JOURNAL OF JEWISH STUDIES, VOL. LXI, NO. 1, SPRING 2010

The Unintentional Killer: Midrashic Layers in the Second Chapter of *

MICHAL BAR-ASHER SIEGAL Yale University

Abstract Mishnah Makkoth, or to be exact the last part of tractate , is imbued with midrashic materials. As previously noted in scholarly re- search the Midrashic sources can be revealed through study of its parallels in the legal Midrashim. This paper will define the reciprocality between the second chapter of Mishnah Makot and Deuteronomy dealing with Deuteronomy verses on the unintentional killer. I will illustrate both overt midrashim quoted in the Mishnah, usually with the underlying verse and the term øîàðù, shene’emar, and covert midrashim. In this last category I address different statements found in the Mishnah whose origins can be explained as emanating from an early midrashic interpre- tation of the biblical verses, even though such a source is not explicitly mentioned in the Mishnah itself. The comparison raises again the ques- tion of the various sources of Mishnah Sanhedrin-Makkoth in particular and of Rabbi’s Mishnah in general.

ifre Deuteronomy on the passage dealing with the unintentional killer S (Deut. 19:1–13), as in other places, often cites our Mishnah, the Mish- nah of Rabbi Judah the Patriarch. At the same time, m.Makkoth, or to be exact the last part of tractate Sanhedrin,1 is imbued with midrashic materials to a greater extent than most other mishnaic tractates. Midrashic passages in tractate Sanhedrin, as J. N. Epstein concluded, ‘originated mostly from the

* Parts of this paper are based on talks given at the J. N. Epstein Prize symposium, in the Department, Hebrew University (January 2002), and the 39th Annual Conference of the Association for Jewish Studies, in Toronto, Canada (December 2007). I am thankful to Profes- sor Menachem Kahana for his many helpful remarks, as well as to Professors Christine Hayes and Steven Fraade for their thorough reading and useful comments on previous versions of this article. English translations of the rabbinic sources are taken (with adjustments according to manuscripts’ versions) from The Mishnah. Translated from the Hebrew, with introduction and brief explanatory notes by Herbert Danby (Oxford / New York, NY: Oxford University Press, 1933); Sifre: a Tannaitic commentary on the . Translated from the Hebrew with introduction and notes by Reuven Hammer (New Haven: Yale University Press, 1986); The . Translated from the Hebrew by (Hoboken, N.J.: Ktav, 1977–1986); Trac- tate Makkot. Translated by Jacob Neusner (Atlanta: Scholars Press, c.1991). 1 The two tractates used to be one, at least in Palestinian traditions, as it still appears in Mishnah manuscripts such as ms. Kaufmann. See J. N. Epstein, Mevo ot le-sifrut ha-Tana im: Mishnah, Tosefta u-midreshe-halakhah, Jerusalem / Tel Aviv, 1957 (in Hebrew), p. 417. THE UNINTENTIONAL KILLER 31 school of ’.2 The midrashic origins of m.Sanhedrin can be re- vealed through a comparative study between the Mishnah and the halakhic Midrashim. The relevance of such a study has been shown in the works of Menachem Kahana, which compare Sifre Zuta Deut. and m.Sanhedrin 3:1;3 Steven Fraade’s work on m.Sanhedrin 2 and Sifre Deut. on ‘the of the king’;4 and the work of Aaron Shemesh showing the close connection between m.Makkoth and Scriptures.5 This paper similarly examines the mutual relations between the second chapter of m.Makkoth and Sifre Deuteronomy on the accidental killer. I shall deal with ‘overt ’ (éåìâ ùøãî) that includes a quoted verse and intro- duces it with the formula ‘as it is said’, øîàðù. I shall also examine ‘concealed Midrash’ (éåîñ ùøãî), i.e. halakhot in the Mishnah whose sources can be explained as a midrash on the biblical verses, though such a midrashic con- nection does not explicitly appear in the Mishnah itself. This examination of the relationship between the Mishnah and the Sifre offers a fresh set of ex- amples of the midrashic nature of this chapter and of tractate Makkoth in general. The results of this survey, in turn, offer an additional perspective on the long-standing deliberation of the nature of the Mishnah–Midrash rela- tionship.

1. m.Makkot 2:6 An example of a that relies explicitly on of biblical verses can be found in m.Makkoth 2:6: øéòì ïéîã÷î ãéæî ãçàå ââåù ãçà äìéçúë øîåà äãåäé øá äñåé §ø åäåâøä äúéî áééçúéðùî ¬íùî ïúåà ïéàéáîå ïéçìåù ïéã úéáå èì÷î åúåà ïéøéæçî úåìâ áééçúéðù éîå ¬åäåøèô äúéî áééçúéð àìù éîå ¬äì øáãîá© §îåâå åèì÷î øéò ìà äãòä åúà åáéùäå §ðù åîå÷îì ®¨äë

Rabbi Yose Bar Yehuda says, ‘To start with, both the intentional killer and the unintentional killer hasten to the city of refuge, and the court sends and fetches them from there. Whoever is deemed worthy of being killed is killed and whoever is not deemed worthy of being killed is let go, and whoever is deemed worthy of exile is brought back to the city of refuge as it is said “and the assembly sends him back to the city of refuge to which he fled”.’ (Num. 35:25) In this example, there is a list of halakhot in which only the last is pro- vided with biblical support. However the same Mishnah is quoted in Sifre

2 J. N. Epstein, Mevo ot le-sifrut ha-Tana im: Mishnah, Tosefta u-midreshe-halakhah, p. 418. 3 M. Kahana, Sifre Zut.a : Muva ot mi-Midrash Tana iH. adash, Jerusalem, 2002 (in Hebrew), pp. 235–236. 4 S. D. Fraade, ‘ “The Torah of the King” (Deut. 17:14–20) in the Temple Scroll and Early Rabbinic Law’, in The Dead Sea Scrolls as Background to Postbiblical and Early Chris- tianity: Papers from an International Conference at St Andrews in 2001, ed. James R. Davila, pp. 25–60 (Studies on the Texts of the Desert of Judah 46; Leiden: E. J. Brill, 2003). 5 A. Shemesh, Elo hen hagolin, MA thesis, Ramat Gan, 1988 (in Hebrew). 32 JOURNAL OF JEWISH STUDIES

