 27 n. 1, which questions the benefit to be found in Maimonides’ systematicity.

[26]     On the casuism of Jewish law, see at length Rabbi Yehuda Brandes, Reshitam Shel Klalei HaPesikah, doctoral dissertation, Hebrew University, Jerusalem 5763. On the principled and logical aspects of this question, see also Rabbi Eliyahu Rahamim Zeini, ‘Logic and Metaphysics in the Exegesis of the Sages,’ in Sefer Higgayon (eds. Moshe Koppel and Udi Merzbach), Bar-Ilan University and Yeshiva University, 5755, pp. 65–78. See also my book Shtei Agalot VeKadur floating, 2nd ed., Jerusalem 5767, gate 3.

[27]     See Twersky, pp. 241–250, for an extensive discussion with sources regarding Maimonides’ language considerations, mainly in the Mishneh Torah and generally.

[28]     See Maimonides’ responsa, Blau, nos. 355 and 447.

[29]     These ideas are perhaps hinted at also in the aforementioned responsum 447, where Maimonides writes that he regrets having written Sefer HaMitzvot in Arabic, since it is not accessible to all Israel.

A similar conjecture was raised by Benzion Bokser (above, n. 4), in his article cited above, p. 87. He, however, attributes the Arabic language to the fact that the work was intended for the common masses rather than for scholars (compare in this regard Rabbi Hilvitz’s remarks in his cited article), whereas my suggestion is that the work had temporary rather than eternal significance. Now see the editor’s note there, pointing out that the Guide of the Perplexed was also written in Arabic. That note refutes Bokser’s conjecture that Arabic was intended for the broad masses uninterested in philosophy. But on my approach it yields an interesting conclusion: the Guide of the Perplexed too was intended only to solve problems that arose in its own time and place, and was not meant for the whole Jewish people in future times and distant places.

[30]     See responsum 447, also printed in the Frankel edition of Sefer HaMitzvot, p. 9; and responsum 368 in the Jerusalem printed edition. This responsum will be mentioned again below.

[31]     Even so, quite a few commentators and decisors did not know Sefer HaMitzvot, and some attribute this to language difficulties. See, for example, the Raavad’s misunderstanding of the basis of Maimonides’ method at the opening of the Laws of Marriage. The principle Maimonides determines there (that betrothal by money is ‘of the words of the scribes’) is explicitly explained in light of Maimonides’ words in the second root, as the Kesef Mishneh notes ad loc., but the Raavad apparently did not know the roots. It is possible that this is precisely why Maimonides regretted using Arabic.

Several authors have already remarked on the language issue underlying this lack of familiarity. See, for example, Responsa Radbaz II, in the ‘Languages of Maimonides,’ no. 257, also cited in Yad Malakhi, Rules of Maimonides, sec. 51. See also Sedei Ḥemed, Ma’arekhet Ḥatan VeKallah, sec. 20. But Shem HaGedolim (entry ‘Maimonides’) already noted in their name that at least with respect to the author of Migdal Oz this claim is incorrect. [If you wish, we can add here an editorial note referring to the discussion of whether Migdal Oz and the Maggid Mishneh knew Arabic and Maimonides’ works that had not yet been interpreted in their time, in Prof. Jacob Spiegel’s article on Migdal Oz and the Maggid Mishneh, which will appear in the jubilee volume.] With great pleasure. Miki.

[32]     A prominent example here is the conjecture of several medieval authorities that the term ‘words of the scribes’ in the second root has a different meaning from the term ‘rabbinic.’ But anyone familiar with the Arabic text of Maimonides knows that there is no basis for this whatever. In Maimonides’ usage these terms clearly alternate. See the article on the second root.

See also R. Hayyim Heller’s introduction to his edition, p. 14 sec. 4, where he discusses the resolution of contradictions between Sefer HaMitzvot and the Mishneh Torah that stem from problems in the different translations.

[33]     The exact time of the composition of Sefer HaMitzvot, and certainly of the roots, is unknown, but the whole codificatory process took more than twenty years, and throughout it Maimonides changed his mind not a little (see also the next note and below). Yuval Sinai rightly noted to me, however, that in the Commentary on the Mishnah there are references to Sefer HaMitzvot—for example, Menahot 4:1 and Hullin 1:5. These were probably later versions, though the matter still requires clarification.

[34]     See, for example, Kin’at Soferim on root 10, where he points to a contradiction between Maimonides’ words in the roots and his words in Sefer HaMitzvot regarding making the anointing oil. See Rabbi Kapach, positive commandment 35 n. 4, and David Henshke, ‘Remnants of Sefer HaMitzvot in the Mishneh Torah,’ cited above in n. 4, especially the text around n. 19. See also Y. Sinai, ‘Apparent Internal Contradictions in Maimonides’ Writings,’ in: HaRambam—Shamranut, Mekoriyut, Mahapekhanut (ed. A. Ravitzky), Jerusalem 5769, p. 157 n. 10, and especially pp. 179–181 (between Mishneh Torah and Sefer HaMitzvot), and pp. 183–186 (between the Commentary on the Mishnah and the Mishneh Torah).

