A. The Textual Practice of Polysemy and Multivocality

This section consists of two Mekhilta de-Rabbi Ishmael passages. Text 1 is legal,Footnote ¹² describing the procedure attending a slave's decision to remain with his master even after the seventh-year release: “He shall be brought to the door or the doorpost; and his master shall pierce his ear with an awl; and he shall serve him for life” (Exodus 21:6). The Mekhilta passage consists of a number of elements: a “you might say ... or perhaps it is merely ...” (אתה אומר... או אינו אלא) argument that concludes that Scripture refers to the right ear; a dispute between Rabbi Judah and Rabbi Meir as to the precise location of the piercing (earlobe or cartilage?); a brief midrash identifying the ear as the locus of punishment because it heard God's commands at Sinai;Footnote ¹³ and a contrast between the scriptural language (awl) and what “Halakhah says” (any instrument). Fraade makes a number of closely packed comments about the passage. The Mekhilta's “preferred understanding is provided first, only to be followed by alternative understandings” (9); the contrast between Scripture and Halakhah is “a bold statement” of the fact that “the rabbinically determined practice may depart from the plain meaning of divinely revealed Scripture” (9); the passage exemplifies the way in which “the various opinions are not simply listed but are editorially given voices and joined in implicit dialogue with one another” (9–10).

The Mekhilta's practice of introducing interpretations that it then goes on to reject is not polysemy, but rather its opposite: There is a correct interpretation, but the Mekhilta rhetorically foregrounds an incorrect one, in order to make explicit the interpretive path not taken. The Mekhilta's contrast between extra-scriptural tradition (Halakhah) and the biblical precept is attested in numerous passages in the Mishnah and Tosefta that contrast halakhot and Scripture. Some of the best known examples are M. ^cOrlah 3:9's division: “The new produce is prohibited by the Torah throughout, the fruit of young trees (^corlah) by Halakhah, and crossbred kinds (kila'yim) by the words of the scribes”; M. Ḥagigah 1:8's characterization of the laws of the Sabbath, the festal offerings, and sacrilege, as “mountains that hang on a hair, for they have few biblical verses and many halakhot”; and the dispute in T. Sukkah, “the willow ritual is an extra-scriptural tradition (halakhah) given to Moses at Sinai. Abba Shaul says: It is from the Torah…” (T. Sukkah 3:2).Footnote ¹⁴ Finally, recording different legal opinions does not constitute polysemy; it is an inclusive editorial practice that recognizes the empirical existence of controverting views, and chooses to record them. It does not reflect a theological commitment to the inherent plurality of meaning in God's word.

Text 2 is an aggadic passage from the Mekhilta consisting of two interpretations of the phrase “house of bondage” (Exodus 20:2), separated by the phrase davar 'aḥer.Footnote ¹⁵ Fraade claims this as an example of the tannaitic tendency to set “multiple interpretations and legal opinions not simply alongside one another but in rhetorical dialogue …” (11), and contrasts this approach with other corpora such as the Hebrew Bible, the New Testament, and the Dead Sea Scrolls, whose editorial practices tend toward univocality. But the derashah in question does not produce such a dialogue since davar 'aḥer marks the juxtaposed interpretations as distinct, editorially demarcated, statements.

B. Early Rabbinic Thematization of Legal Multivocality

The first two texts in this section, texts 3 and 4 (reproduced in the appendix), are drawn from Sifre Deuteronomy pis. 48. The first is a short midrash on Proverbs 27:7, which likens a Torah scholar to a sieve: “Just as a sieve separates flour, bran, and meal, similarly [a disciple] sits and sorts words of Torah and weighs them: so-and-so permits, so-and-so prohibits, so-and-so declares impure, so-and-so declares pure.” The second links Amos 8:12 (“They shall wander about seeking the word of the Lord, but they shall not find it”) with a person traveling from city to city to find the answer to an obscure legal question—corresponding to Amos's wandering in search of the word of the Lord.Footnote ¹⁶ Rabbi Shimon bar Yoḥai then asks whether the prophet foretells the forgetting of the Torah (“the word of the Lord”) and, having shown that Deuteronomy 31:21 forecloses this possibility, glosses forgetting as a situation in which “so-and-so prohibits, so-and-so permits, so-and-so declares impure, so-and-so declares pure, and one cannot find a clear rule.” In keeping with his thesis, Fraade finds here a valorization of the opposing legal positions, since all of the contradictory opinions “are to be preserved through the dual process of sorting and memorizing through repetition,” such that “the halakhic tradition need not be stripped of its multivocality” (14).