Deuteronomy, this time with biblical verses supporting each stage of the pro- cess: ïéá ââåùá ïéá ùôðä úà âøåää ®§åà äãåäé §øá éñåé §ø äéä ïëéî ïúåà ïéàéáîå ïéçìåù ïéã úéá ®èì÷îä éøòì ïéîã÷î ìëä ãéæéîá íéøáã© §åâå åç÷ìå åøéò éð÷æ åçìùå §ðù ¬åäåâøä äúéî áééçúðùî ®íùî çöøä úà äãòä åìéöäå §ðù ¬åäåøèô äúéî áééçúð àìù éî ®¨áé ¬èé §ðù ¬åîå÷îì åúåà ïéøéæçî úåìâ áééçúðù éî ®¨äë ¬äì øáãîá© §åâå ®¨íù© §åâå äîù ñð øùà åèì÷î øéò ìà äãòä åúåà åáéùäå

From this, Rabbi Yose b. Rabbi Yehuda used to say: ‘The killer of a person, whether intentionally or unintentionally, all hasten to the . The court sends and fetches them from there. Whoever is deemed worthy of killing is killed as it is said: “the elders of his town shall send for him, bring him back from the city, and hand him over to the avenger of blood to die” (Deut. 19:12). Whoever is not deemed worthy of killing is let go as it is said: “The assembly must protect the one accused of murder from the avenger of blood” (Num. 35:25), and whoever is deemed worthy of exile is brought back to the city of refuge as it is said “and the assembly sends him back to the city of refuge to which he fled”.’ (Num. 35:25) The parallel Sifre, therefore, fleshes out the biblical origins of the mishnaic statements. According to this, the laws described in the Mishnah are actually derived from two biblical texts—Deuteronomy 19 and Numbers 35—and are not the rephrasing of a single biblical source. The Sifre’s citation of this Mishnah is unique in that it is the only instance in the Sifre’s treatment of the unintentional killer in which a halakha derived from Deuteronomy appears side by side with halakhot derived from the . This goes against the usual practice of the Sifre in this section, in which the focus is solely on halakhot derived from Deuteronomy, while ignor- ing any additional halakhot on the same matter that are mentioned elsewhere in the Bible. It is possible that this exception was caused by the fact that here the Sifre is quoting a pre-existent Mishnah that had already combined the verses from Deuteronomy and Numbers to create a coherent juristic process, although without citing the verses themselves.

2. The axe and the wood, m.Makkot 2:1 In contrast to this last example in which a verse was explicitly mentioned in the Mishnah, the next example focuses on a mishnah that does not explicitly quote a biblical verse. Instead, the wording of the verse can be detected in the phrasing of the mishnaic halakha. Deuteronomy 19:5 reads: õòä úøëì ïæøâá åãé äçãðå íéöò áèçì øòéá åäòø úà àáé øÖàå íéøòä úçà ìà ñåðé àåä úîå åäòø úà àöîå õòä ïî ìæøáä ìÖðå ®éçå äìàä

A man may go into the forest with his neighbour to cut wood, and as he swings his axe to fell a tree, the metal may fly off from the wood and hit his neighbour and THE UNINTENTIONAL KILLER 33

kill him. That man may flee to one of these cities and save his life. The question that arises from this verse is well-known. What is the situation described by the words: ‘the metal may fly off from the wood and hit his neigh- bour and kill him’? Does ‘the wood’ refer to the wooden handle so that the metal axe-head flies off of the handle as the axe is swung, or does ‘the wood’ refer to the tree (as it does throughout the verse) so that the metal axe-head flies off the handle upon striking the tree? M.Makkoth (2:1) probably reflects this same interpretive question and its subsequent clarification: íéøîåà íéîëçå äìåâ åðéà øîåà éáø âøäå åúð÷éî ìæøáä èîùéð äìåâ åðéà íéøîåà íéîëçå äìåâ øîåà éáø òé÷áúîä õòä ïî »äìåâ

If the metal slips/flies off from its handle, Rabbi says: ‘he does not go into exile’, and the sages say: ‘he does go into exile.’ From the wood that is being chopped, Rabbi says: ‘he goes into exile’ and the sages say: ‘he does not go into exile.’ The Mishnah advances two possible readings of the scenario described by the verse and attaches a dispute to each reading.6 Understanding ‘the wood’ as referring to the axe handle, the sages view the incident as an accidental homi- cide requiring flight to a city of refuge; but Rabbi disagrees. This indicates that in the sages’ view, the verse must refer to an axe-head that flies off its handle. Rabbi, however, understands ‘the wood’ as referring to the wood that is being chopped, and views this incident as an accidental homicide requiring flight to the city of refuge; the sages disagree. This indicates that in the view of Rabbi, the biblical verse must refer to an axe-head that flies from the wood that is being chopped. A close examination reveals that as in m.Makkoth 2:6, the scriptural back- ground plays an important role in the formation of these legal rulings. Ac- cording to this mishnah, both the sages and Rabbi interpret the verse in a limiting way: only the case that each one maintains is described in the verse requires exile and any other case—including the alternative case proposed by their partner in the dispute—does not require exile. The assumption that a dispute in biblical interpretation underlies the legal dispute in the mishnah is made more explicit in the Sifre: ò÷áúîä õòä ïî øîåà éáø ¬ò÷áîä õòä ïî ìæøáä ìùðå

‘And the metal flew off from the wood’—[from the wood] that chops. Rabbi says: from the wood that is being chopped. On the other hand, we find in the Tosefta 2:1: §ø ¬äãåäé §ø éøáãÐäìåâ åðéà äæ éøä ò÷áúîä õòä ïî úÌòKÙa äæúéð äìåâ äæá §îåà

‘If a splinter flies from the wood that is being chopped [and kills] he does not go into exile’, says Rabbi Yehuda; Rabbi says, ‘in this case he goes into exile.’

6 Tannaitic passages in both discuss a few options for the dispute over the biblical interpretation that underlies this Mishnah. See below. 34 JOURNAL OF JEWISH STUDIES

And in 2:6 we find: §îåà §ø ¬äãåäé §ø éøáãÐäìåâ äæ éøä ò÷áúîä õòä ïî ìæøáä ìùð äìåâ åðéà äæá

