Four Ways to Derive the Thirty-Nine Avot Melakhot

The Torah never defines specifically what, melakha, labor, on Shabbat entails, but the Mishnah already has an exact list of 39 categories of labor prohibited on Shabbat.^[1] A comparison of the structure of the Yerushalmi and Bavli sugyot highlight the different ways the Amoraim conceptualized melakha in contrast to the Mishnah.^[2] 

Rabbi Dr. Richard Hidary

 by anshie.com

How Does the Torah Define Melacha?

In the Ten Commandments, the Torah issues a general prohibition: “The seventh day is a Sabbath of the Lord your God, you shall not do any work (melakha),”^[3] on penalty of death.^[4] The Torah, however, does not systematically define the parameters of such work other than providing a few examples including:

• gathering and preparing food,^[5]
• gathering fuel,^[6]
• burning a fire,^[7]
• doing agricultural work,^[8]
• and working a slave or an animal.^[9]

The Prophets further mention performing business transactions^[10] and carrying wares to the gates for selling.^[11] The sages and communities of later generations were left with the responsibility to define more precisely the parameters of Sabbath observance. Mishnah Hagigah 1:8 appropriately comments that “the laws of Shabbat…are like mountains (or “desert plants”)^[12] hanging from a strand.”^[13]

The Thirty-Nine Melakhot in the Mishnah

Rabbinic sources provide a systematic delineation of prohibited actions at various levels of liability—the highest and most important being the thirty-nine avot (principle labors; singular av).^[14] Mishnah Shabbat 7:2 lists these thirty-nine principle labors, which include activities relating to agriculture and food preparation, clothing manufacturing, preparation of parchment and writing, building, kindling a fire, and carrying. However, the Mishnah provides no source for its list of activities, nor does it explain how it arrived at the number thirty-nine.

Gilat’s Suggestion: A Standard List of Work
Yitzhak Gilat demonstrates that the list of avot belongs to a secondary stratum of the Mishnah and all Tannaim did not agree with it.^[15] He further notes that there is nothing in this list to indicate any relationship with the work of the mishkan (Tabernacle) as the Talmud claims; in fact, many of the activities listed in Mishnah Shabbat 7:2 bear little connection with any equivalents in the mishkan.^[16] Rather, he argues that this list is derived from standard lists of work activities, such as the one found at Tosefta Berakhot 6:2, where Ben Zoma details all the activities that Adam had to do simply to eat bread: “He plowed, harvested, sheaved, thrashed, winnowed, sifted, ground, sieved, kneaded, and baked.”^[17] The Talmud, in fact, designates this group of avot, “the process of making bread.”^[18]

This conception of melacha as the processes involved in the manufacturing of basic goods has its basis in the Torah, which prohibits field activities, food preparation and fire. The Mishnah simply specifies what these activities are—in part based on common practice, prior tradition, and its own innovation.^[19] The Mishnah’s enumeration may have no further textual proof than this. The number thirty-nine does not seem to have any symbolic significance; it may simply be the sum total of items in this grouping of lists. Certainly, the form of the number as “forty-less-one” rather than “thirty-nine” is simply the typical style of rabbinic Hebrew.^[20]

The 39 Melachot in Other Rabbinic Texts

Unlike the Mishnah, which rarely cites sources, other rabbinic texts make various attempts to derive the number thirty-nine and the specific contents of the avot from the Torah. We can find four major strategies in the Talmud for this derivation, each with its own history and development.

I – Derivation from Shabbat’s Juxtaposition with the Mishkan: Exodus 35:1
Exodus 35 begins Moses’ instructions to the Israelites about how to build the mishkan. As a prelude to the mishkan instructions, verses 2-3 mention the prohibition of work on Shabbat. Mekhilta d’R. Ishmael (Vayakhel, Shabata 1) derives from this juxtaposition that the building of the mishkan must pause for Shabbat. The Mekhilta continues:

He [Moses] told them, ‘These are the utterances’ (Exodus 35:1) – Rabbi said: “This comes to include the thirty-nine principle labors [prohibited on Shabbat] that Moses told them orally.”^[21]

The midrash infers that, “the utterances,” in verse 1 also relate to the Sabbath and its plural form refers to an oral instruction given by Moses regarding the thirty-nine avot. There is no attempt here to derive the specific number or contents of the avot from the wording; rather, this verse alludes to an oral tradition given by Moses delineating the avot.^[22]

While the Mekhilta simply accepts the thirty-nine avot as a part of ancient oral law, the Talmudim make various attempts to derive the number itself from this verse.

