A Call for Order: Maimonides and the Mishnah

Yaakov Taubes

With Daf Yomi having started Massekhet Gittin, many students will be examining the question as to why Gittin, the section of the Talmud dealing with divorce law, comes before Kiddushin, the section dealing with marriage; we would expect Kiddushin to come first, since marriage must obviously precede divorce. Many have either quipped, or answered genuinely, by invoking the talmudic principle that ‘God creates the cure before the disease’ and thus divorce law, which can serve as the basis of the cure for an unsuccessful marriage, comes first.^[1] Whether this is said as a somewhat inappropriate joke that belittles the pain that can accompany divorce, or is suggested more honestly as a reflection of the fact that perhaps unlike in other religious traditions, a couple needn’t be trapped in a bad marriage according to Jewish law because the Torah sanctions and even requires divorce in some scenarios, it is not really a satisfying answer to the question regarding the order of the tractates of the Mishnah/Talmud. Other answers to this particular question, which some may find more satisfying, have been suggested.^[2] 

If we take a step back, however, we can ponder if this is really a valid question in the first place. Put differently, does the order of the massekhtot really matter such that we should be concerned that something seems out of place? There is certainly no logical reason to believe that one is supposed to learn the massekhtot in a particular order, be it the way they are traditionally arranged, or in any other order, nor, we might add, was there ever an agreed upon order in which one is supposed to study Mishnah/Talmud, at least before the advent of Daf Yomi. While the Talmud itself does occasionally comment on the juxtaposition of one massekhta to another, it is not clear that the provided explanation impacts our understanding of the overall order (more on that below). As is well known, while built around the Mishnah, the Talmud certainly did not feel itself bound to the topics or themes contained therein, and often diverged and digressed independent of it.^[3]

This question is seemingly addressed by the Talmud itself, which states “ein seder la-Mishnah,” – “there is no order to the Mishnah,” – meaning, as the Gemara clarifies, that there is no order of two different massekhtot, though within any given massekhta there is a seder to the presentation of the subjects.^[4] The Iggeret Rav Sherira Ga’on, an early medieval letter which charts the history of the Oral Law, understood this to mean simply that there is no formal order to the massekhtot, and we do not really know which was taught first.^[5] The mishnayot in most massekhtot require knowledge from elsewhere to understand the material contained therein, and learning the mishnayot in order does not necessarily alleviate that problem. In fact, the very first mishnah, Berakhot 1:1, asks what time the evening Keri’at Shema may be recited, and answers that the earliest time is the same time that the kohanim may resume eating their terumah (after having purified themselves from a ritual impurity). The specifics as to when the kohanim may in fact continue to eat their terumah are discussed elsewhere, and the student of this first mishnah is required to either have studied that section – out of order – or to have a commentary explain those specifics, to be able to understand the reference here.^[6] Studying the massekhtot in order, whatever order that is, thus does not necessarily facilitate easier comprehension, which may mean that questions or problems relating to the order are not really so outstanding.^[7] To put the point differently, the question of why Gittin appears before Kiddushin is relevant only to someone studying the massekhtot in a specific order like followers of the Daf Yomi. Nobody would ask an individual or a yeshiva choosing to learn Massekhet Gittin one year why they weren’t starting with Kiddushin instead.^[8]

This is not to say, however, that there is no organizing principle behind the order of the massekhtot at all, or that such a structure is without merit. The order may be useful as a convention for memorization, for locating a given massekhta and being able to identify its place among others, or for some other practical reason. Additionally, some Rishonim contended with the few places where the Talmud itself noted the placement and juxtaposition of a given massekhta or massekhtot (e.g., Sotah, Nazir, and Shevuot), by admitting that some tractates were indeed placed specifically and intentionally, yet still maintained that most were not, as expressed by the Talmud’s statement ein seder la-Mishnah.^[9]

So, while we may prefer that the massekhtot of Rosh Ha-Shanah, Yoma, and Sukkah, for example, appear consecutively, in accordance with the order of the holidays on our calendar which they discuss,^[10] we can acknowledge that our ordering convention may have been based on other factors, and that since ultimately the order does not really matter, the issue should not really bother us.^[11] Our initial question as to why Gittin precedes Kiddushin, then, may be interesting to ponder, but it is not necessarily one that requires a sophisticated answer, as conventions in general can have different reasons behind them, such as ease of access, memorization, and the like. 