Already in Yad Malakhi, Rules of Maimonides, sec. 23, it was written that one should not infer that Maimonides changed his mind. Maimonides himself wrote in a responsum (Blau ed., no. 217) that one should not infer contradictions between the Commentary on the Mishnah and the Mishneh Torah, since at times he changed his mind after clarifying the issue, and the decisive text is what appears in the Mishneh Torah. We find this explicitly also in Ma’aseh Nissim, no. 7, s.v. however, where R. Abraham son of Maimonides pointed to a contradiction between Sefer HaMitzvot and the Mishneh Torah and wrote that Maimonides changed his mind, and that the law follows what he wrote in the Mishneh Torah. It should be noted that he himself relies on the responsum of Maimonides that I cited above concerning the Commentary on the Mishnah. See also Mitzvah Berah, the commandments of prayer, chapter ‘Facing the Temple.’ More on the contradictions between Sefer HaMitzvot and the Mishneh Torah appears below at the end of the article.

[35]     For a list of editions and translations of Sefer HaMitzvot, see a short survey in the introduction to Mitzvot HaMelekh, ch. 6, and also in R. Hayyim Heller’s introduction to his edition. In greater detail, see Israel Jacob Dienstag, ‘Ein HaMitzvot—A Bio-Bibliographic Lexicon of Researchers on Maimonides’ Sefer HaMitzvot and Its Commentators,’ Talpiyot 9 (5725–5730), pp. 663–759 (also published as a book, New York 5729); idem, ‘Maimonides’ Sefer HaMitzvot: A Bibliography of Editions, Translations, Explanations,’ Areshet 5 (5732), pp. 34–80.

On the problems inherent in translation itself and in the various translations, see the translators’ introductions, and also Moses Goshen-Gottstein, ‘On the Appearance of the Translation of Maimonides’ Sefer HaMitzvot,’ Sinai 44 (Tishrei–Adar II, 5719), pp. 185–187 (this article serves as an introduction to his popular translation published by Mossad Harav Kook).

Among commentators and scholars there is also a question which of these translations Nachmanides had before him when composing his glosses. See a detailed and comprehensive discussion in Hayyim Heller, ‘A Wonder Concerning Nachmanides,’ HaPardes 12, issue 1, sec. 1 (Nisan 5698), pp. 11–16.

[36]     See Rabbi Chavel’s words in the introduction to his edition of Sefer HaMitzvot. Nachmanides set himself the goal of defending halakhic authorities against attacks by later sages. This is also true of his book Milḥamot Hashem, whose main purpose was to defend Rif from Ba’al HaMaor’s objections.

[37]     In the Shabtai Frankel edition of Sefer HaMitzvot, the responsa dealing with the roots and Sefer HaMitzvot are printed around Maimonides’ own words.

[38]     It should be noted that RYFP possessed a highly developed research instinct, and his words are always written after comprehensive and meticulous examination of all the relevant sources. These features, like reflective engagement in Jewish law itself, are relatively rare in traditional rabbinic literature, and they are what give this exceptional and important work its uniqueness and quality.

[39]     See HaMitzvah VeHaMikra, by R. A. Z. Rabinovitz, Jerusalem 5748; see also the beginning of Pekudei Yesharim (above n. 4); and Rabbi Weitzman’s Mitzvah Berah (ibid.). In research literature there are treatments mainly of the first and second roots (as in Halbertal’s article in Tarbiz 59; in Yekutiel Yehudah Neubauer’s book cited above in n. 2, and the sources cited there; likewise in several articles by David Henshke: ‘Prohibitions for Which One Does Not Receive Lashes According to Maimonides,’ HaMa’ayan 24:2 [5744], pp. 33–40; ‘On Legal Reality in Maimonides’ Teaching,’ Sinai 92 [5743], pp. 228–239; ‘On Maimonides’ Distinction Between Torah-Level and Rabbinic,’ Sinai 102 [5748], pp. 205–212; ‘Secondary Rabbinic Prohibitions,’ Sinai 108 [5751], pp. 55–63; ‘On the Foundations of Maimonides’ Conception of Jewish Law,’ Shnaton HaMishpat HaIvri 20 [5755–5757], pp. 103–149, especially ch. 5). See also the article of the honored jubilarian, ‘On Rabbinic Laws Whose Force Is Torah-Level in Maimonides’ Teaching,’ Sinai 56 (5753), pp. 61–72 (also in the collection cited above in n. 10). See also Shlomo Kassirer and Shlomo Glicksberg, MiSinai LeLishkat HaGazit, Ramat Gan 5768. Also my articles: ‘The Logical Status of the Modes of Exegesis,’ Tzohar 12 (Tishrei 5763), pp. 9–21; and ‘Induction and Analogy in Jewish Law,’ Tzohar 15 (Summer 5763), pp. 23–34. See also my book, due to appear God willing in 5769 as the fourth volume in the quartet Shtei Agalot VeKadur Porēaḥ, and more. Treatments of the roots as a whole can be found mainly in Ben-Menahem’s aforementioned articles (n. 4), and also in Zvi Karl’s cited article.

[40]     Not a few medieval authorities, however, doubted and even disputed the absoluteness of this number, beginning with Nachmanides at the opening of his glosses to the roots, Tashbetz, Ralbag, and others. See a fascinating scholarly survey and discussion of this in Gutman’s book, especially chs. 2–4. He argues there that the number 613 is not the result of the various numerological calculations that lead to it; rather, such a tradition already existed, and those numerological calculations were created following it. In chapter 4 he also argues that Saadia Gaon did not know those numerological calculations and therefore labored to invent others (though it should be noted that Saadia Gaon’s way in several places is to propose alternatives to the Sages even when he did know their statements. For example, in his commentary to the baraita of the hermeneutical rules at the beginning of Sifra, he proposes different examples of the thirteen rules rather than using only those brought by the Sages).