• Yerushalmi Shabbat 7, 2 (9b) includes two explanations based on gematria .^[23]
  • Hanina from Sepphoris in the name of R. Abahu counts: אלה=36, “utterance” adds one, and the plural form, “the utterances,” adds two more for a total of 39.
  • The sages of Caesarea instead allow a ח to replace the ה of אלה for a total of 39.

• Bavli Shabbat 70a seems to include the first derivation—that of the amora R. Hanina—within a baraita (i.e., a Tannaitic source) of R. Natan; however, it does not make the gematria explicit: “דברים, הדברים – אלה הדברים – אלו שלשים ותשע מלאכות שנאמרו למשה בסיני.”

• Bavli Shabbat 97b also repeats the same wording but as a baraita in the name of Rabbi, the same authority mentioned in the Mekhilta. It is entirely possible that the Bavli texts merely repeat in slightly different wording the sense of Rabbi’s midrash in the Mekhilta that included no gematria.^[24] If, however, the Bavli means to imply the gematria of R. Abahu in its formulation “דברים, הדברים, אלה הדברים,” then it is interestingly inserting this gematria back into Rabbi’s original midrash that made no effort to derive the exact number.

II – The Number of Vessels in the Mishkan
Midrash ha-Gadol^[25] makes a stronger connection between the mishkan verses and the avot by presenting the thirty-nine mishkan vessels listed at Exodus 35:11-19 as the source for the same number of avot. This source, like the previous one, derives only the number of avot from the mishkan but not the specific types of labor. In other words, the principle Shabbat labors bear no necessary resemblance to the list of vessels of the mishkan; but at least we have the number thirty-nine. Rabbi Bin Nun impressively came upon this midrashic derivation independently and even added a significant insight by finding thirty-nine vessels listed in Exodus 39:33-41 as well.^[26]

III – Counting the Number of Words Related to Work
Yerushalmi Shabbat 7, 2 (9b) quotes R. Shmuel bar Nahman in the name of R. Yohanan who claims that the thirty-nine avot correspond to forty-less-one times the word “melakha” is written in the Torah. Bavli Shabbat 49b similarly cites in the name of R. Shimon b’R. Yose ben Lakonia that the avot correspond to the instances of the word “melakha”^[27] in the Torah (section C in the chart below).

Despite slight differences in the names of the tradents and the formulation, the Yerushalmi and Bavli sugyot show an overall affinity. Both the Yerushalmi and the Bavli question whether Genesis 39:11 and another verse (Genesis 2:2 or Exodus 36:7) should be included in this count. In both sugyot, an amora asks a question, another amora rejects the basis of the question since the answer can be easily checked in a Torah, and the question is then reformulated and left unanswered (section D). This overlapping structure suggests that the sugyot in each Talmud share a common skeletal origin.

The idea to base the thirty-nine avot on the number of occurrences of the word melakha may have arisen from activity of the early transmitters of the biblical text. Y. Shekalim 5, 1 (48c) states that the Soferim (lit. counters), “made the Torah into many enumerations…avot melakhot are forty-less-one.” As pointed out by Rabbi Hoffman, these Soferim are also mentioned in B. Kiddushin 30a as those who count the number of letters, words, and verses in the Bible. These may have been predecessors of the Masoretes who also recorded such counts as well as counting occurrences of word spellings throughout the Bible.^[28] As they were counting occurrences of words, they also connected these counts with relevant halakhot.

Be that as it may, this count is very difficult to calculate since various permutations of the work melakha occur well over thirty-nine times in the Torah. Many commentators ancient and modern have discussed this problem.^[29] Most recently, Rav Yoel Bin Nun has brilliantly offered a possible new solution by counting only “melakha” and “melakhto” ^[30] but not “melekhet,” yielding forty instances, in which case the ensuing discussion about how to count individual verses makes sense as an attempt to remove one to get thirty-nine.