‘If the metal flies as a result of the [blow to the] wood that is being chopped he goes into exile’, says R. Yehudah; Rabbi says, ‘in this case he does not go into exile.’ The contradictions are apparent. First, there is a contradiction between the Tosefta, on the one hand, and the Mishnah and Sifre, on the other hand, con- cerning the opinion of Rabbi when the metal flies as a result of the blow to the wood that is being chopped. The Mishnah and the Sifre claim that Rabbi (as opposed to the sages) rules that exile is necessary in this case, while according to the Tosefta, Rabbi holds that exile is not necessary in this case. Second, the first part of the Tosefta passage includes an additional case in which a splinter flies off from the chopped tree and kills. This case introduces a new element, the splinter, and it is not clear where this case fits in relation to the halakhot we find in the Mishnah and the Sifre Deuteronomy. It seems that just as the need to interpret the ambiguous biblical verse underlies the legal disputes in the Mishnah, so too the Tosefta’s additional splinter tradition arises from this same need. Medieval commentaries on the Mishnah confront the question of how to understand the two cases listed in the Mishnah (‘if the metal flies/slips from the handle . . . from the wood that is being chopped ...’), in light of the Tosefta and the biblical ambiguity. Maimonides7 understood the second clause in the Mishnah (‘from the wood that is being chopped, Rabbi says: ‘he goes into exile’ and the sages say: ‘he does not go into exile”) as governed by the protasis articulated in the first clause, ìæøáä èîùð, ‘If the metal slips/flies’ either from its handle, the first case in the Mishnah, or from the wood that is being chopped, the second case in the Mishnah. On Maimonides’ reading, the second case deals with the metal head flying off as a result of hitting the wood that is being chopped, just like the second case in the Tosefta. The first case in the Tosefta, according to this reading, is a completely new situation involving a splinter and is not related to the cases described in the Mishnah. ,8 on the other hand, understood the Mishnah as if the protasis in the first clause (‘if the metal slips/flies’) does not carry over to the second clause. Thus, the second case no longer deals with the metal head of the axe flying from the wood that is being chopped, but rather with a splinter flying from the wood that is being chopped as a result of the axe’s blow. The Mishnah’s ‘from the wood that is being chopped’, ò÷áúîä õòä ïî, should be understood as follows: ‘[and if a splinter flew] from the wood that is being chopped’. Accord- ing to this reading, the second case in the Mishnah is parallel to the Tosefta’s halakha regarding a splinter. As De Vries pointed out,9 the Targumim suggest these two interpretations

7 Maimonides, Mishne Torah, Hilkhot Rotse ah, 6:15. 8 b.Makkoth, 7b. ¯ 9 B. De Vries, Meh. k. arim be-sifrut ha-Talmud, Jerusalem, 1968, p. 27 (= De Vries). THE UNINTENTIONAL KILLER 35 in their translations of this verse. For ìùðå, Onqelos has ‘óìúùéå’, pulled out, implying that the splinter was pulled out from the wood because of the axe, similarly to Rashi’s explanation. By contrast, Yonathan has ‘èåîùéå’, dropped, implying that the metal dropped from the wood, a reading analogous to Maimonides’ view. Maimonides’ explanation seems like the simpler reading of the Mishnah, though the discrepancy in Rabbi’s opinions remains, while Rashi’s expla- nation aligns the Tosefta with the Mishnah’s halakha, solving the above- mentioned problem with the opinion of Rabbi.10 In other words, if we adopt Maimonides’ reading of the Mishnah we have an explicit attribution to R. Yehudah of the view that exile should result when an axe-head flies from wood that is being chopped and kills, which stands in direct contradiction to the view attributed to him in the Tosefta. By contrast, if we adopt Rashi’s reading, Rabbi in the Mishnah does not express an opinion on the question of an axe-head that flies from wood that is being chopped, eliminating any contradiction with the view attributed to him in the Tosefta. Modern scholars have similarly disagreed over the meaning of the Mishnah and Sifre and their relation to the Tosefta. Chanoch Albeck11 and Yosef Tzvi Dünner12 adopted Maimonides’ reading of the Mishnah, in spite of the prob- lems it causes for Rabbi’s view;13 while Lieberman14 and De Vries15 preferred Rashi’s explanation. Lieberman pointed to the word äæá, in this case, in both toseftan halakhot to support a connection between the two cases. One case is the äëåôä ú÷åìçî (reversed argument) of the other case, thereby contra- dicting the suggestion that the first toseftan case is an entirely new scenario. However, De Vries noted that according to Rashi’s explanation the word ïî, from, which appears in both cases of the Tosefta, would then mean two dif- ferent things in both halakhot: ò÷áúîä õòä ïî úò÷á äæúð,‘If a splinter flies from the wood that is being chopped ’; and ò÷áúîä õòä ïî ìæøáä ìùð, ‘If the metal flies as a result of the [blow to the] wood that is being chopped ’. De Vries and Epstein16 went even further and identified the opinion of Rabbi Yehudah in the Tosefta with the anonymous sages of the Mishnah. In both Talmudim we find baraitot that testify to different interpretive tra- ditions of this same biblical verse, and again these different interpretations differentiate between Rabbi and the sages. Furthermore it is made explicit that as before, the ambiguity of the biblical verses underlies the dispute in the

10 Both opinions appear in Rabbenu Hananel’s exegesis. 11 In his supplements to his commentary on the Mishnah, Ch. Albeck, Shishah sidre Mishnah, Jerusalem, 1975 (=Albeck), proposed to emend the Tosefta to: ò÷áîä õòä ïî úò÷á äæúéð, If a splinter flies from the wood that chops, to align the Tosefta with the Mishnah more easily. 12 J. H. Dünner, H. idushe ha-Ritsad, Jerusalem, 1981. (=Dünner). 13 Dünner tries to solve this problem using a third case he finds in the Tosefta, and claims: åú÷ìçîå éáøã àáéìà éàðú, ‘there are two opposing tannaitic views [as to Rabbi’s opinion in this case].’ 14 S. Lieberman, Tosefet , vol. 2, New York / Jerusalem, 1999, p. 164. 15 Ibid.p.25. 16 J. N. Epstein, Mevo ot le-sifrut ha-Tana im: Mishnah, Tosefta u-midreshe-halakhah, p. 194, discusses the Tosefta as part of his treatment of the in the b. Makkoth 7b. 36 JOURNAL OF JEWISH STUDIES

Mishnah. B.Makkoth 7b reads: àì àìäå åöòî ìæøáä ìùðå øîàð éëå íéîëçì §ø ïäì øîà àéðú äî äìòîì õò øîàðå äèîì õò øîàð ãåòå õòä ïî àìà øîàð õòä ïî äìòîì øåîàä õò óà ò÷áúîä õòä ïî äèîì øåîàä õò åùøã ãçà àø÷î ïäéðùå áø øîà éùà øá äééç áø øîà ®ò÷áúîä ïðáøå áéúë ìù†ðå úøåñîì íà ùé øáñ éáø õòä ïî ìæøáä ìùðå ïðéø÷ ìùðå àø÷îì íà ùé éøáñ