On the total number of the commandments, see also Rabbi Barukh Brenner, ‘Ralbag’s Attitude Toward Maimonides’ Method in Counting the Commandments,’ Ma’aliyot 20 (5759), pp. 228–242 (also available on the view website). See also Mitzvot HaMelekh, introduction, and RYFP’s introduction to Saadia Gaon’s Sefer HaMitzvot. See also The Secret, around n. 24 (on Ibn Ezra and R. Judah ibn Balaam).

[41]     For a comparison of lists of commandments among the various enumerators, see HaMitzvah VeHaMikra throughout. In the second part of the book he even proposes his own list.

[42]     See, for example, a systematic comparison between the methods of the medieval authorities with respect to the roots in HaMitzvah VeHaMikra, from p. 17 onward. For the list of roots according to Ibn Ezra, see Ben-Menahem, The Secret. For Ralbag, see Asael Ben-Or, Sefer HaMitzvot Shel Ralbag Al Pi Peirusho LaTorah, pp. 6–9 (available on the ‘Keren Or’ website).

[43]     See, for example, the end of the fifth, eighth, and ninth roots; more extremely, the end of the twelfth and thirteenth roots; and the middle of the fourteenth root.

[44]     See Halbertal on this, and the article on the second root.

[45]     See the article on the first root. As can be understood from here, these first two roots provide excellent ground for examining the interpretive dispute between Maimonides and Nachmanides and its philosophical roots. This is the issue underlying the fourth book in the quartet Shtei Agalot VeKadur Porēaḥ, due to appear soon, God willing.

[46]     See, for example, the twelfth and fourteenth roots and others. In the thirteenth root it appears that Maimonides brought no proof for his words at all, but closer examination shows that there is there a proof from the language of Scripture. See the article on the thirteenth root.

[47]     See also Pekudei Yesharim (above n. 4), vol. 1, pp. 111–112.

[48]     Beyond his remarks in roots 5–11, which Nachmanides disputes, compare also Maimonides’ words in the fourteenth root, where he proves the enumeration of capital punishments imposed by the court from the language of the Mishnah, such as How is the procedure of those who are burned?, and the like.

[49]     Ben-Menahem (see Individuation), however, argues that the Sages too were aware of, and even engaged in, aspects of the enumeration of the commandments (‘individuation of laws,’ in his terminology). See also The Secret, p. 185, where he cites Ibn Ezra’s rejection of proof from the words of the Sages regarding the enumeration of the commandments.

[50]     See his book Hazal—Emunot VeDe’ot, Jerusalem 5743, p. 302. See, however, Ben-Menahem’s comments in Individuation, p. 99, and the fuller discussion in the article on the ninth root, where a detailed discussion is conducted concerning the significance of duplicated commands in the Torah. On Ibn Ezra’s method, see The Secret, ch. 4.

[51]     Thus, for example, Karl (above n. 4) understood the matter, and therefore concluded that Maimonides did not provide a truly systematic answer to the question which commandments are to be counted and which are not. The reason is that the words of the Sages caused him to deviate from the principled rules he himself established. But this is mistaken, as will be seen immediately. Karl also claims there that Maimonides does not distinguish between plain meaning and exegesis, and that his method is therefore inconsistent. This claim is puzzling, especially in light of the second root (which Karl himself also discusses). It seems that Karl himself does not correctly distinguish between plain meaning and exegesis, and that is the source of his error.

[52]     It is possible that one can see here an Aristotelian influence. For Aristotle, science is not determined through empirical experiment but through a priori reason (and therefore there is no sharp distinction between philosophy and science). For example, Aristotle determined that the speed at which bodies fall to the earth depends on their weight—something easily refuted by a simple experiment—and so on.

[53]     See, for example, the article on root 11, where I distinguish between various kinds of relations that can exist between parts of a commandment themselves and between them and the whole. According to that distinction, fringes and tefillin belong to different systematic categories. If so, the words of the Sages are not the essential reason for the exceptions, but only a sign of the existence of a substantive reason.

[54]     In the responsum of R. Yehoshua the Nagid (Ratzhabi ed., Jerusalem 5749, p. 34; see also the honored jubilarian in Yad Peshutah, introduction to the Laws of Repentance, Sefer Mada, vol. 2, p. 453), he reports that the number 14 was especially beloved to Maimonides. The explanation he offers is that this is the reduced numerical value of 248 and 365. Therefore it is no wonder that it is the number of books in the Mishneh Torah, the number of roots, and the number of hermeneutical rules enumerated at the beginning of the second root (13 plus extension). The honored jubilarian there (and also in Yad Peshutah, introduction and enumeration of the commandments, p. 68) discusses the place of the number 14 in Maimonides’ teaching generally. My thanks to the editor (Rabbi Haber) for the sources for this note.

[55]     Twersky, p. 180, claims that this is Maimonides’ policy in the Mishneh Torah as well.

[56]     Further detail on the roots may be found in my aforementioned articles on each of them. Here I present only the principal points for discussion and the essence of the conclusions with respect to each one.

[57]     See also my articles in Tzohar 12 and 15, and in Middah Tovah on parashiyot Yitro and Mishpatim, 5765.

[58]     In fact, this is the only halakhic consequence that appears explicitly in Maimonides’ writings for the principle laid down in this root, namely that commandments derived through exegesis have rabbinic force. It is still important to note that there is such a consequence, for it proves that Maimonides’ intention in this root is to formulate a principle that also has halakhic implications. The absence of an explicit command, such as when something is learned through exegesis, means that one does not punish for the prohibition so derived.