I do not see how these sources support Fraade's conclusion. Text 3 states that the disciple must sort, sieve-like, through the teachings he has received, in Shlomo Naeh's words, “classifying mixed materials—an activity without which the Torah he has acquired is not fit for use, like flour that requires sifting.”Footnote ¹⁷ In other words, the disciple has heard various traditions and now must sift through them and order them so as to make them suited for legal discussion. It is not clear what the disciple is to do with the teachings that are trapped, as it were, in the sieve. The Sifre does not elaborate, and its silence on this point is significant, since this is the heart of the matter from a legal-pluralist perspective. If they are discarded, the passage is outright anti-pluralist. But even if they are to be preserved, the resulting pluralism is weak: it acknowledges the existence of opposing traditions, but does not valorize this reality in any way. Text 4, in Fraade's own words, “problematizes this proliferation of rabbinic teachings...”—nota bene, it problematizes this pluralism; it does not celebrate it. Despite that, Fraade tries to salvage a polysemic position by characterizing the Sifre's problematization as undertaken “... specifically in terms of memory” (14), with the sorting activity functioning as a guarantor of the teachings' preservation for later generations. This is, in a sense, true, but the derashah's estimation of this state of affairs is significantly different than Fraade suggests. It is not the case that Shimon bar Yoḥai positions Amos 8:12 (“they shall not find the word of the Lord”) and Deuteronomy 31:21 (“It will not be forgotten from the mouth of their offspring”) as two contradictory views whose resolution is that, in his words, “the sorting and (repeated) recitation of the conflicting opinions will ensure that they are remembered” (14). Rather, the forgetting of the Torah is introduced as a possible but ultimately untenable interpretation of Amos 8:12 (since it contradicts Deuteronomy 31:32), only to be replaced by the correct interpretation. According to the Sifre, then, “They shall not find [the word of the Lord]” means that they will not find a single, authoritative legal position because “so-and-so prohibits, so-and-so permits, so-and-so declares impure, so-and-so declares pure.” The passage, in other words, is not advocating the preservation of contradictory legal views (which, to repeat, would not constitute pluralism or polysemy in any case), but lamenting a state of affairs in which the proliferation of such views impedes a definitive and unambiguous legal ruling, equating it with Amos's dire prophecy that “they shall not find the word of the Lord.”Footnote ¹⁸

Fraade discusses texts 6 and 7 (reproduced in the appendix) jointly, though they are concerned with different matters. Tosefta ^cEduyot 1:1 (text 6) explains the ordering of the oral testimonies as a rabbinic response to an imminent situation in which “a person will go looking for a teaching from the words of Torah and will not find it, from the words of scribes and will not find it.” Mishnah ^cEduyot 1:4-6 (text 7) asks, “Why do they record the words of Shammai and Hillel for naught?”Footnote ¹⁹ and “Why do they record the words of an individual among (those of) the majority, whereas the Halakhah may only be according to the words of the majority?” The Mishnah answers the first question with a call for jurisprudential humility (“to teach future generations that a person should not persist in his argument”), the second with two incommensurable answers: one anonymous, that the minority view is preserved as a precedent for a future ruling, another, attributed to Rabbi Judah, so that future legists know to discount that position.

Text 6 is not apposite, as it merely justifies the organization of Halakhot by tradent (“what is of the House of Shammai and what is of the House of Hillel?”), as opposed to the thematic structure that prevails in other Mishnah and Tosefta tractates. Fraade argues that the preservation of the traditions of both Hillel and Shammai (likely a synecdoche for irreconcilable rabbinic views) shows that the editorial process involved “collecting and ordering the conflicting rulings ... without stripping them of their multivocality” (19). However, multivocality is not a characteristic of individual rulings, and so it cannot be stripped from them. Multivocality inheres to a collection of rulings that preserves many positions—and even then only in a weak editorial sense that has long been recognized as a hallmark of rabbinic literature. Tosefta ^cEduyot recognizes the prima facie oddness of preserving both legal positions, such that the relative openness of the tannaitic sources stands out against the backdrop of, say, the Dead Sea Scrolls. But for an investigation of “talmudic” polysemy in tannaitic sources, the operative comparison is to the Babylonian Talmud, and here the contrast is stark. In the anonymous stratum of the Bavli, the conflicting views of the Houses of Hillel and Shammai give rise to a clear statement of the multivocality of received traditions: “These and these are the words of the living God” (B. ^cEruvin 13b). Not so the Mishnah, which eschews theological justification for this practice.