It has been taught: Rabbi said to the sages, ‘Does it say “and the iron slips from its wood”? Doesn’t it say “from the wood”. Moreover “wood” appears twice. Just as in the first instance the reference is to the tree that is being cut down so in the second case it is to the tree that is being cut down.’ Rabbi Hiyya bar Ashi said Rav said, ‘Both parties interpret the same verse of scripture, namely “And the iron flies/slips from the wood”. Rabbi maintains that the unvocalised letters of the text are determinative so we may [revocalise and] read the word as “and was hurled away”, while the Rabbis hold that the vocalisation of the letters of the text is determinative so that we can only read “and slipped”.’ As Dünner noticed, this BT passage has two parts. First, there is the tannaitic baraita that presents two arguments from scripture in support of Rabbi’s opinion in the Mishnah. Both arguments refer to the words õòä ïî (‘from the wood ’), further noting that the word ‘wood’ (õò) in our verse appears ear- lier in the verse to refer to the tree being chopped. Thus, the interpretation of õò as the handle in its second appearance would assume a change in the meaning of õò within a single verse. The second argument suggests that if Scripture wanted to refer to the handle it would have said ‘its wood’. The fact that it does not do so means that it does not refer to the handle, but to the wood that is being chopped. The second part of the talmudic passage presents an amoraic suggestion that the views of Rabbi and the sages rely on different understandings of the verb ìùðå,‘fly off’. This amoraic explanation accords well with the Tosefta, in line with Rashi’s interpretation of the connection between the Mishnah and the Tosefta, since, according to this explanation, Rabbi reads the verb ìùð,‘fell off ’, in the ìòéô conjugation: ‘made the splinter fly’. The tannaitic part of the talmudic passage, in contrast, is in line more with the view of Maimonides. There is no mention of a splinter; Rabbi’s view is explicated with reference only to the metal, the axe, and the wood that the metal flies from—whether handle or tree.17 Nonetheless, I would like to offer a new understanding of the relationship between the sources we have been discussing. As mentioned above, the Sifre supplies us with the biblical interpretive disagreement behind the halakhic formulation of the Mishnah. The Tosefta in this case is acting as its name

17 An additional PT passage (Makkoth 31c) similarly suggests that the word ìùðå, fell off,is the basis of both opinions in the Mishnah. The passage is so unclear, however, that Lieberman claims it supports Maimonides’ reading, while Dünner sees it as following Rashi’s. THE UNINTENTIONAL KILLER 37 might suggest—it adds to the information in the Mishnah. In contrast to scholars who attempt to create a link between the opinion of Rabbi Yehuda in the Tosefta and the sages in the Mishnah, I would posit that the Tosefta does not refer at all to the opinion of the sages in the Mishnah. In the Mishnah, the sages understood the biblical words ‘flew off from the wood’ as refer- ring to the metal head of the axe flying off from the wooden handle. The wood is the wood that chops, namely, the axe’s handle. However, the two opinions in the Tosefta deal with a case in which the wood is understood as the wood that is being chopped. This can clearly be seen from the phrasing of the Tosefta: ¨¡© ò÷áúîä õòä ïî úÌòKÙa äæúéð,‘If a splinter flies from the wood that is being chopped; ¨¡© ò÷áúîä õòä ïî ìæøáä ìùð, If the metal flies from the [blow to the] wood that is being chopped ’. As opposed to the Mish- nah, where one opinion interprets the wood as the axe handle itself, in the Tosefta both opinions interpret the word ‘wood’ in the verse as ‘the wood that is being chopped’ or the tree. The Tosefta’s repeated phrase ¨¡© ò÷áúîä õòä, ‘the wood that is being chopped ’, suggests that the toseftan traditions consis- tently interpret ‘wood’ (õò) as ‘wood that is being chopped’. However, within the ‘wood that is being chopped’ school of interpretation there is room for two possible views. According to one, what flies off from the wood that is be- ing chopped is the axehead. According to the other, what flies off from the wood that is being chopped is a splinter. Now we can resolve the apparent contradiction between the Mishnah and the Tosefta. The Mishnah draws a first-order distinction between those who understand ‘wood’ in Deut. 19:15 as the axe handle (the sages) and those who understand ‘wood’ as the wood that is being chopped, or tree (Rabbi). The Tosefta, however, makes a second-order distinction among those who favour the ‘wood that is being chopped’ view. The question then becomes what flies off? Some in the ‘wood that is being chopped’ school maintain that it is the metal handle (R. Yehudah) while others maintain that it is a splinter that flies off (the sages). The following diagram summarises the different views on the necessity of punishment by exile in the Mishnah, Sifre, and Tosefta, and the biblical exe-

gesis that underlies these different views:

®íéîëçÐò÷áîä õòä ïî

º

®éáøÐúò÷á äæúð ò÷áúîä º ®äãåäé éáøÐìæøáä ìùð 38 JOURNAL OF JEWISH STUDIES

From  the chopping wood (metal head flew from the handle the of the axe)—The sages in the Mishnah

wood ²

the wood that is  a splinter—Rabbi

being chopped ² The metal head itself (flew from the blow to the wood)—Rabbi Yehuda

In conclusion, according to my explanation, the sages in the Mishnah and the first opinion in the Sifre understood the verse as describing a case in which the metal flew off the chopping wood, i.e. the metal head flew from the handle of the axe, and killed someone. By contrast, R. Yehudaand Rabbi understood the verse as describing a case in which the wood itself is the reason for the deadly accident. But they differ on the exact nature of the accident: either the metal axe head flew off as a result of the blow to the tree (R. Yehudah), or a splinter flew from the wood that is being chopped while the metal head was chopping it (Rabbi). The conclusion of Epstein and De Vries should thus be rejected: the anonymous sages in the Mishnah are not equal to R. Yehuda. On the contrary, they differ in their basic reading of the verse.