A number of other medieval rabbinic figures, however, had a more limited view of ein seder la-Mishnah, instead claiming that it only refers to the way the massekhtot were originally taught, but certainly they were organized at a later date into the order we have now. As proof, they cite the very same Talmudic passages dismissed above as exceptions by those who denied that there was a reason-based order. In this latter camp’s view, the concept of ein seder la-Mishnah is a pedagogical one, stating the idea that the correct understanding of one massekhta is not contingent upon a statement from another. However, from a literary point of view, there is certainly a reason for the order of the massekhtot, and thus the question as to the reasoning for the placement of Gittin before Kiddushin remains a valid one.^[12]

While one can find scattered comments among proponents of this second approach to explain the placement of individual massekhtot, the most prominent and well-known authority, and the only one to systematically clarify the placement of each and every massekhta, is Maimonides. In his Introduction to the Commentary on the Mishnah, Maimonides explains the sequence of the six sedarim (orders) of the Mishnah as well as the individual massekhtot within them. He argues for a well-reasoned and organized flow of the massekhtot based both on verses in the Torah and logic, which dictated to the redactors of the Mishnah precisely how everything should be laid out. 

When it comes to the third order, Seder Nashim (where Gittin and Kiddushin are found), Maimonides first explains that it opens with Yevamot, the massekhta dealing with levirate marriage, instead of Ketubot, the massekhta which deals more broadly with marriage contracts (a seemingly questionable decision when we consider that a marriage contract must have been established before a levirate marriage would even be relevant), because marriage in general is voluntary while a levirate marriage can be forced, as the beit din instructs the brother of the deceased childless husband to either marry the widow or release her. Maimonides then writes, “And starting with matters that are compelled before [discussing] matters that are not compelled is correct and fitting.” As many have noted, this reasoning is far from straightforward, especially in light of the cases in the Talmud where marriage (and divorce) can be compelled, but it serves as an example of Maimonides’s commitment to finding an underlying reason for the placement.

At the end of his discussion of Seder Nashim, Maimonides addresses our initial question – why is Gittin before Kiddushin – head on. Having explained why Ketubot comes right after Yevamot, since they both discuss the subject of sexual union, he acknowledges that Kiddushin should still have preceded Gittin. He answers that the order is following the sequence of the verses in the Torah (Deut. 24:2) which describe a situation in which a woman who is divorced and then marries another man (and the subsequent prohibition for the original husband to take his ex-wife back after she had been married to another man), and in which divorce is discussed before marriage. He notes that the Talmud specifically derives laws from this particular sequence, invoking the Talmudic phrase “Entering (marriage) is compared to leaving (divorce),”^[13] which demonstrates that these verses and their order are halakhically significant. Once again, later commentaries have questioned this reasoning on multiple fronts, but what concerns us here is that Maimonides unquestionably sees a clear intentionality and logic to the placement of the massekhtot.^[14]

Now, some readers may be aware of a simple and straightforward answer to the question with which we started. Gittin comes before Kiddushin because the massekhtot are organized within each seder according to the number of chapters contained in each given massekhta, starting with those that have more and moving sequentially to those with less. Thus, the order of Seder Nashim is: Yevamot (16 chapters), Ketubot (13), Nedarim (11), Nazir (9), Sotah (9), Gittin (9), Kiddushin (4). Yevamot leads off Seder Nashim because it has the most chapters, and Gittin comes before Kiddushin because it has more chapters. This idea was first noted (at least in print) by Abraham Geiger (1810-1874), a German rabbi and scholar, and an important figure in the Wissenschaft des Judentums.^[15] This solution was well received, and one can still hear it invoked in both popular and religious settings. One can still question why an order based on chapter quantity was chosen (as mentioned above, perhaps it was easier to arrange on a shelf?), but it does not seem to necessarily be worth searching for any other underlying reasons.