[59]     There I argued that Maimonides identifies the distinction between Written Torah and Oral Torah with the distinction between Torah-level and rabbinic. I also showed there from several sources in Maimonides’ writings that even a law transmitted as a law given to Moses at Sinai has the status of ‘words of the scribes,’ and again the reason is that there is no explicit command for it in the Torah. In that case, however, there is a command, only it was transmitted orally. In the article I show the halakhic and theoretical implications of the difference between a law given to Moses at Sinai and laws derived through exegesis, although Maimonides calls both categories ‘words of the scribes.’ The gist of the claim is that a law given to Moses at Sinai is a commandment for which there is a command but no objective content (repair or corruption, in reality or in the person), at least none meeting the threshold of Torah-level law. By contrast, laws derived through exegesis have content but no explicit command. We shall see this distinction again below.

[60]     It should be noted that Gutman, at the end of chapter 3, challenges Maimonides’ words in this root and raises lengthy objections to the distinctions that emerge from it. My article contains responses to all his objections (not mentioned there, because at that time I had not yet encountered his words).

[61]     And this is probably where Karl erred in his aforementioned article, where he took it for granted that this rule applies only to negations. There is an explicit passage in Sotah that relates to positive formulations as ‘negations’ (as Nachmanides cites in his glosses).

[62]     In the introduction to Mitzvot HaMelekh, ch. 1 n. 2, it was suggested that the commandment of settling the Land of Israel was not counted by Maimonides because it constitutes a preparatory act for the land-dependent commandments. From this the author argues that according to Maimonides too it is a Torah-level commandment, but one not to be counted by force of the tenth root. In my humble opinion this is mistaken, because Maimonides does count full commandments that serve as preparation.

[63]     This root is not identical with the seventh. The ‘parts’ here are not different applications of the commandment, but components that, every time one fulfills the commandment, must all be performed together.

[64]     See also my book Shtei Agalot VeKadur Porēaḥ, gate 2, and my article ‘On Commandments and Parts of Commandments—On the Philosophical Nature of Concepts in Jewish Law,’ Akdamot 21 (5768), pp. 160–175.

[65]     One classification is that of Hanina Ben-Menahem in his two aforementioned articles. The second is my own classification in the introductory article to the roots, which was written before Ben-Menahem’s works were brought to my attention.

[66]     In The Secret, Ben-Menahem notes that Ibn Ezra does not regard the rabbinic division of the commandments as binding tradition, since there is no one correct or incorrect division; everything depends on the aims of the various divisions. See also Gutman, p. 48, and below.

[67]     In the Mishneh Torah, of course, all the commandments appear, even those that are not counted. In the introduction to Mitzvot HaMelekh, chapter 1 n. 3, the author suggested that counted commandments appear in the Mishneh Torah in the language of ‘positive commandment,’ whereas uncounted commandments are described by Maimonides merely as ‘a commandment.’ He adduces several proofs. At the end of the note there, however, several counterexamples are brought, and the matter requires investigation.

I later saw that the honored jubilarian raises a similar suggestion in Yad Peshutah to Laws of Shema 1:3, and also in chapter 2 of his article ‘Commands, Obligations, and Goals in Maimonides’ Teaching,’ in Sefer HaYovel LiKhevod HaGrid Soloveitchik, Jerusalem 5744, pp. 599–612 (= the article also appears in the collection cited above in n. 2). See also below in the discussion of the commandment of repentance.

[68]     See the introductory article to the roots.

[69]     With respect to this root, however, there is room for discussion, because some interpret it as a principle of classification (similar to the seventh or eleventh root), from within the conception that a preparatory act for a commandment is itself a full obligation, and that the only reason it is not counted is that it is included in the commandment for which it prepares. I already noted this above in connection with the suggestion of the author of Mitzvot HaMelekh regarding the commandment to settle the Land of Israel.

[70]     See on this Yaakov Levinger, Darkhei HaMaḥshavah HaHalakhit Shel HaRambam, Jerusalem 5725, chapter 3 sec. C. There he distinguishes between ‘commandments’ and ‘their laws,’ referring mainly to the seventh root. He claims that this principle underlies all the roots from the tenth onward (see there p. 79 and n. 34), but this is not necessary. The tenth root deals with preparatory acts, which on their face are not parts of the commandment itself (though see the article on the tenth root). And in the fourteenth root he assumes that punishment is part of the commandment, which is not necessary. Certainly one cannot infer this from the order of the roots, since after the seventh root come also the eighth and ninth, which according to everyone are not included in this category. See there on p. 83 also applications to the first root (regarding rabbinic laws and ramifications of Torah laws; see my articles on the first two roots).

[71]     Here too, the reason for the separation I adopted was to clarify the treatment of the roots and examine the relations among them. As can be seen in the aforementioned articles, the roots of redundancy are discussed as a group, and each clarifies the others. The same applies to the roots of classification and categorization.

[72]     See also Gutman, chapter 3 (though not in a legal context, and not in this terminology). Also, in the broader context of classification and individuation, see Twersky, pp. 209–213.

[73]     And after him Joseph Raz. See Individuation, n. 1; and The Secret, n. 14.

[74]     See also my book Shtei Agalot VeKadur Porēaḥ, gate 2, where I discuss individuation of objects and of concepts.

[75]     Individuation, chapter 2.

[76]     The legal philosopher Ronald Dworkin criticizes meta-legal individuation precisely on this point. See Ben-Menahem, Individuation, n. 23.

[77]     See Ben-Menahem, Individuation, p. 99.

[78]     The Secret, p. 187.

[79]     It seems to me that the intention is only to the fifth and eighth roots. See below.