3. The Order of the Halakhot in the Mishnah The connection between the halakhot in the Mishnah and biblical verses is apparent in other ways as well. For example, the order that halakhot are pre- sented in is often influenced by the order of elements in the biblical text, as can be seen in the following mishnayot.18 M.Makkoth 2:2 reads: íàå äìåâ íùì ñðëéì ÷æéðì úåùø éù íà âøäå åøéöçì ïáàä úà ÷øæ ÷æéðì úåùø øòéä äî øòéá åäòø úà àáé øùàå §ðù äìåâ åðéà åàì ñðëéì ÷æéðì úåùø ïéàù úéáä ìòá øöç àöé íùì ñðëäìå ÷éæîìå úà äëîä áàä àöé úåùø íéöò úáèç äî þøîåàü ìåàù àáà ®íùì ïéã úéá çåìùå åãéîìúá äãåøä áøäå åðá

[A man] threw a stone into his own courtyard and killed another: if he that was injured had the right to enter thither, the other must escape into exile; but if he had not the right, the other need not escape into exile; for it is written, ‘As when a man goeth into the forest with his neighbour’. Since the forest is a place into which the injurer and the injured have the right to enter. This excludes [from the law of unwitting murder] the courtyard of a householder into which the injured had not the right to enter. Abba Saul [says], ‘Just as the chopping of wood is an act of free choice [the law of unwitting murder applies to every act of free choice]—this excludes the father that smites his son or the teacher that chastises his pupils or

18 I am thankful to Professor Menachem Kahana for pointing this out to me. THE UNINTENTIONAL KILLER 39

the agent of the court. The first opinion is explicitly based on Deut. 19:5: øòéá åäòø úà àáé øùàå, ‘As when a man goeth into the forest with his neighbour’. The opinion of Abba Saul does not contain an explicit biblical quotation, but the phrase úáéèç äî íéöò clearly assumes the biblical phrase íéöò áåèçì, ‘to chop wood’. A par- allel in the Sifre makes this explicit: ÷éæîìå ÷æéðì úåùø øòé äî þä ¬èé §áãü øòéá åäòø úà àáé øùàå ñðëéäì ÷éæîìå ÷æéðì úåùø ïéàù úéáä ìòá øöç ÙzàÌöéü íùì ñðëäì áà àöé úåùø íéöò úáéèç äî øîåà ìåàù äáà íéöò áåèçì þíùì ïéã úéá çåìùå åãéîìú úà äãåøä áøäå åðá úà äëîä

‘As when a man goeth into the forest with his neighbour’. Just as both the one who was injured and the one who has inflicted the injury have the right to enter there [so does this apply to any other place where both of them had the right to enter]19 [thus excluding the injurer’s private courtyard which he had the right to enter while the injuree had not].20 ‘To chop wood’. Abba Saul says, ‘Just as chopping wood is a voluntary action [so does this apply to any other voluntary action]21 excluding a father who smites his son or the teacher that chastises his pupil or the agent of the court.’ The order of the halakhot in the mishnah, therefore, is based on the order of specific elements in the verse, as fleshed out in the halakhic midrash. The edi- tor of the Mishnah, wishing to shape his treatise as an independent halakhic composition, left out the biblical quote íéöò áåèçì at the beginning of Abba Saul’s saying. Instead, he settled for a contextual connection between the two consecutive halakhot, both dealing with úåùø (‘permission’) and the excep- tions to the rule: úéáä ìòá øöç àöé ® ® ® ÷éæîìå ÷æéðì úåùø ùé and úáéèç §åëå áàä àöé úåùø íéöò. In a similar way one can explain the order of the halakhot in m.Makkoth 2:5: ïéøñåîå ®§âå úùìùå êøãä êì ïéëú §ðù åæì åæî íéëøã ïäì úåðååëîå øîåà øéàî §ø åéìà åøáãéå êøãá åðâøäé àîù íéîëç éãéîìú éðù åì çöåøä øáã äæå §ðù åîöò éãé ìò øáãî àåä

And roads were made ready from one to the other as it is written, ‘Thou shalt pre- pare thee the way and divide [the borders of thy land].’ And they used to appoint for him two disciples of the sages lest [the avenger of blood] slays him on the way, that they may speak unto him. R. Meir says, ‘even he himself may speak unto him for it is written, “This is the word of the murderer”.’

19 These clarifying words, íùì ñðëéì ÷éæîìå ÷æéðì úåùø åì ùéù ìë óà, appear only in marginal additions (Oxford Bodl. 150 Uri 119; Berlin OR. 4 1594 33) and later versions (first printing, Venice, 1546, Yalkut Talmud Tora Oxford 2638, etc.). 20 Missing in ms. Vatican Ebr. 32.3 due to homoioteleuton, but completed according to all other manuscripts. 21 Here is the same textual situation as in footnote 19. This suggests that the literary formula in the more original versions was only comprised of ®®®úàöé®®®äîrather than óà®®®äî ® ® ® úàöé ® ® ®. 40 JOURNAL OF JEWISH STUDIES

In this case, the two verses are explicitly mentioned in the Mishnah, the first taken from Deut. 19:3 and the second from the next verse, 19:4: »êéäìà äåäé êìéçðé øùà ¬êöøà ìåáâÐúà úùìùå ¬êøãä ¬êì ïéëú øÖà ºéçå äîÖñåðéÐ øùà ¬çöøä øáã äæå ®çöøÐìë äîÖñåðì ¬äéäå íùìù ìîúî åì àðùÐàì àåäå ¬úòãÐéìáá åäòøÐúà äëé

Build roads to them and divide into three parts the land the lord your God is giv- ing you as an inheritance, so that anyone who kills a man may flee there. This is the rule concerning the man who kills another and flees there to save his life—one who kills his neighbour unintentionally, without malice aforethought. The order of the halakhot in this Mishnah was thus influenced by the order of the biblical rules.22

4. äîù (Whither), m.Makkoth 2:7 Another example of a midrash can be found in m.Makkoth 2:7: åðéà âøäù ìåãâ ïäëå ìåãâ ïäë âøåää ìåãâ ïäë àìá åðéã øîâéð ïåîî úåãéòì àìå äåöî úåãòì àì àöåé åðéà íìåòì íùî àöåé ìàøùéÐàáöÐøù åìéôàå åì ïéëéøö ìàøùé åìéôà úåùôð úåãòì àìå íùå åúøéã éäú íù äîù §ðù íìåòì íùî àöåé åðéà äéåøö ïá áàåéë ®åúøåá÷ àäú íùå åúúéî àäú

One who was judged while no High priest was in office, or one who killed a high priest, or a high priest that killed, [all of these] will never leave [the city of refuge]. He may not come forth to bear testimony that is a religious duty or testi- mony in a non-capital case or testimony in a capital case. Even if Israel is in need of him, and even if he is the captain of the host of Israel as Joab ben Zeruiah, he may never go out from there, as it is said: ‘whither’—there shall be his dwelling place, there shall be his death, there shall be his burial. The triple halakhot at the end of the Mishnah, åúúéî àäú íùå åúøéã éäú íù åúøåá÷ àäú íùå, can be understood in light of the midrash in the Sifre:23 àäú íù åúãéøé àäú íù íéîòô §â äîù äîù äîù øîåì ãåîìú äî åúøåá÷ àäú íù åúúéî

Why does it say ‘whither, whither, whither’ three times: there will be his place of impoverishment,24 there will be his death place, there will be his burial place.