Text 7 does not deal with polysemy in any sense, and is problematic for the claim of legal pluralism. Its question “why do they record the words of the individual among (those of) the majority for naught?” is driven by the view, explicitly stated, that the minority's position is wrong, the majority's right. To be sure, the Mishnah is playing a sophisticated game here. The anonymous (majority) view is that today's minority position may be tomorrow's majority ruling (M. ^cEduyot 1:5); while Rabbi Judah's (minority) position is that the minority view is preserved so that future tradents of this view not gain unwarranted currency. As Moshe Halbertal has written, “The editor thus created a kind of circular paradox: according to the majority view, minority opinions were mentioned in the Mishnah so that they could be revived in the future; whereas in the eyes of the minority, minority opinions were mentioned so they could be rejected forever.”Footnote ²⁰ All the same, both the anonymous view and Rabbi Judah hold that there is a right and a wrong position, differing only on whether the two might be reversed at a future time. The fundamental orientation is non-pluralistic: the conflicting views do not merit preservation in their own right. It is only the theoretical possibility that in some future time the rejected position will find favor that justifies its preservation.

C. The Multivocality of Revelation

In the final section of his essay, Fraade argues that the Sinai theophany was a site of polysemic revelation. Text 9, from the Mekhilta de-Rabbi Ishmael,Footnote ²¹ describes the voices at Sinai allowing “each person to hear them according to his capacity” (לפי כחו), suggesting a multivocality embedded in the theophany. But hearing “according to one's capacity” has nothing to do with divine polysemy, and everything to do with the varying cognitive abilities of the auditors. As Richard Hidary notes, “this midrash does not necessarily reflect multiple conflicting interpretations but rather deeper or more superficial understandings of the single message of the voice.”Footnote ²² But even this may be more than the text allows. For as Fraade notes, the “midrashic passage is attentive to the plural forms of ‘thunderings’ and ‘lightenings’ in the sound and light show that accompanied revelation at Mt. Sinai” since “the Hebrew word employed for ‘thunderings’ (kolot) can also mean ‘voices’...” (24).Footnote ²³ Fraade's reading is attentive to the nuances of the biblical text—identifying the plural kolot as the trigger for the plural understandings the Mekhilta attributes to the Israelites—and his analysis is subtle and convincing. But it also works against his broader argument. For the Mekhilta is not making the polysemic claim that a single divine voice spoke at Sinai but produced many meanings, but rather that there were plural voices from the outset—kolot and not kol.

Text 11, from Sifre Deuteronomy,Footnote ²⁴ states that for each word God spoke at Sinai “Israel would observe it and would know how much midrash could be derived from it, how many laws (halakhot) could be derived from it, how many a fortiori arguments (kalin va-ḥamurin) could be derived from it, how many arguments by verbal analogy (gezeirot shavvot) could be derived from it.” As with the previous passage, Fraade offers a penetrating interpretation of the passage. “The Israelites are able to discern the plenitude of meanings that will one day be derived from each divine utterance through rabbinic ... hermeneutical methods of exegesis. Although those methods ... are expressed in plural forms, there is no suggestion here that they are in discord with one another or arise in the context of debate” (27; emphasis added). Fraade nonetheless counterposes this passage to Boyarin's claim that it is only in the Talmud that we find “thematized and theorized for the first time the theological principle ... that the divine language produces manifold and different meanings.”Footnote ²⁵ But the Sifre Deuteronomy passage is not concerned with plurality of meaning, but rather with bridging, indeed effacing, the historical chasm stretching from revelation to belated rabbinic interpretation. It asserts, in Fraade's words, that rabbinic conclusions were already visible “at the originary moment of divine utterance and human reception” (27). It is not, then, the plurality of rabbinic positions here addressed, but rather the gap between the biblical word and the later rabbinic interpretation.