We are thus left with a question regarding all those who sought to find logic in the placement of the massekhtot. Why did they go to the trouble when there was a seemingly more obvious solution? It does not seem probable that the chapter order is a coincidence. This question is particularly acute regarding Maimonides, who not only was comprehensive in his explanations, but also had to rely on reasoning that was far from intuitive, as noted above. It seems unlikely that he was not aware of the chapter lengths, as he goes out of his way to record how many chapters appear in the Mishnah altogether, and one would have expected him to notice this obvious sequence within the orders.^[16]

To answer this question, it must first be noted that the organizational principle of “most chapters to least chapters” does not hold completely true across the board. In Seder Nezikin, the first three massekhtot, Bava Kamma, Bava Metzia, and Bava Batra, have 10 chapters each, and are followed by Sanhedrin (11), Makkot (3), Shevuot (8), Eiduyot (8), Avodah Zarah (5), Avot (5), and Horayot (3); this seder is therefore apparently not solely organized based on the number of chapters in each massekhta. This problem can be solved, though, by citing the Talmudic statement that the first three massekhtot originally constituted one single larger massekhta, known as Nezikin, which thus consisted of 30 chapters and therefore was fit to start the seder.^[17] Furthermore, many have argued that Makkot is actually part of Sanhedrin, as is indeed implied by the Gemara’s first comment there after the mishnah; this would negate the issue of Makkot preceding Shevuot despite being shorter, and the seder can now be said to be arranged according to the quantity of the massekhta’s chapters. However, while the first three massekhtot are almost universally viewed as one, Maimonides (among others) specifically disagrees with those who place Makkot as part of Sanhedrin and not as a separate massekhta.^[18] In this arrangement, then, the massekhtot in Nezikin are not arranged completely in the order of most chapters to least.^[19]

In Seder Zera’im, the matter is even more complicated. The order is arranged as follows: Berakhot (9), Pei’ah (8), Demai (7), Kilayim (9), Shevi’it (10), Terumot (11), Ma’aseir (5), Ma’aser Sheini (5), Hallah (4), Orlah (3), and Bikkurim (4). This arrangement of the massekhtot is very clearly not based on chapter length, and there do not appear to be any good solutions to this problem.^[20] Maimonides may well have been aware that many of the orders are indeed organized according to the quantity of chapters in its massekhtot, but he saw enough exceptions to keep him from concluding that this was the primary reasoning behind the placement. Additionally, there is a further issue which the quantity of the chapters’ organizational structure does not deal with, i.e., the question of where to place the massekhtot that have the exact same number of chapters. In Seder Mo’eid, for example, the order of the massekhtot based upon the number of chapters is as follows: Shabbat (24), Eiruvin (10), Pesahim (10), Shekalim (8), Yoma (8), Sukkah (5), Beitzah (5), Rosh Ha-Shanah (4) Ta’anit (4), Megillah (4), Mo’eid Katan (3), and Hagigah (3). As we can see, after Shabbat, every massekhta is tied with (at least) one other regarding chapter length. In Seder Nashim, Nazir, Sotah, and Gittin each have five chapters, making their order difficult to determine, and there are “ties” in the other orders as well.^[21] Relying upon an organizational structure based solely on the quantity of chapters would obviously not explain these cases.^[22] Between the exceptions and the ties, Maimonides (and others who believed the placement of the massekhtot to be logic-based) may have come to the conclusion that there must be a better explanation for the placement of each massekhta within a  specific seder, one that utilizes better reasoning than simple chapter length.

There may be one more piece to analyzing this question, at least as concerns Maimonides. Many students are surprised when they first discover that the massekhtot are seemingly organized based on the number of their chapters. It seems like a somewhat crude (and intellectually unsatisfying) system, and does not really express what we think of as a modern organizational arrangement. We would certainly never expect a library to organize books based on how many chapters or pages they contain or a teacher to assign texts to read in the order of their respective length. Lest anyone suggest that it would be anachronistic to expect Maimonides to come up with such an idea, it must be pointed out that there is another important book, albeit from a different faith, that is also (more or less) organized based on chapter length: the Quran, the holy scripture of Islam. After the first introductory chapter, the chapters that follow are not arranged in the chronological order of revelation or any other chronology. The only easily discernible organizational system is that the chapters are arranged roughly in order of their length, with chapters containing more verses preceding those with fewer verses. Many traditional Muslims believe that the chapter order was set by the prophet Muhammad, and subsequent attempts to reorganize the chapters according to chronology or another system were often met with resistance.^[23]

Maimonides certainly had some knowledge about the content of the Quran. Besides the fact that he spent much time living in Muslim lands, he references verses from the Quran in his Epistle to Yemen, written as advice and inspiration to the Jewish community that was suffering there. Perhaps Maimonides was also aware of the Quran’s organizational structure which appeared to be random and not based on any logic perceivable to outsiders. Indeed, in coming up with his ordering principles, he may even have been motivated to show that the organized structure of the Mishnah was not based on such a simple convention.