22 For recent work on this topic, see also A. Shemesh, ‘The Scriptural Background of the Penal Code in the Rule of the Community and Damascus Document’, DSD 15 (2008), pp. 191–224 (= Aharon Shemesh, ‘The Penal Code from Qumran and Early Midrash’, Meghillot 5–6 (2008), pp. 245–268; in Hebrew). Shemesh argues that the order of laws in the Penal Code follows that of the scriptural passages upon which it is based. 23 As well as the parallel in the Tosefta and the Mechilta de-Rashbi. 24 This version is unique and only found in the best Sifre manuscripts (åúãéøé:VaticanEbr. 32.3; Oxford Bodl. 150 Uri 119 (corrected in margin to åúøéã); åúãéø: Berlin OR. 4 1594 33 (cor- rected in margin to åúøéã); and London BL Add. 16.406). All other Sifre versions, as well as all textual versions of this exegesis in the parallels in the Mishnah, Tosefta, and Mekhilta de-Rashbi THE UNINTENTIONAL KILLER 41

It becomes clear, therefore, that the halakha in the Mishnah is in fact based on a midrash that relies on the three occurrences of the word äîù, whither, in the verse. However, the midrash itself is problematic. Where does the word äîù appear three times? In Numbers 35 it appears five times (verses 6, 11, 15, 25, 26) and in Deut. 19 it appears twice (3, 4). According to the printed editions of the Mishnah, as well some of the Sifre’s commentators such as Rabbenu Hillel25 and R. David Pardo,26 the Sifre’s tradition was understood to refer to the verses in the book of Numbers. But in manuscripts the word äîù is quoted without its biblical context, so another explanation can be offered. In contrast, the marginal comments in the Berlin manuscript27 list the two Deuteronomy verses (19:3, 4) and add verse 19:12 where we have åúåà åç÷ìå íùî,‘they took him from there’, which of course is not the same word as äîù. Perhaps the midrash relies on a different biblical version, one that had three occurrences of the word äîù in the Deuteronomy verses, but such a version is not documented in any of the manuscripts we currently have. In its favour, this latter explanation at least recognises that, as previously mentioned, Sifre Deuteronomy deals almost exclusively with verses from the book of Deuteronomy. Thus I agree that the midrash here, which is the basis for the Mishnah’s tradition, relies on the three occurrences of the word in the book of Deuteronomy. However, I submit that the third occurrence of äîù appears in Deut. 4:42, not 19:12: çöåø äîù ñåðì ® ® ® íéøò ùìù äùî ìéãáé æà

Then Moses set aside three cities . . . for a killer to flee whither. This claim is further supported by other passages in the Sifre that demonstrate that the Midrash views the verses dealing with cities of refuge in Deut. 4 and Deut. 19 as one unit.

5. The Number of Cities of Refuge In regard to the question of the number of cities of refuge, the Sifre treats these verses as part of a single, organic literary unit. While the verses in Numbers 35, Joshua 20, and Deut. 19 tell of six cities of refuge, Deut. 4:41–43 mentions only three.28 The Mishnah (2:4) has a final count of six cities of refuge, and does not mention additional cities:

read åúøéã. It is possible that the Sifre here is preserving the original version of the exegesis. If taking åúãéøé to mean the impoverishment of the murderer (see Even Shoshan’s dictionary, entry äãéøé), the version åúãéøé matches the context better. The Sifre version lists three negative things, in a descending order, that happen to the murderer in the city of refuge. The unique use of åúãéøé, without the usual åéñëðî ãøé (stemming from the succinct literary formula of the exegesis, a one- word consequence for each occurrence of the word äîù), and an early textual corruption can easily explain the change åúøéã Ð åúãéø Ð åúãéøé. 25 Hillel ben Eliakim, de-ve Rav,K. oliditsk. i, Shakhna ben H. ayim (ed.), Jerusalem, 1961. 26 David Samuel ben Jacob Pardo, Sefer Sifre de-ve Rav, Salonika, 1799. 27 Ms. Berlin Or. 4 1594 33. 28 Exodus 21:13 does not refer to the number of the cities. 42 JOURNAL OF JEWISH STUDIES

ïòðë õøàáù ùìùìå ïãøéä øáòáù ùåìùì èì÷î øéòì ïéìåâ ïëéàì åðúú íéøòä ùìù úàåü ïãøéì øáòî åðúú íéøòä ùåìù úà §ðù ùåìù åéä àì ìàøùé õøàáù ùìù åøçáéð àìù ãò §åâå þïòðë õøàá úåèìå÷ ïúùù åäéù äðééäú èì÷î éøò ùù §ðù úåèìå÷ ïãøéä øáòáù úçàá

Whither may they go into exile, to a city of refuge? To the three that are in the Transjordan or to the three that are in the land of Canaan, as it is written, ‘Ye shall give three cities beyond the Jordan [River] [and three cities shall ye give in the land of Canaan].’ The three cities in TransJordan granted no right of asylum until the three were chosen in the land of Israel, for it is written, ‘There shall be for you six cities of refuge’—meaning that all six of them together grant a right of asylum. However Sifre Deuteronomy (185), which quotes the Tosefta (3:10) as well as the midrashim of the school of Rabbi Ishmael (Sifre Num. 160 and Mekhilta Deut. [Midrash Tannaim 19:9]), counts nine cities of refuge in some way or other,29 adding the three cities in the verse in Deut. 4 to the six cities men- tionedinDeut.19: §â §îåà äúà ïëéî §åâå úôñéå §åâå úàæä äååöîä ìë úà øåîùú éë ùåìù ãåò ùéøôä õøàì åàáùëå ïãøéä øáòá äùî ùéøôä íéøò òùú éøä §â ãåòå ùù éøä §â ìò §â ®§â ãåò ïéùéøôî àåáì ãéúòìå

‘If thou shalt keep all this commandment . . . then shalt thou add [three cities more for thee, besides these three] (Deut. 19:9).’ Hence you conclude that Moses set aside three cities in the TransJordan, and after they came to the land [of Israel] he set aside three more, and in the future they will set aside an additional three. Three and three make six and three more makes nine. We can approach the äîù midrash in the Mishnah in light of this tradition, which regards the verses in Deut. 4 and 19 as a single literary unit in con- nection with the number of the cities of refuge, adding the three mentioned in Deut. 4 to the six in Deut. 19. The midrash in the Sifre and the Mishnah which lists three occurrences of the word äîù is treating, in a similar way, the verses in Deut. 4 and 19 as a whole unit, adding the occurrence in 4:42 to the two found in Deut. 19:3 and 19:4. If this is correct, our study of the midrashic layer in the Mishnah reveals an interesting phenomenon. The Mishnah in one case (2:2), when dealing with the number of cities of refuge, treats the verses in Deut. 4 as a repetition of what is described in Deut. 19 and does not add the three mentioned there to the six in Deut. 19. On the other hand, the midrash in Mishnah Makkoth