[80]     The fourteenth root, however, can also be included in this category, but that would require more detailed study of Maimonides’ words there. See the article on this root.

[81]     See at greater length in the article on the third root.

[82]     In The Secret, n. 38, an explicit division of all Maimonides’ roots according to the four categories proposed there appears.

[83]     See the beginning of the article on the eleventh root, where I pointed to the tautological and empty character of these roots, since if something is part of a commandment, then by definition it is included in the general commandment; and if we do not know it is part of a commandment, the rule Maimonides set down does not help us at all.

There I explained that Maimonides’ intention is to establish a determination with clear logical content, by no means empty. To understand this we must distinguish between two divisions of halakhic material: the halakhic division and the division of the enumeration of the commandments (to some extent this distinction parallels Ben-Menahem’s distinction between meta-legal individuation and legal individuation). These two divisions certainly do not overlap completely (Ben-Menahem, The Secret, p. 185, cites Ibn Ezra’s rejection of proofs from the words of the Sages regarding meta-legal division, while accepting their authority only regarding legal division).

For example, according to most views, Maimonides too sees laws derived by exegesis as laws whose halakhic force is Torah-level in every respect; yet they are not counted in his enumeration of the commandments. According to these views, we have here laws that are independent commandments halakhically, but parts of a commandment from the standpoint of the enumeration. Below I shall clarify this further when discussing the relation of the enumeration of the commandments to Scripture.

Another example is the prohibition of neshekh and tarbit, which in the enumeration of the commandments appears as one prohibition—prohibition 235—whereas in Maimonides’ laws he states that the lender violates two distinct prohibitions; see the beginning of ch. 4 of the Laws of Loans and Borrowing. See also the article on root 11.

We can now understand that in his words on the eleventh root, Maimonides means to say that if a certain act is part of a commandment in the halakhic sense, then it will be included in that commandment also from the standpoint of the enumeration of the commandments. It is now clear that this is not a tautology, since meta-legal division did not have to overlap with legal division. Maimonides’ innovation is that here there is overlap between these two divisions. Below I shall return to the distinction between them.

[84]     Ben-Menahem uses Maimonides’ formulation in the ninth root, that one should count the matters that are prohibited and the matters that are commanded, and infers from this that the principle divides the commandments according to states of action. Every state of action is counted as one commandment, even if several different commands are involved in it.

This definition, however, is not exact, for several reasons: (a) Sometimes the very same state of action is counted as two different prohibitions (and certainly halakhically it is considered as two different prohibitions; see the example of interest in the previous note). When one reads closely, one sees that Maimonides relates to states of action only as a measure of the number of lashes incurred, not as a measure of the number of prohibitions. (b) As RYFP himself notes in his comments on the sixth root, if the criterion were states of action, then even in the case of duplication between a prohibition and a positive commandment we should count only one commandment. For example, there is both a positive commandment and a prohibition regarding labor on the Sabbath. Their practical content is completely identical, and nevertheless Maimonides counts them as two different commandments. See my article on the sixth root. (c) As we shall see below, the division is a combination of different contents (not necessarily states of action) and the way the commands appear in the Torah (this emerges from Maimonides’ own words in the ninth root).

Maimonides’ intention in the ninth root is only to exclude counting repeated commands that command the same thing, not commands whose practical content is the same but which command different things (such as interest. In my article I argued that ‘tarbit’ is the prohibition against increasing the lender’s wealth in this way, whereas ‘neshekh’ is the prohibition against biting the borrower). Maimonides is not dealing here with different commands that happen to have the same practical content, but with identical commands in the conceptual sense (simple duplication). In any event, it is clear that even according to my explanation, this is a rule of individuation of the second kind.

[85]     Here too the thirteenth root is classified in the second category, which is not substantive. In substantive classification it belongs to the roots of redundancy. There would, however, be room to see it as a root of classification and categorization if we view the additional offerings of each day of a festival as parts of the general commandment of festival additional offerings rather than as a repetition of it each time. See the article on the thirteenth root, where I argued that there are temporal series of commandments of both kinds.

[86]     Here too I shall not address the proposal of the author of Mitzvah Berah in understanding the concept ‘commandment,’ for the reason noted above.

[87]     See at length Yaakov Levinger, Darkhei HaMaḥshavah HaHalakhit Shel HaRambam, chapter 3. The distinction between positive commandments and prohibitions is also discussed in Aharon Shemesh, ‘Toward a History of the Meaning of the Terms “Positive Commandment” and “Prohibitory Commandment,”’ Tarbiz 72, issues 1–2 (5763), pp. 133–149. See my critique of his remarks in the article on the sixth root.

[88]     This depends on the question whether coercion is not regarded as if one acted. See Jerusalem Talmud Gittin 7:6 and Kiddushin 3:2, where R. Yoḥanan and Resh Lakish dispute this (and some medieval authorities reverse the version on the basis of the Babylonian Talmud).

[89]     See more detail in the article on the ninth root.

[90]     See, however, the reservation in Maimonides’ responsum 447 (Blau ed., ‘Responsum to form on the question of the enumeration of the commandments’). See also David Henshke, ‘Prohibitions for Which One Does Not Receive Lashes According to Maimonides’ (above n. 40).

[91]     The source is the discussion in the Babylonian Talmud, Sanhedrin 63a.

[92]     Maimonides’ words here imply that this is Torah law, which requires investigation.