The Yerushalmi (section E) includes a variant of this method:

“R. Yose b. R. Bun in the name of R. Shmuel bar Nahmani: [The thirty-nine avot] correspond to forty-less-one occurrences of ‘`avodah’ and ‘melakha’ regarding the mishkan.”^[31]

As Gilat points out, this opinion actually combines aspects of this strategy with that of the next one, which focuses on work in the mishkan.^[32]

IV – Categories of Work Performed in the Mishkan
Very few Tannaitic sources connect the avot with the various specific types of work activity performed in the mishkan.^[33] Amoraic sources expand the relationship between the avot and the mishkan to a much greater extent.

Tannaitic Sources
To begin with Tannaitic sources:

• Mishnah Shabbat 11:2 derives the manner of transferring from one private domain to another over a public domain based on how the Levites would transfer planks of the mishkan from one wagon to another.

• Mishnah Shabbat 12:3 states that one who writes two letters on Shabbat is liable for the av of writing. R. Yose explains: “They made liable for writing one who makes two letters [as a minimum] because this is how they would mark the planks of the mishkan to know which one goes next to which.”

• Tosefta Shabbat 11:2 states: “Rabban Shimon ben Gamaliel says, ‘Even one who strikes a hammer against an anvil during work is liable for that is how they would beat metal sheets for the mishkan’”

These texts look to the manner in which a certain activity was done in the mishkan in order to define a parameter of the av; but they do not derive the av itself from a mishkan equivalent.

Amoraic Sources
The Yerushalmi states:

“All of the principle manners of work (avot) were learned from the mishkan. What plowing was done in the mishkan?…What dying was there in the mishkan?…What tying was there in the mishkan?…What tanning was done in the mishkan?… What erasing was done in the mishkan?…What building was there in the mishkan?”^[34]

The Yerushalmi only explicitly connects these few activities with the mishkan but it does begin with the broad statements that all of the avot derive from the mishkan.

By basing all of the avot on the mishkan, the Yerushalmi draws a much closer analogy between the two than the Tannaitic sources do. But the Bavli goes even further in emphasizing and building far-reaching definitions on the basis of this conceptualization.  This is reflected in its programmatic statement in the primary discussion of the source for the avot as well as in its conceptualization of the definition of melakha.

Let us compare the loci classici for the origin of the avot in both the Yerushalmi and Bavli. (Below is the English Text – see Appendix for Hebrew text.)  

Both sugyot open with the same question (section A). Sections C and D, as we analyzed above, are more or less equivalent between the two Talmuds and focus on counting the occurrences of melakha in the Torah. Section E, which counts words for work found in the mishkan context, appears only in the Yerushalmi. Section F appears in the Yerushalmi sugya and has no parallel in the corresponding Bavli sugya at 49b.

Most significantly, sections B and G appear only in the Bavli. R. Hanina (bar Hama) says that the avot correspond with the work performed in the mishkan and section G cites a baraita to back up this opinion. The Yerushalmi does include a similar statement in the continuation of the chapter, as cited above. However, I think it is significant that only the Bavli cites this opinion in this central sugya and that only the Bavli quotes this baraita.

Analyzing the overall structure of the Bavli sugya, we see that it omits any mention of the gematria derivation here, it cites the third strategy but leaves the question about that count unanswered, and it advocates decisively for the fourth strategy. It thus begins and concludes with the strategy that it clearly prefers. The Yerushalmi, on the other hand, does cite snippets of this strategy in the ensuing discussion but shows no preference for it and does not even mention it in this primary sugya.

The Bavli’s emphasis on the connection between the avot and the mishkan activity likely relates to its overall tendency towards greater conceptualization and systematization.^[37] The Bavli’s project of conceptualization also has far-reaching effects on defining conditions for liability. Tannaitic sources require that one have intention to perform the forbidden act in order to be liable to bring a sin offering.^[38] But only the Bavli expands this to also require that the act be work of craftsmanship. ^[39] This Bavli innovation may stem from its conception of avot as deriving from the work performed in assembling the mishkan.^[40]

Conclusion

In sum, Mishnah Shabbat 7:2 presents these lists of creative activities without any hint as to their source, and other Tannaitic texts make only slight reference to an analogy between these avot and the mishkan activities. Amoraic sources, on the other hand, present four different strategies for the derivation of the avot and of the number thirty-nine. These various “derivations” come after the legislation of the system of the avot; nevertheless, they help to define and refine the conceptualization of the avot over the generations of the halakha’s development. These laws, which were once likened to “mountains hanging from a strand,” have now been attached to the Torah with a fourfold cord.