Aside from the possible connection to the Quran, why was Maimonides so set on finding internal reasoning in the ordering system of the Mishnah? Presumably, his concern rested on his understanding of the circumstances which led to the recording and redaction of the Oral Law to begin with. In the beginning of the Introduction to the Commentary on the Mishnah, Maimonides explains that R. Yehudah Ha-Nasi, “the holy prince” who organized the Mishnah, received the tradition from the elders before him, as part of a chain going back to Moses on Mount Sinai. R. Yehudah Ha-Nasi is described as wise, eloquent, and wealthy, and as having gathered all the statements of the great rabbis, recording and turning them into a great legal code. It would not make sense for such a momentous work, which was divided into precise sedarim and then subdivided into fixed massekhtot, to be lacking in coherence in terms of the order of those massekhtot. We can further suggest that this may have been especially important to Maimonides in light of his comments in the introduction to his Mishneh Torah, where he states that R. Yehudah Ha-Nasi wrote down the Mishnah in order to prevent it from being forgotten, as material presented in a logical order is presumably easier to remember. 

To be sure, Maimonides did not think that the order of the Mishnah was perfect or exclusively correct. He specifically chose to organize his own Mishneh Torah following an order which he thought was better suited for study, as opposed to the order of the original Mishnah. We might note in this context that in his Book of Women’s Law (Nashim) there, the laws of marriage indeed precede the laws of divorce (and levirate marriage), which is certainly more intuitive, however logical or Torah-based the order of the Mishnah was. This is not to say that Maimonides saw the Mishnah’s order as flawed, but perhaps he saw how his code, which was different than that of the Mishnah, could be organized differently and more effectively in terms of the goals he was pursuing.^[24]

Additionally, Maimonides, among other rabbinic authorities, went to great lengths to prove the authenticity and reliability of the Mishnah as the source for authoritative law going back to Sinai. This claim had to be defended against Karaites, Muslims, and later Christians, who denied it with different motivations and reasons. A seemingly haphazard and inconsistent organization system would simply not do. Maimonides firmly believed that just as the Written Law was given by God to man in a clearly intentional and discernible order, the same had to be true for the Oral Law as organized and recorded by R. Yehudah Ha-Nasi and other rabbis.

In this vein, it is worth noting that in addition to the comments of Maimonides regarding the Mishnah cited above, we find another programmatic statement of his that demonstrates what may have been at stake here. In a letter to R. Phineas ben Meshullam, a judge in Alexandria who had criticized the Mishneh Torah, Maimonides explained the difference between the Mishnah as a hibbur, a monolithic code, and the Talmud as a peirush, a discursive commentary.^[25] It is clear from Maimonides’s statement that he saw his own Mishneh Torah, a thorough and well-organized code, to be a hibbur and modeled after the Mishnah.^[26] If the Mishnah set the standard for what a code of Jewish Law should look like, it could not have a haphazard or crude organization system, which is perhaps why Maimonides rejected the descending order of chapters theory.^[27]

For those dissatisfied with Maimonides’s explanation, we may look at the order of the massekhtot in Nashim and conclude that Gittin comes before Kiddushin because it indeed has more chapters. But does that negate the possibility that the order has a deeper wisdom behind it? Students of the Torah are certainly aware that biblical verses, rabbinic passages, etc., can have multiple explanations that are far from mutually exclusive. The idea behind Maimonides’s argument, namely that the Mishnah seeks to reflect the idea and sequence contained within the Written Torah upon which it is based and from which all its laws come, can certainly be accepted. Ultimately, whether one believes this idea can be used to compellingly explain the Mishnah’s order as Maimonides contended, may depend on to what extent one is bothered by the question in the first place.

[1] See Megillah 13b.

[2] See Tosefot Yom Tov to Gittin 1:1.