29 The Tosefta quoted in the Sifre later brings additional opinions that expand the number of cities further to twelve and even fifteen. In light of this complex textual problem, Sifre Zuta Numbers 35 seems to reach the conclusion that there is actually no specific fixed number of cities that should be set aside, and one can add cities as needed. The Mekhilta Deuteronomy,onthe other hand, in two passages (Midrash Tanna im 19:7, 9) seems to polemicise with this expansive approach, and stresses the prohibition to add more cities of refuge, For a source critical approach to this biblical issue, see ÷ìç ¬íéøáã øôñì àåáî ºêåúá ¬§èì÷îä éøò ìù ïäéúåãìåúì§ ¬àôåø ®à ç¢îùú íéìùåøé ¬êùîä é÷øôå ïåùàø. THE UNINTENTIONAL KILLER 43

2:7 adds up the appearances of the words in both paragraphs as if each rep- resented another future stipulation for the unintentional killer, based on the assumption that the two paragraphs are a literary unit. The Deut. 4 passage is not viewed as a parallel passage to Deut. 19, but rather as the beginning of the issue continued in Deuteronomy 19.30 This finding is further confirmation of the diverse midrashic sources found in the Mishnah, and the different in- terpretive approaches to the biblical verses utilised by these sources and taken up by the Mishnah.

6. ‘Divide the borders of thy land’: m.Makkoth 2:5 At times one can demonstrate the lateness of the midrash in the Sifre in rela- tion to that apparent in the Mishnah. Such is the case in m. Makkoth 2:5: §åâå ¢úùìùå êøãä êì ïéëú¢ §ðù åæì åæî íéëøã ïäì úåðååëîå

And roads were made ready from one to the other as it is written, ‘Thou shalt prepare thee the way and divide the borders of thy land into three’. (Deut. 19:3) The Sifre (180) on this verse reads: ¬äëåúì úåùøåôî åäéù úåàèøèñ êì ïéëú ¢êøãä êì ïéëú¢ úåðååëî åéäé àìà úåøæåôî åéäé àìù ¢êöøà ìåáâ úà úùìùå¢ íøëáù úåøåù éúùë

‘Thou shalt prepare thee the way.’ Prepare highways that lead directly into [the city of refuge]. ‘And divide the borders of thy land into three parts.’ They must not be scattered; rather they must be straight like that between two rows [of vines] in a vineyard. The connection between the two midrashim is evident in the treatment of the issue of creating roads and the common use of the word úåðååëî.However, the mishnah is presented as one exegesis on the entire verse, while the Sifre has two different exegeses on two different parts of the verse. In the Sifre the word úåðååëî most likely refers to the cities themselves, and in the Mishnah it refers to the roads. The midrash in the Mishnah is a difficult one. Why is there a need to create roads between the cities of refuge? Sifre Deut. (180, 183) specifically claims that a murderer does not flee from one city to the other, but rather, once he

30 If the Mishnah in m. Makkot 2:7 regarded Deut. 4 and 19 as parallel passages and not as a literary unit, then the number of äîù occurrences treated in the midrash should have included the five occurrences in the Numbers 35 passage as well. 44 JOURNAL OF JEWISH STUDIES reaches one city of refuge he should be considered safe.31 Albeck32 attempts to solve this question by correcting the Mishnah to say åæì åæå instead of åæî åæì, meaning that both the roads and the cities are úåðååëî,‘aligned ’, and not that the roads are connected to each other. Alternatively, he suggests that the original mishnah did not have the words åæì åæî at all, in line with the citation of this mishnah in the later Deuteronomy Rabbah (ad Deut. 4:42). I would like to suggest another explanation for this difficult mishnah. Read- ing this mishnah sequentially with the previous mishnah (2:4) reveals another referent for the words åæì åæî: ïòðë õøàáù ùìùìå ïãøéä øáòáù ùåìùì èì÷î øéòì ïéìåâ ïëéàì øáòáù ùåìù åéä àì ìàøùé õøàáù ùìù åøçáéð àìù ãò ® ® ® êøãä êì ïéëú §ðù åæì åæî íéëøã ïäì úåðååëîå®®®úåèìå÷ïãøéä §âå úùìùå

Whither may they go into exile, to a city of refuge? To the three that are in the Transjordan or to the three that are in the land of Canaan . . . The three cities in the Transjordan granted no right of asylum until the three were chosen in the land of Israel . . . And roads were made ready from one to the other as it is written, ‘Thou shalt prepare thee the way and divide the borders of thy land into three.’ When read in this way, the sentence in m. Makkoth 2:5 states that roads should be set åæì åæî—between the previously mentioned three cities in Tran- sjordan and the three cities in the land of Israel. The singular demonstra- tive form åæ referring to a plural noun íéøò (‘the cities’) does not raise any grammatical problems, since the demonstratives here are used as reciprocal pronouns.33 In such constructions, it is a cross-linguistical phenomenon that singular terms can refer to plural entities (for example, in English one can say ‘they all kissed one another’ even when there are more than two persons involved). The reason for setting the roads between the cities is specified in the rest of the verse: úùìùå, ‘divide in three’. Laying roads from the cities in the East to the cities in the West would create a division of the land of Israel into three parts, and this is the reason for the biblical reference to a tripartite division. A similar desire to use the cities as reference points for a tripartite division of

31 Albeck, following Rabenu Hillel, R. Eliezer Nah. um (Perush Sifre, M. Kahana, ed., Jerusalem 1992), R. Suliman Uhana (Hagahot ve-he arot al Mekhilta, Sifra, ve-Sifre, JTS manuscript MS 9519), and the Pseudo-Rabad (Commentary to Sifre Deuteronomy, H. W. Basser, ed., Atlanta 1994), argued that these passages in the Sifre refer to the case when the murder is committed in a city of refuge. They thus read the Sifre in light of m.Makkoth 2:7. However, I tend to agree here with Ish Shalom’s reading (M. Friedmann, Sifre debe Rav im Tosefot Me ir ayin, Jerusalem, 1967). The Sifre does not mention a murder taking place in the city, but just states in a general way that the city of refuge should be the last stop in a man’s flight, and that he should not be forced to run to another city. 32 In the supplements to his Mishnah commentary. 33 See for example m. 11:5: åæì åæî íéìèìèî åæá åæ úåøåù÷ úåðéôñ (‘if ships are tied together, one may carry from one to another’); this halakha is by no means restricted to only two ships. For the use of the repeated demonstratives åæ¯äæ as reciprocal pronouns in , see M. H. Segal, A Grammar of Mishnaic Hebrew, Oxford 1927, p. 208. I thank Elitzur Bar-Asher Siegal for discussing this issue with me. THE UNINTENTIONAL KILLER 45 the land is also present in t. Makkot 3:3: íëùì ïåøáçîë íåøãì ïåøáçî àäéù úåùìåùî åäéù úùìùå ùã÷ì íëùîë íëùì ïåøáçîå