 ___________________

Appendix – Hebrew Table of Comparison

___________________

Rabbi Dr. Richard Hidary is an associate professor of Judaic studies at Yeshiva University where he teaches courses in Bible, Second Temple Jewish history, Dead Sea Scrolls, Talmud, Midrash, and Jewish ethics. He is also the Distinguished Rabbinic Fellow at Congregation Shearith Israel. Richard studied at Yeshivat Har Etzion, received his ordination from the Chief Rabbinate of Israel and a PhD from NYU. His book, Dispute for the Sake of Heaven: Legal Pluralism in the Talmud, has been published by Brown University, and his articles appear in Encyclopedia Judaica, Encyclopedia of the Bible and Its Reception, AJS Review, Dine Israel, Okimta, and Conversations. Rabbi Hidary also runs the websites teachtorah.org, pizmonim.org, and rabbinics.org.

[1] In previous posts to TheTorah.com, my teacher Rabbi Dr. Yoel Bin Bun and Rabbi Evan Hoffman have offered learned analyses of the origins of thirty-nine principle labors (avot melakhot) that are prohibited on Shabbat. Rabbi Bin Nun, “The Textual Source for the 39 Melachot of Shabbat,” and Rabbi Hoffman, “40 Minus 1 Melachot of Shabbat.” My piece will be in conversation with theirs.

[2] I would like to thank the editors at TheTorah.com as well as Samuel Berkovitz for their helpful comments and corrections.

[3] וְי֙וֹם֙ הַשְּׁבִיעִ֔֜י שַׁבָּ֣֖ת׀ לַי-הֹוָ֣ה אֱ-לֹהֶ֑֔יךָ לֹֽ֣א־תַעֲשֶׂ֣֨ה כָל־מְלָאכָ֡֜ה

Exod. 20:10. All biblical translations are from NJPS.

[4] Ibid. 31:14 and 25:2.

[5] Exod. 16:22-29.

[6] Num. 15:32-36. Interpreters provide various explanations for the precise Sabbath violation involved in this pericope. Philo, Moses 2.220, writes that gathering wood connects to the act of making a fire and furthermore that one may not move wood. See further at Samuel Belkin, Philo and the Oral Law: The Philonic Interpretation of Biblical Law in Relation to the Palestinian Halakah (Cambridge: Harvard University Press, 1968), 198; and Lawrence Schiffman, The Halakhah at Qumran (Leiden: Brill, 1975), 118. Onkelos translates the prohibited act as binding, while Targum Pseudo-Jonathan accuses the sinner of uprooting trees. Similarly, B. Shabbat 96b records suggestions by various rabbis that the wood gatherer was guilty of detaching the branches from the trees, of carrying the wood in public domain, or of tying them up into bundles. See also Yitzhak D. Gilat, Studies in the Development of the Halakha (Jerusalem: Bar-Ilan University Press, 1992) (Hebrew), 33 n. 9; and idem, “Introduction: The Sabbath in the Bible,” in Yad la-Talmud: Selected Chapters, ed. E. E. Urbach (Jerusalem: 1984), 47; and the TABS essay, “What Melacha did the Wood-Gatherer Violate?”  

[7] Exod. 35:3.

[8] Exod. 34:21.

[9] Exod. 20:10, 23:12, Deut. 5:14.

[10] Amos 8:5 and Isa. 58:13.