[93]     As already noted above, a commandment for which we have no command will not be counted, although it can still be a full obligation (some have suggested a linguistic distinction between the two categories, between ‘positive commandment’ and merely ‘commandment’). See also the article of the honored jubilarian, ‘Commands, Obligations, and Goals in Maimonides’ Teaching.’ See also his article ‘From the Mouth of Tradition, From the Mouth of Teaching,’ Sinai 58 (5755), pp. 26–41, and in the collection cited above in n. 2.

[94]     See the article on the ninth root. The basis of this distinction is set out in R. Elhanan Wasserman’s Kovetz Ma’amarim, in the essay ‘Repentance.’ See also the honored jubilarian’s article ‘Commands, Obligations, and Goals in Maimonides’ Teaching,’ where he gives many examples of obligations that are not commandments but consequences of aims and purposes to which the Torah points.

In the background of all this it is important to note that Maimonides consistently upholds the view that all commandments have reasons. Therefore, in his method it is obvious that every commandment has an aspect of content besides the aspect of command. See my articles on the fifth and ninth roots. Maimonides also explained the reasons for the commandments in detail in part III of the Guide of the Perplexed. For a survey of the reasons for all the commandments, and of the omission of reasons for about fifty commandments, see HaRambam KePosek VeKefilosof, chapter 4 devoted to this matter. See also Rabbi Y. L. Maimon’s remark in his book Rabbi Moshe ben Maimon, pp. 137–141, discussing whether those reasons reflect the rule of intellect or of feeling.

[95]     See, for example, the seventh root, where he speaks only about parts of a commandment that appear explicitly in the Torah; similarly in the tenth root regarding preparatory acts that are written in the Torah; and in the fifth root, where he speaks only about reasons written explicitly in the Torah.

[96]     Compare Mitzvah Berah at the beginning of part 1, where he discusses at length the dispute between Maimonides and the Geonim concerning the force of the obligation to keep commandments observed before the giving of the Torah. As he shows there, according to Maimonides the force depends solely on the command at Sinai (on this see also my article ‘On Causing a Secular Jew to Sin,’ Tzohar 25 [Spring 5766], pp. 9–20, especially ch. 3).

[97]     See the Commentary on the Mishnah, Kelim 17.

[98]     See, however, Henshke’s challenge to this assumption in his article in HaMa’ayan cited above in n. 40.

[99]     See my article on the second root and, in greater detail, the fourth book in the quartet.

[100]    Regarding the question whether there are indeed Torah-level commandments here, see the articles on the first and second roots.

[101]    See the beginning of the article on the twelfth root.

[102]    See the honored jubilarian’s article ‘Commands, Obligations, and Goals,’ cited above in n. 70; also the article on the twelfth root, and briefly the article on the eighth root. Yaakov Levinger discussed this at length in Darkhei HaMaḥshavah HaHalakhit Shel HaRambam, chapter 3 sec. B (subsections 2–3). There he distinguishes between two different types: religious-legal determinations and addressless directives. I shall not need that distinction here.

[103]    See a detailed survey of all commandments of this type at the end of the article on the twelfth root. See also Levinger in the previous note. In his HaRambam KeFilosof VeKeposek, pp. 73–74, he discusses this too, distinguishing between procedural commandments expressed in the language of ‘law’ and commandments imposing obligations, written in the language of ‘commandment.’ All the procedural commandments do not appear in Maimonides’ list of necessary commandments at the end of the list of positive commandments in Sefer HaMitzvot.

[104]    See The Secret, p. 194.

[105]    On this matter see RYFP’s introduction, the introduction to Mitzvot HaMelekh, ch. 7, the introduction by Rabbi Weitzman to his book Mitzvah Berah, Gutman, and the Wikipedia entry ‘Taryag Mitzvot.’ The aim also affected the arrangement of the enumerations of the commandments (whether according to the order of the Torah portions or some other order). See Twersky, pp. 187–188.

[106]    On the other hand, from the standpoint of Torah study, it is known that the Chafetz Chaim greatly encouraged study of sacrificial commandments not applicable in our time, and even established a special kollel for that purpose. There the aim was that the Temple would soon be rebuilt, yet the fact remains that in his enumeration of the commandments he did not include those commandments.

[107]    See a brief discussion of these remarks in The Secret, pp. 183–184.

[108]    In Mitzvot HaMelekh, however, at the end of chapter 1 of the introduction, the author notes words of the Vilna Gaon itself cited in Aderet Eliyahu, beginning of parashat Ki Tissa, where he does touch on the enumeration of the commandments and even uses Maimonides’ third root (that commandments not applicable for future generations are not counted).

[109]    He also argues there that the whole Torah consists of commandments, all of them general categories, and one should not distinguish between primary and secondary matters in this respect (somewhat like the well-known criticism of Maimonides’ thirteen principles).

[110]    See the introductory article to the roots.

[111]    See also Gutman, chapter 3, especially pp. 45–46 and thereabouts.

[112]    Mabit already noted this in his introduction to Kiryat Sefer, ch. 6. There, however, he proposes consequences according to Maimonides (with regard to lashes), and in my humble opinion they are incorrect; see the discussion immediately below. Pri Megadim also discusses this at the beginning of his ‘general introduction’ to his book, and RYFP in his general introduction to part 1. See also Mitzvot HaMelekh, introduction, ch. 2 (he concludes like RYFP and adduces several examples). The main points of their remarks are brought below.

[113]    Jerusalem, no. 368, printed in the Frankel edition of Sefer HaMitzvot before Maimonides’ introduction.

[114]    The relevant roots: 7, 9, and 10–12.

[115]    I proved this in the article on the ninth root, where I showed that Maimonides has a singular method on this matter.