‘And you will divide’. That they be divided into three parts. Let the part from He- bron to the south be as from Hebron to Shekhem and from Hebron to Shekhem as from Shekhem to Qadesh. The Tosefta refers to the equal distance between the three western cities as the fulfillment of the verses’ request to split the land into three parts. The Tosefta does not deal with the Transjordanian cities, which were probably not so equally set apart.34 But the main point is similar—the locations of the cities of refuge are a fulfillment of the verse’s úùìùå. In my reading of the Mishnah, the roads laid between the cities in the East and the cities in the West are what create the division, and not the cities themselves. Supporting this reading of the Mishnah is the biblical quote in the Mishnah itself: §âå úùìùå êøãä êì ïéëú. Were the Mishnah dealing only with laying roads into the cities or from one city to the other there would have been no need to quote the rest of the verse—úùìùå. According to my reading, the úùìùå element is crucial, as the roads are what create the division of the land into three parts. If my reading is correct, then the relationship between the Mishnah and the Sifre on these verses should be re-examined. The Mishnah applies the entire verse to the roads (úùìùå êøãä êì ïéëú) as if the second clause describes the manner in which the first requirement should be fulfilled, meaning that setting roads (between the eastern cities and the western cities) will divide the land into three parts. But the Sifre has two different exegeses: êøãä êì ïéëú, ‘you shall set roads into the cities’, and êöøà ìåáâ úà úùìùå, ‘[the cities] must be straight like two rows in a vineyard’. Of course, the second passage in the Sifre is a difficult one, since there does not seem to be a clear connection between the verse’s úùìùå and the alignment of the cities. A parallel Tosefta passage has the same exegesis (t.Makkoth 2:2) but it is not connected to this part of the verse (úùìùå) and lacks the Sifre’s seemingly unnecessary àìù ® ® ® àìà úåøæåôî åéäé, ‘theymust notbescattered;rather...’. DeVries35 suggested that the Sifre’s exegesis originally dealt with the word ïéëú and this is why it uses the adjective úåðååëî. While I agree, I would add that the original midrash was the one in the Mishnah, where it is clear that the mishnaic verb úåðååëî is based on the biblical verb ïéëú, both sharing the same root ï¢åë. According to my proposal, the problems in the Sifre’s exegesis, as well as the common use of the adjective úåðååëî found in both the Mishnah and the Tosefta, suggest that the midrash in the Sifre is a re-worked version of the midrash in the Mishnah. The more difficult midrash in the Mishnah, in this case, should enjoy the status of lectio difficilior potior. According to this,

34 See Yohanan Aharoni, At.las Kart.a li-tek. ufat ha-Mik. ra, Jerusalem, 1995, p. 73. If we follow the scholarly identification of the Transjordanian cities, we must accept that the division created by roads set between the western and eastern cities does not create three equal parts. 35 Ibid. p. 106. 46 JOURNAL OF JEWISH STUDIES the Sifre no longer understood the original midrash in the Mishnah, which alluded to laying roads from the three Transjordanian cities to the western cities. When understood, as previously discussed, as a statement in favour of laying roads, in general, between the refuge cities, the midrash was perceived as problematic. Therefore, the Sifre split the midrash into two parts, creating two new exegeses. The first re-worked the Mishnah, so that it no longer meant laying roads from one city to the other. The new exegesis now only states the need to lay roads into the city. The second exegesis was created by using (1) the exegesis also found in the Tosefta on the neat alignment of cities, together with (2) the úåðååëî, ‘aligned’, from the Mishnah, and (3) with the addition of ® ® ® àìà úåøæåôî åéäé àìù, ‘theymustnotbescattered;rather...’.The result is a midrash that is not strongly related to the verse it is interpreting. In conclusion, if my reading here of the Mishnah is correct, then this is a case where the comparison between the midrashic layer in the Mishnah and the midrash in the Sifre proves the earlier date of the midrash in the Mishnah. Of course, no general conclusions can be deduced here regarding the overall lateness of the Sifre in relation to the Mishnah. The complexity revealed in this short study only indicates that each example should be examined sepa- rately. While in this case the midrash in the Mishnah proves to be earlier than the version in the Sifre, in other cases, as shown above, the Sifre supplies us with the midrash that underlies the later Mishnaic ruling. Thus, one should be wary of broad conclusions concerning the literary corpora in which these materials are located. Rather, our conclusions are limited to local, specific materials that are found within these broader compositions.

7. Summary I chose to discuss a few cases from the second chapter of Mishnah Makkoth in which a covert midrashic layer seems to serve as the basis for the halakhot of the Mishnah. This midrashic layer is evident when comparing the Mish- nah, which is mostly based on the traditions studied in the school of Rabbi Akiva, with the Sifre Deuteronomy, the legal midrash which originated in the Rabbi Akiva midrashic school. I do not intend to claim here that the midrash found in the Sifre is the same midrash that stood as the basis for what we find in the Mishnah. On the contrary, in some cases we can demonstrate a different midrashic stratum in the two sources. One such example is the above mentioned ruling regarding the number of cities of refuge, which is six in the Mishnah and nine to fifteen in the Sifre. In other cases one can demonstrate the lateness of the midrash in the Sifre in relation to the midrash apparent in the Mishnah, as seen from the last example. However, this early midrashic layer in the Mishnah is often close to the midrash in the Sifre and sheds light on the development of rabbinic sources in the different literary collections, and this is why the comparison between the two corpora is extremely benefi- cial at times. Finally, the comparison of the early midrashic layer in the Mishnah and the midrashic material found in the legal midrashim, raises again the ques- THE UNINTENTIONAL KILLER 47 tion of the various sources of Mishnah Sanhedrin-Makkoth in particular and of Rabbi’s Mishnah in general. The simplistic, scholarly question of chrono- logical priority of Mishnah or Midrash, and associated assumptions about the separate nature of halakha in these works, must take into account stud- ies that demonstrate a more complex relationship in which the two often in- tersect. This is especially true in tractates such as Sanhedrin, in which the midrashic layer is more perceptible. Much more remains to be done using this methodological tool.