[11] Jer. 17:21-22. Neh. 10:32 and 13:15-23 mentions also buying food and wares as well as pressing wine and loading animals. On the biblical Sabbath prohibitions, see further at Gilat, Studies, 33; and idem, “Introduction,” 45-51; Admiel Kosman, “On the History of the Category of ‘Ovadin DeChol’ and Yom-tov and its Relationship to the Category of ‘Shevut’ Prohibitions,” (Hebrew; PhD diss., Bar-Ilan University, 1993), 9-12; and Gerhard F. Hasel and W.G.C. Murdoch, “The Sabbath in the Prophetic and Historical Literature of the Old Testament,” in The Sabbath in Scripture and History, eds. Kenneth Albert Strand and Daniel Andre Augsburger, (USA, 1982), 44-56.

[12] This follows  ms. Parma; see the insightful analysis of Michal Bar-Asher Siegal, “Mountains Hanging by a Strand? Re-reading Mishnah Hagiga 1:8,” Journal of Ancient Judaism 4 (2013), 235-256.

[13] הלכות שבת חגיגות והמעילות הרי הם כהררים (כחררים) התלויין בשערה שהן מקרא מועט והלכות מרובות

[14] For the prohibitions in the category of shevut, see Richard Hidary, “’One May Come to Repair Musical Instruments’: Rabbinic Authority and the History of the Shevut Laws,” Jewish Studies, an Internet Journal  (forthcoming).

[15] Gilat, Studies, 39-43. (Professor Yitzhak D. Gilat (1919-1997) was a Professor of Talmud at Bar-Ilan University. He studied in Lithuanian Yeshivot and in Yeshivat Hevron in Jerusalem and graduated from the Hebrew University. He served as guest professor at Tel-Aviv University, at Yeshiva University in New York and at the Shechter Rabbinical Seminary in the United States. He published many books and articles in the field of Talmudic literature and history of the halakha.)  See also Rabbi Bin Nun’s discussion of R. Eliezer as well as Rabbi Hoffman’s insightful discussion and his citation from Midrash Lekach Tov. 

[16] See ibid., 36-39.

[17]

בן זומא כשראה אוכלסין בהר הבית או’ ברוך מי שברא את אלו לשמשני כמה יגע אדם הראשון ולא טעם לוגמה אחת עד שזר’ וחרש וקצר ועמר ודש וזרה וברר וטחן והרקיד ולש ואפה ואחר כך אכל ואני עומד בשחרית ומוצא אני את כל אילו לפני כמה יגע אדם הראשון ולא לבש חלוק עד שגזז ולבן ונפס וצבע וטווה וארג ואחר כך לבש ואני עומד בשחרית ומוצא את כל אילו לפני כמה אומניות שוקדות ומשכימות ואני עומד בשחרית ומוצא כל אילו לפני

See also Y. Shekalim 5, 1 (48c), and ibid., 42.

[18] B. Shabbat 74b and Y. Shabbat 7:2, 10b.

[19] See Gilat, Studies, 35 n. 15, citing R. Eliezer of Metz, “Scripture left it up to the sages [to determine that] these are fitting to be called melakha.”

[20] Abraham Goldberg, Commentary to the Mishna Shabbat (Jerusalem: Jewish Theological Seminary of America, 1976) (Hebrew), 137; and Shmuel Safrai and Ze’ev Safrai, Mishnat Eretz Israel (Jerusalem: E.M. Liphshitz College Publishing House, 2008), 259. For this reason, I would not accept Rabbi Hoffman’s suggestion that this number is chosen to relate it to forty-less-one lashes. I do agree, however, that the rabbis enacted great leniencies in the Shabbat laws compared to Second Temple sectarian interpretations. See further at Hidary, “‘One May Come to Repair Musical Instruments.’”

[21] ויאמר אליהם אלה הדברים וגו’. רבי אומר, להביא שלשים ותשע אבות המלאכות שאמר להם משה על פה

[22] See comment at I. H. Weiss, Mekhilta (Wien: Buchdruckerei Von Jacob Schlossberg, 1865), 111.

[23] Section F in the table below.

[24] Perhaps an original version of the Bavli had only the wording of the Mekhilta without “דברים, הדברים.” But even without this conjecture, these midrashim may simply be pointing out the superfluity of the words אלה הדברים so as to understand why Rabbi would find in them an allusion to an oral teaching by Moshe. See also Rabbi Bin Nun’s article and the next note.

[25] Midrash ha-Gadol is a 14^th cent. C.E. Yemenite anthology that includes much ancient material that has been otherwise lost.