Text 12 (see appendix), also from Sifre Deuteronomy,Footnote ²⁶ interprets Deuteronomy 33:2 as a series of statements describing the Sinai theophany, each associated with another language: “The Lord came [בא] from Sinai” with Hebrew; “and dawned from Seir upon us” with “Roman”; “he shone forth [הופיע] from Mount Paran” with Arabic; and “he approached [אתה] from Rivevot-Kodesh” with Aramaic. Taken together, the verses teach that God's revelation to Israel was “not just in one language” but in four. The precise meaning of this derashah is elusive. As Fraade acknowledges “[i]t is unclear here whether the four languages were uttered/heard simultaneously or in succession.” This is a critical issue, since if the statements are made in succession, the midrash is not polysemic. Not surprisingly, Fraade suggests the four utterances were heard simultaneously, on the basis of two considerations. One is the subsequent interpretation—introduced by davar 'aḥer—that explains the fourfold statement in geographic terms: God's revelation was “not just from one direction ... but from four directions, with Sinai, Seir, and Paran constituting three, and a fourth introduced by Habakkuk 3:3, ‘God comes from Teman.’” Another is that “[t]his is clearly the understanding in the later tradition of God's speech having ‘split’ into seventy languages at Sinai” (31 n. 101). The latter is inadmissible on logical grounds. All scholars agree that polysemy is attested in post-tannaitic strata; Fraade claims to show that this is also the case for the tannaitic sources. Invoking continuity between earlier and later traditions to justify a polysemic reading of Sifre Deuteronomy pis. 343 is nothing more than petitio principii (i.e., begging the question). Fraade is asserting the very proposition (tannaitic and post-tannaitic sources demonstrate continuity with regard to polysemy) that requires demonstration.

The analogy between the two derashot is also untenable but on different grounds. If the second (geographic) derashah speaks to the simultaneity of the first (four languages), it reflects the view of the editor who juxtaposed the two, not the meaning of the first text. But even this is not the case, since the editor marks the difference between the two by interposing the phrase davar 'aḥer between them. As noted above, this phrase indicates that the editors understand the derashot to be substantively distinct. And finally, because the two derashot differ in a critical respect: the first understands Sinai, Seir, Paran, and Rivevot-Kodesh to be synonyms—four reports of the same event framed in different terms. In the second, Sinai, Seir, Paran and (imported from Habakkuk) Teman are different locations, so that each stich of the verse refers to a discrete geographic aspect of revelation. The simultaneity of the distinct geographic locations speaks to the geographic—not linguistic—scope of God's revelation, and therefore cannot be blithely transferred to the neighboring derashah.

But let us assume, for the sake of argument, that the Sifre Deuteronomy understands the four languages to have been spoken simultaneously. What exactly does this teach us? Less, I fear, than Fraade would have us believe. For one thing, the tetraglossic revelation consists of four enunciations of the same statement, not four disparate and incommensurable divine dicta—so this is not “another articulation of the rabbinic conception of the multivocality of divine revelation” (31). Though speaking four statements at the same time is wondrous (if that is what the Sifre in fact describes), a four-fold enunciation of the same content in different languages is not polysemy, but rather polyglossia. Fraade tries to circumvent this difficulty by asserting that, while polyglossia and polysemy are not identical “to the extent that they are both associated with divine speech and its immediate human reception, they are complementary expressions of revelatory multivocality” (29 n. 93). But this is a conceptual error. Any statement, whether divine or human, can, in classic translation theory, find expression in—that is, be translated into—many languages, without thereby denoting or even connoting polysemy. Moreover, the derashah is responding to the four stiches of Deuteronomy 33:2: “The Lord came from Sinai, and dawned from Seir upon us; he shone forth from Mount Paran, and approached from Rivevot-Kodesh.” The invocation of four languages (not polysemic in any case) is an exegetical attempt to account for a perceived redundancy in one verse, not a statement of principle about the nature of divine speech or of the biblical text, as such.Footnote ²⁷