[3] It must be pointed out that alternate orders of the tractates exist in other traditions and manuscripts. See J. N. Epstein, Introduction to the Mishnaic Text (Heb.), (Jerusalem, Israel), 980-1000, who carefully surveys the evidence for varied orders of the massekhtot among different communities and rabbinic authorities. As he notes, even the order of the six sedarim is not agreed upon in early rabbinic texts. While our order (mentioned in the Bavli) is Zera’im, Mo’eid, Nashim, Nezikin, Kodashim, and Tahorot, other texts have the order as Nashim, Zera’im, Tahorot, Mo’eid, Kodashim, and Nezikin. The question of the order of the sedarim would seem to have even less of a ramification. In any event, this article will focus on the order of the massekhtot as accepted nowadays and attested to in most traditional sources, but will note exceptions where relevant.

[4] See Bava Kamma 102a and Avodah Zarah 7a.

[5] Me’iri (Beit Ha-Bechirah to Nazir 2b) likewise writes that the order may be derekh mikreh, happenstance, and that students learn what their heart desires without concern for the order. Also see Tosefot Rabbeinu Peretz there. Despite their remarks, Rav Sherira Gaon and Me’iri still tried to offer a rationale for the order of some of the massekhtot.   

[6] Interestingly, an early commentary on the Gemara called Sefer Ha-Maftei’ach, authored by R. Nissim b. Jacob of Kairouan, was written to address this very problem. As the author notes, the passages of the Talmud are interwoven, with one often relying on another, and his work strives to provide the necessary background material to enable one to proceed through the Talmud.

[7] There are places in the Talmud where we are told that a given mishnah is actually based on a lesson from an earlier one. See, for example, Taanit 2a. But the fact that the Talmud points this out suggests that it was not a given that the earlier one had already been studied, and that it is no longer necessary to study the earlier teaching to understand the current topic.

[8] We might add that traditional Talmud study does not necessarily give precedence to an early chapter over a later one in a given massekhta. In many yeshivot, students typically learn a particular chapter from a massekhta and do not necessarily start from the beginning. While material later in a massekhta can often build on material presented earlier, this is not always the case. Indeed, there are some massekhtot in which the order of the chapters actually differs among different commentaries, but again, the order does not necessarily affect understanding.

[9] See Ramban, Rashba, and Ritva to Shevuot 2b for further discussion.

[10] Indeed, the Vilna Shas printing placed these massekhtot in the order in which they appear on the calendar, and many sets of Talmud remain organized in this way. However, this is not the order of the Mishnah, nor is it the order followed by the Daf Yomi program.

[11] It is worth pointing out that similar questions have been raised about the ordering of the books of the Bible. Different ordering systems exist within varied Jewish and Christian traditions, based on factors such as chronology, level of holiness, genre, and others. This topic has received much scholarly attention, but is somewhat beyond the scope of the present study.

[12] See Tosafot to Bava Kamma 102a, s.v. ein, to Bava Metzi’a 2a, s.v. shnayim, and to Bava Batra 2a, s.v. hashutafin, and the discussion in the Rishonim cited above.  

[13] Kiddushin 5a.

[14] As some have noted, the verse which precedes the one cited by the Talmud, Deut. 24:1, opens with “Ki yikah ish ishah…,” “When a man takes a woman,” which is referring to marriage (and, we might add, is a primary source for the mitzvah of kiddushin). It is only in the next verse that divorce precedes marriage; the Torah thus clearly places marriage before divorce, which somewhat undermines Maimonides’s argument. Presumably, Maimonides is picking up on the language of the Gemara, not the sequence of the verses, but his argument is certainly not as strong as it may seem. For an analysis of this and many other issues in the Introduction, as well as citations to the various commentators who have addressed them, see the version of the commentary edited and annotated by Ezriel Wolodarksy (Jerusalem, 2018) which includes helpful footnotes and insights.. 

[15] See citations and discussion in Chanan Gafni, Conceptions of the Oral Law in Modern Jewish Scholarship (Heb)., (Jerusalem, 2019), 216-218. Gafni also quotes contemporary scholars who either embraced or rejected Geiger’s theory.