[26] I disagree, however, that this derivation is more original than any of the others or that this is the original meaning of the Mekhilta and its parallels at B. Shabbat 70a and 97b (see the previous note) or of R. Hanina bar Hama in B. Shabbat 49b. The meaning of R. Hanina bar Hama’s statement is clearly explicated at the end of the sugya as belonging to the fourth strategy of deriving the avot from the types of work activities performed in the mishkan; see further analysis below for strategy IV.

[27] See further discussion below and in Rabbi Bin Nun’s article.

[28] See Lea Himmelfarb, “The Identity of the First Masoretes,” Sefarad 67, no. 1 (2007), 37-50.

[29] See references at Gilat, Studies, 34 n. 12.

[30] This explanation does not sit well with the manuscript evidence since no manuscript or early printing includes מלאכתו in the list of what to count. Nevertheless, there is enough confusion in the textual transmission reflected in the many textual variants to warrant a proposal that attempts to reconstruct a lost original formulation. See variants below, n. 37.

[31] For an enumeration, see Saul Lieberman, Ha-Yerushalmi ki-fshuto (Jerusalem: Hotsa’at Darom, 1934), 128, citing Israel Elnecave, Menorat Hama’or (New York, 1843), 3:603-604.

[32] Gilat, Studies, 34.

[33] See Goldberg, Commentary to the Mishna Shabbat, 243 and 38 n. 10; and Safrai and Safrai, Mishnat Eretz Israel, 352.

[34] Y. Shabbat 7, 2  (9b-10d). For the last instance, Ed. Venice reads, “mikdash,” but this does not fit the context; see parallel at Y. Shabbat 12, 1 (13c), which reads, “mishkan.”

[35] Translation based on Jacob Neusner, ed., The Talmud of the Land of Israel: A Preliminary Translation and Explanation, 35 vols. (Chicago: University of Chicago Press, 1982).

[36] Translation based on Isidore Epstein, ed., Babylonian Talmud  (Soncino Press, 1971).

[37] See Leib Moscovitz, Talmudic Reasoning: From Casuistics to Conceptualization (Tübingen: Mohr Siebeck, 2002); and Hidary, “‘One May Come to Repair Musical Instruments.’”

[38] T. Beṣah 2:18; Mekhilta De-R. Shimon bar Yoḥai 12:16; and see Gilat, Studies, 53 n. 100.

[39] B. Betzah 13b and see Gilat, Studies, 48-59, who shows that various other conditions that are only implicit in the Mishnah and explained to some degree in the Yerushalmi are only fully conceptualized and applied in the Bavli.

[40] On the conceptual and textual linkage between Shabbat and the mishkan generally, see Nehama Leibowitz, New Studies in Shemot (Ha’omanim, 1995), 471-86; and Moshe Shamah, Recalling the Covenant: A Contemporary Commentary on the Five Books of the Torah (Jersey City: Ktav, 2011), 462-63 and 501-3.

[41] Text from MSS Cambridge – Westminster College G.F. Talmudica II 2. Major variants are indicated in the notes.

[42] MSS Oxford 366, Munich 95, and Ed. Soncino add “bar Hama.”

[43] MS Oxford 366 reads, “מלאכה ומלאכת.” MS Munich 95 reads, “מלאכה ומלאכות.” MS Vatican 108 reads, “מלאכה ומלאכה.” Eds. Soncino and Venice reads, “מלאכה מלאכתן ומלאכות.” Ed. Vilna reads, “מלאכה מלאכתו ומלאכת.” See discussion above.

[44] MS Oxford 366 also reads, “מיבעיא.” MS Munich 95 has, “מסתפקא,” MS Vatican 108 has, “מיסתפקא,” and Ed. Soncino reads, “מספקא.”

[45] MS Vatican 108 also reads, “ויבא הביתה,” but it is omitted in MSS Oxford 366, Munich 95, and Ed. Soncino.

[46] Lieberman, Ha-Yerushalmi ki-fshuto, 128, cites an alternate version that makes this line similar to B. Shabbat 70a and 97b, cited above.

[47] All other witnesses omit, “כיצד.”