[116]    See on this the article on the sixth root.

[117]    His proof from Ran on Nedarim 8a is rejected in the article on the second root.

[118]    See Moses Halbertal, ‘Maimonides’ Sefer HaMitzvot: The Architecture of Jewish Law and Its Theory of Interpretation,’ Tarbiz 59 (5750), p. 457. See also Twersky’s book, especially chs. 1 and 4. Hanina Ben-Menahem also noted this in The Secret. Ben-Menahem compared Ibn Ezra and Maimonides, but it seems that he (like others) did not fully identify the aspect presented here and all its implications.

[119]    In this respect there is some similarity to Bahag, who places his enumeration of the commandments at the beginning of his halakhic work. See, however, Gutman, ch. 1, where he argues that unlike Maimonides, Bahag probably did not intend his enumeration of the commandments as a skeleton or introduction to his work. In his opinion, originally it was with Bahag a separate composition, perhaps a remnant of the earlier tradition mentioned above.

[120]    See his aforementioned article in Tarbiz 59.

[121]    Levinger, however, in Darkhei HaMaḥshavah HaHalakhit Shel HaRambam, p. 120, raises the conjecture that matters not entering the rigid structure of the Mishneh Torah were omitted. This is a far-fetched and implausible conjecture, and Twersky already commented on this (see his book, p. 181, around n. 98).

[122]    And indeed even with regard to these two roots there are many disputes whether they are intended in general to make a halakhic claim (= that rabbinic commandments or those derived through exegesis are commandments of lower halakhic status), or whether these are merely meta-halakhic claims with no halakhic consequence. See the articles on those roots.

[123]    I already noted above that there is not necessarily a dispute here. Every aim in the enumeration of the commandments will yield a different count and a different classification.

[124]    The Secret, end of p. 185.

[125]    Ibid., p. 188.

[126]    See on this HaMitzvah VeHaMikra, p. 50.

[127]    Bahag’s method is harder to present in general terms, and this depends also on the different textual versions of his work.

In the article on the seventh root I pointed to the conceptions underlying these approaches. I argued there that Maimonides is a rationalist, who relies on his generalizations, and therefore includes several details that appear separately in the Torah within one commandment. Nachmanides (and others), by contrast, are empiricists; that is, they cling to the bare facts without trusting our human processing of them. There are two ways to do this: either to treat the whole complex as one unit (like Nachmanides), or to treat every detail written in the Torah as an independent commandment (like Saadia Gaon).

[128]    In the article on the twelfth root I cited several commentators who took this principle literally and therefore erred in applying it. The details of the discussion and the different reasonings are complex, and I refer the reader there. The gist is that in the twelfth root Maimonides intends to establish his general aim in the enumeration of the commandments and to say that the considerations are also systematic and not only substantive. See there the implications of this understanding for his words.

[129]    See on this the beginning of the honored jubilarian’s article, ‘Commands, Obligations, and Goals in Maimonides’ Teaching,’ where he proposes a different solution for understanding the commandment of belief, namely that the commandment is to deepen and study, not the mere state of believing. See all his remarks there.

[130]    See at length in the article on the twelfth root. After writing this, I found that Gutman had already pointed to this dilemma in chapter 3, from p. 47 onward. He presents it this way: apparently there was an early traditional enumeration of 613 commandments (which he attempts to prove throughout the discussion there, and there is much to support it). The number itself has no substantive meaning; it is simply what emerged for the bearers of that tradition. For example, he writes, no one will insist on the division of the Talmud into sixty tractates. That is simply what came out; anyone who wishes may divide them differently. Maimonides therefore could have taken one of two paths: either adopt both the total number and the commandments included in it (as transmitted in that tradition, like Bahag and the authors of the Azharot), or build a new enumeration, but then not commit himself specifically to the number 613. Maimonides tried to have it both ways.

[131]    I already noted above, however, that it is possible Maimonides really learns nothing from the wording of the Sages, and that his proofs are based on the content of their statements rather than on the fact that the wording speaks of two commandments or one commandment.

[132]    Here too an advocate could argue that Maimonides’ definition of the term ‘Torah-level’ as an expression denoting the source of the law (namely, that it appears explicitly in Scripture) resolves the difficulty. This definition too is systematic rather than substantive in nature. But this definition appears also in Maimonides’ responsa and introductions, not only in preparation for constructing the Mishneh Torah, and therefore it seems that he understands these terms this way on the substantive plane as well.

[133]    Not every two commandments that are distinct on the halakhic plane will appear as two separate commandments in the enumeration of the commandments. See the example of neshekh and tarbit mentioned above.

[134]    See Yad Peshutah, pp. 70–74 and thereabouts. There he proposes to resolve the contradiction that I shall bring here through a constraint deriving from Maimonides’ systematic aim in his book and his relation to the language of Scripture (and see there many more examples). Here I propose a rather similar, though not identical, solution.

[135]    See a survey in Henshke’s article, ‘Remnants of Sefer HaMitzvot’ (cited above in n. 4). There Henshke discusses the relations among Maimonides’ three enumerations of the commandments: the one in Sefer HaMitzvot, the one appearing in the introduction to the Mishneh Torah, and the one appearing in parts at the beginning of the various halakhic units. On contradictions among the enumerations, see also Rabbi B. Z. Benedict, ‘On Maimonides’ Formulations in the Laws of Sabbatical and Jubilee Years,’ Torah SheBe’al Peh 28 (5747), p. 134 n. 12. See also S. Ettinger, ‘Two Notes on Maimonides’ Method in the Mishneh Torah,’ Sinai 106 (5750), pp. 229–242, especially p. 231; Levinger, HaRambam KeFilosof VeKeposek, p. 267. See also what R. Hayyim Heller wrote on this at the beginning of his edition of Sefer HaMitzvot, p. 14.