[16] Maimonides’s own interest in chapter breakdown can be demonstrated by the fact that he included precisely one thousand chapters in the Mishneh Torah, surely no coincidence, and a further sign of Maimonides’s exactitude. See Marc Shapiro, Studies in Maimonides and his Interpreters, (Chicago, 2008).

[17] Ri Migash actually understood the Talmud to mean that all of Seder Nezikin originally constituted one massekhta. See citation in Ramban to Shevuot 2b and his rejection of this explanation.

[18] This is made clear by Maimonides’s tally of the massekhtot in Nezikin as eight, and his overall tally of the massekhtot as sixty-one; both of those numbers work only if one counts Sanhedrin and Makkot as distinct massekhtot. 

[19] It should be noted here that while Massekhet Avot is usually printed as having six chapters, thus making it longer than Avodah Zarah which precedes it and only has five chapters, this is because the sixth chapter of Avot is really a collection of talmudic passages that were added on at a much later point, and is not included by Maimonides or other medieval authorities as an actual part of the massekhta; hence our identification above of Avot containing 5 chapters. Additionally, while the order of Seder Kodashim would seem to work according to the quantity of chapters system, there is one small issue. In most standard editions, Massekhet Tamid includes seven chapters, which would make it longer than Keritot and Me’ilah which precede it and have six chapters each. However, Maimonides, in accordance with other medieval traditions, only has six chapters in Tamid, thereby avoiding the problem. Medieval and later authorities who had seven chapters would have further reason to question the order of chapters theory, a point that Geiger himself noted.   

[20] Independent of Maimonides, various explanations have been given for this discrepancy. See R. Reuven Margoliyot’s Yesod Ha-Mishnah Va-Arichatah, where he suggests that Seder Zera’im was a later addition as a separate order, its contents having initially been spread throughout other orders, with Berakhot in Mo’eid, Pei’ah in Nezikin, and others throughout Kodashim and Tahorot. This is an intriguing but ultimately unprovable thesis. See Epstein, Introduction, 985-988, who attempts a more systematic explanation but ultimately acknowledges that every system has exceptions for one reason or another. 

[21] It is worth noting that Maimonides’s arrangement actually has Gittin before Sotah, while the standard order has the reverse.  

[22] Epstein (see above) notes these “ties,” and adds that most of the diversity in massekhta order among different authorities is within massekhtot with the same number of chapters. While he sees this as a proof to the theory, one could also see it as a problem, as it acknowledges a certain amount of arbitrariness in the system, something which many find difficult to accept.     

[23] See Neal Robinson, Discovering the Qur’an: A Contemporary Approach to a Veiled Text, (Washington DC, 2003), 256-270. Robinson notes earlier scholars who noted the similarity between the Mishnah and the Quran in this respect, but disagrees, arguing that the distinctions in genre and style are too great to consider a comparison seriously. He also dismisses scholars who have compared this organization style to that of early Islamic poetry. More fruitful, he argues, is a comparison to the Pauline epistles of the New Testament which, with a few exceptions which Robinson explains away, are also in order of most chapters to fewest. See Robinson at length for other attempts to explain and arrange the chapters of the Quran. My thanks to Dr. Ari Gordon for his help with this reference.

[24] See further below and Isidore Twersky, Introduction to the Code of Maimonides (Mishneh Torah), (New Haven, 1980), 238-245, where he elaborates on the significance of Maimonides diverging from the Mishnah’s order.    

[25] See citation, translation, and analysis in Isidore Twersky, Introduction to the Code of Maimonides (Mishneh Torah), (New Haven, 1980), 30-37. 

[26] Similar statements can be found in his introduction to Sefer Ha-Mitzvot. See Isidore Twersky, “Some Non-Halakhic Aspects of the Mishneh Torah,” Jewish and Medieval and Renaissance Studies, ed. Alexander Altmann, (Cambridge, MA), 106-111.

[27] See Menahem Kahana, “The arrangement of the Orders of the Mishnah,” Tarbitz 76 1-2, (2007), 29-40, who discusses several of the ideas mentioned in this paper, although he does not focus on Maimonides. He notes that a clear editorial hand was involved in the formation of the Mishnah’s structure into 6 orders and 60 tractates, numbers that may have been decided upon before all the material had even been collected and organized. Perhaps a similar argument could be made regarding the order of the tractates themselves.