Fraade's last example, text 13 (reproduced in the appendix), is the Tosefta's discussion of the phrase “masters of assemblies” (בעלי אסופות) which occurs in Ecclesiastes 12:11: “The sayings of the wise are like goads, and like nails firmly fixed are the masters of the assemblies, given by one shepherd.” This source, Fraade notes, is “central to Boyarin's denial of any rabbinic thematization of polysemy and pluralism prior to the post-amoraic stratum of the Babylonian Talmud” (31). Here the key issue is the passage's later reworking in the Babylonian Talmud. In the Tosefta, the masters of the assemblies “enter and sit in multiple assemblies, declaring what is impure [to be] impure, and what is pure [to be] pure...” The Talmud's masters of the assemblies, in contrast, are disciples of the sages who sit in assembly where “these forbid and those permit; these declare impure and those declare pure; these declare unfit and those declare fit” (b. Ḥagigah 3b). Shlomo Naeh has offered a subtle philological analysis of the two passages, arguing that the Tosefta is concerned with legal monovocality and needs to justify the preservation in memory of the earlier controversies of Hillel and Shammai, for example. The Talmud, in contrast, lionizes legal controversy among the sages.Footnote ²⁸ Boyarin, for his part, has incorporated Naeh's insights into a broader pattern of Babylonian revisions of earlier traditions that he locates in the anonymous stam. Fraade does not engage this line of argument, however, as his “focus, by contrast, will be on T. Sotah 7:11-12 in its own right and not in the shadow of B. Hagigah 3b” (31 n. 102). But that is precisely Boyarin's point, as I understand it: without “the shadow of the Bavli,” no one would think that the Tosefta was addressing the question of polysemy. As noted, the first scene offers a nostalgic picture of the masters of the assemblies “declaring what is impure [to be] impure, and what is pure [to be] pure.” The second warns of the danger of legal controversy: “lest a person think to himself, ‘Since the House of Shammai declares impure and the House of Hillel declares pure, so-and-so prohibits and so-and-so permits, why should I henceforth learn Torah?’” After a brief and implacably obscure derashah (perhaps to Deuteronomy 1:1), the passage addresses a hypothetical student who despairs of Torah study due to the current plurality of legal positions, allaying his concerns with the claim that “All the[se] words were given by one shepherd. One God created them, one benefactor [Moses] gave them ... .” The Tosefta passage concludes with an exhortation to preserve in memory the words of both Hillel and Shammai, “the words of those who declare impure and the words of those who declare pure.”

Fraade offers a relatively lengthy analysis of the passage and concludes that “the multivocality of the ‘words of sages/Torah’ is originary and not belated to divine revelation,” since “[a]ll the words of the sages, whether in consensus or dissensus, were ultimately created, given, and uttered by a single God and human intermediary” (36). True enough, and text 13 may well represent the closest thing to polysemy in the tannaitic corpus. But even here, Fraade's conclusion glosses over significant issues. First, the Tosefta by no means celebrates this state of affairs. To the contrary, the existence of different legal positions is a problem that could drive people away from Torah study. Second, the Tosefta contrasts the reality of conflicting legal views with the (idealized) halakhic uniformity of days gone by, in which the masters of the assemblies would sit in judgment “declaring what is impure [to be] impure, and what is pure [to be] pure.” This is a critical point: under ideal circumstances halakhic decisions should be decided by a monovocal, legal realist approach—things that are (in reality) pure, should be declared pure, and things that are (in reality) impure, should be declared impure. The legal landscape, however, has changed, and the Tosefta must acknowledge that different schools espouse different positions. The identification of a single divine source for “all the[se] words” is a postlapsarian concession to the contemporary reality of competing legal positions. The Tosefta can only implore the reader to preserve both sides of halakhic controversies, and thus keep the divergent views from being lost to future generations. Grudging legal pluralism is the last recourse against the potentially deleterious effect of halakhic discord on future Torah students. How far this is from a principled affirmation of the inherent polysemy of God's word, and from the robust legal pluralism of sages who “promoted—even celebrated—an ideology and intellectual culture that encouraged … rabbinic groups or individuals to ‘agree to disagree’ and to ‘teach the controversy’ when it could not be resolved” (4).

* * *

Having carefully examined the material Fraade has marshaled in support of tannaitic polysemy, I conclude that if Fraade's careful analysis of the tannaitic sources has failed to provide a clear example of polysemy or of valorized legal pluralism, it is because these are not to be found until a significantly later stratum of the rabbinic corpus.