On contradictions within the Mishneh Torah itself, see Twersky, pp. 232–238.

[136]    The verse And you shall return to the Lord your God (Deut. 30:2) is interpreted by Maimonides (Laws of Repentance 7:5) as prophecy and promise, not as command. Nachmanides, in his commentary there (on verse 11), indeed disagrees with Maimonides on this point.

[137]    See on this matter also the distinction noted above between the language ‘positive commandment’ (a counted commandment written in the Torah) and the language ‘commandment’ (an obligation not appearing in the Torah as a commandment). See also chapter 2 of the honored jubilarian’s article ‘Commands, Obligations, and Goals in Maimonides’ Teaching,’ and throughout that article.

On commandments grounded in reason rather than command, see my book Enosh Kaḥatzir, Jerusalem 5768, gate 5 ch. 3; and my articles: ‘Commandment, Reason, and the Will of God,’ Tzohar 30 (Elul 5767), pp. 15–25; and in the daily study booklet of the Be’er Gar’in, no. 2, Yeruham 5764 (there I argued that this is the characteristic of all the laws appearing in Maimonides’ Laws of Repentance, but this is not the place to elaborate).

[138]    There is here a certain innovation, namely that the enumeration of the commandments that introduces the halakhic units in the Mishneh Torah is from the outset not entirely congruent with the enumeration and character of the commandments in Sefer HaMitzvot. This is only a possible suggestion, and the matter still requires clarification.

[139]    Bokser, in his article (above n. 4), explains this by saying that Sefer HaMitzvot was written for the broad public and does not have a philosophical aim. On the contradictions between Sefer HaMitzvot and the Mishneh Torah see also Y. Sinai’s article cited above, p. 157 n. 10, especially pp. 179–181. See also Henshke, ‘On Maimonides’ Halakhic Thinking’ (above n. 4).

[140]    See examples in Rabbi B. Z. Benedict, ‘On Maimonides’ Way in Sefer HaMitzvot,’ Torah SheBe’al Peh 9 (5757), pp. 93–110. See there an explicit source from Maimonides’ own responsum (Jerusalem 5694, no. 69, pp. 69–70). From Maimonides’ words it emerges that at times he is not precise in his language and law in Sefer HaMitzvot because his aim is to present the principal laws. Even according to my first approach, this can be explained by saying that Maimonides wanted to present in Sefer HaMitzvot all the headings of the Mishneh Torah so that nothing would be omitted, but he did not see a need to be exact about the contents, since he would reexamine them when writing his great halakhic work. Maimonides hinted at this in his introduction when he wrote: My intention in this work is not to explain the law of any of the commandments, but only their enumeration. This is also stated explicitly in Kin’at Soferim at the end of root 14. See also the introduction to Mitzvot HaMelekh, chs. 4–5, where several examples are brought. See also Twersky, p. 178, and Henshke’s article ‘On Maimonides’ Halakhic Thinking’ (above n. 4).

R. Moses ibn Tibbon, however, in the translator’s introduction writes that he preferred his translation of Sefer HaMitzvot to that of R. Abraham HaLevi Ḥasdai because according to Ḥasdai’s translation several contradictions remain between Sefer HaMitzvot and the Mishneh Torah, and he notes that he possessed written testimony that Maimonides himself wrote that everything ought to be in harmony. Presumably, however, the intention is not to every detail literally.

[141]    Above, however, I proposed a different explanation for the use of Arabic.

[142] With regard to rules of individuation, this is easier to understand. Systematic classification and categorization can serve as an indication of substantive classification. At the end of the article on the twelfth root I brought an analogy from philosophy of science, which sees a connection between considerations of elegance and efficiency and substantive considerations. In science too, the more convenient and elegant theory is also considered more correct. That is, the systematic efficiency of classification and generalization serves as an indication of the truth of the generalization (= the law or scientific theory). This issue is discussed at length in my book Et Asher Yeshno VeAsher Einenu, Jerusalem 5766, gate 2.

This also connects to the rationalist features in Maimonides’ interpretation, which sees his a priori intellectual reasonings as measures for understanding the objective situation (= understanding the commandments themselves). Therefore he allows himself to generalize and to establish causal relations among commandments. Nachmanides, as an empiricist, adheres more closely to biblical and rabbinic text and places less trust in a priori considerations. Therefore he opposes Maimonides’ generalizations, especially when they seem to run counter to Talmudic material (= the facts, in the scientific analogy). The rationalist-empiricist dispute about methods of interpretation runs like a thread through several of the disputes between Maimonides and Nachmanides in the various roots. See the introductory article to the roots, especially the article on the seventh root, and others.

All this concerns rules of individuation, which deal with classification, differentiation, and categorization. Here there is some analogy between science and Jewish law. But with respect to the problem of the definitional commandments, matters are much more difficult. There we are not speaking of methods of classification and arrangement (individuation), but of filtering rules. That is, these commandments are not commands at all halakhically, and systematic elegance—even if it can serve as an indication of substantive classification—cannot turn them into commandments in the halakhic sense.

[143] For example, see his article ‘Commands, Obligations, and Goals in Maimonides’ Teaching,’ in chapter 2, where he makes use of the fifth root, and immediately afterward (in n. 5) also the fourth root, and many other such examples.