Featured
What Is Midrash?
Analyzing midrashic terminology offers a fresh perspective.
Bíblia de Cervera, Deuteronómio (fl 118v), Tempera, gold, and ink on parchment, illuminated by Josef Asarfati (1299–1300). Wikimedia Commons. Public Domain
By Ishay Rosen-Zvi
“What is midrash” is the title of the never-ending magnum opus that I have been conspiring to write for almost two decades now.1 In the meantime, I am collecting more and more material and writing various philological articles on specific midrashic terms and techniques, articles that will someday hopefully join into a larger thesis and in the interim are read by a devoted audience of a few dozen colleagues and students (no complaints; this is our profession). I will try to explain here what it is that I dream about, why it is potentially significant, and why it is so difficult to complete. And maybe this will be a push for me to finally bring things to a happy ending.
The nature of Midrash has troubled great minds since the inception of Jewish Studies as a discipline. In 1844, the masterful Abraham Geiger, who was also one of the leaders of the nascent Reform movement, diagnosed the Sages with “an extremely muddled (gertrübt) exegetical sense.”2 Subsequent scholars tried their hand at explaining or even justifying the interpretations that Midrash creates. Zacharias Frankel hypothesized that we are missing a large piece of the puzzle. Isaac Hirsch Weiss reinstated a traditional explanation that the Midrash was revealed at Sinai, putting the divine logic beyond our ken. Other traditionalists, like the Malbim, R. Meir Leibush Wisser, claimed that Midrash actually reveals the hidden depth of Scripture’s sense. Some, on the other hand, argued that Midrash is not interpretation at all, but a textual tack that fastens rabbinic ideas and stories to the biblical text. What all of these explanatory forays have in common is the notion that ultimately it is impossible to understand the nature of Midrash as exegesis.
The mid-twentieth century saw a major breakthrough with the publication of Isaac Heinemann’s Hebrew study Darkhei ha-Aggadah (The Methods of Aggadah).3 Unlike some of his predecessors, Heinemann counted Midrash as interpretation, albeit a type that operates with its own principles and assumptions. He coined the term “creative philology” to express how Midrash operates. It is “philological” in the sense that it pays acute attention to the original text’s nuances, and it is “creative” in that it can uproot the text from its surroundings and plant it in new contexts, or water it with imagination and let it sprout new meanings. Heinemann’s major contribution was to take Midrash seriously as interpretation and to typologize its methods. His work, though, was very much a product of a particular moment in the history of literary theory. He viewed Midrash through the lens of Romanticism, especially the cycle of Stefan George, in which the space between text and interpreter collapses.
Susan Handelman was the first to systematically apply post-structuralist methods to Tannaitic Midrash. In her 1983 Slayers of Moses: The Emergence of Rabbinic Interpretation in Modern Literary Theory she claimed that the midrash is the definitive model of Derrida’s différance, in which the game of signifiers never ends.4 In his Midrash and Theory, David Stern criticized this reading arguing that the Midrash does not assume that the text is empty of meaning, but rather that it explodes with meanings, which the darshan is set to unearth, an assumption with a completely different theological and hermeneutic baring.5
James Kugel has laid out fundamental principles of ancient Jewish interpretation.6 He not only brought biblical interpretation to the forefront but also helped contextualize Midrash as part of a broader intellectual trend. Here, however, our focus is on what distinguishes Midrash from earlier modes of exegesis.
Daniel Boyarin’s 1990 Intertextuality and the Reading of Midrash argued that practitioners of Midrash see the entirety of the biblical canon as the immediate relevant context for understanding every word of the Torah.7 Structurally speaking, while the modern reader approaches the Torah as parole, a collection of utterances to be understood on their own terms, the ancient darshan viewed it also as langue, a lexicon from which one draws and recombines phrases to say something that has never been said before.
Azzan Yadin-Israel has developed similar hermeneutic tools to examine legal, halakhic midrashim of the two schools. He proposed that for R. Ishmael, Torah is conceived as both code and cipher. The way the Midrashim of this school talk about Scripture indicates that it is an active partner in interpretation: “Scripture comes to distinguish,” “Scripture comes to teach,” “until Scripture expressly decides,” and so on. The darshan is the apt pupil, who expectantly looks for a sign from Scripture to know how to proceed.
Rabbi Akiva’s school, in contrast, is trying to hitch tradition to Scripture by any suitable device. Its primary motivation is the desire to fuse the written and oral Torah into a single system.8 Note however that the vast majority of the midrashim of Rabbi Yishmael’s school consist of exposition on verses without any kind of marking. Yadin-Israel’s discovery thus may belong more to the realm of rhetorical strategy than to hermeneutics.
Maren Niehoff and Yakir Paz have also studied the distinctive approaches of the schools. They compare the Midrash to the Scholia, brief marginal notations made by Hellenistic scholars on the Homeric corpus which developed in the Alexandrian Museum from the third century BCE. Paz analyzed the similarities between the ancient Alexandrian commentators and the hermeneutic activity of the darshanim, especially in the simpler exegetical techniques of Rabbi Yishmael’s school. The idea that one can resolve a local ambiguity with express statements from elsewhere, bears an unmistakable resemblance to a principle in Homeric interpretation, “to clarify Homer from Homer.” Niehoff has demonstrated the profound influence of the Homeric Scholia on Jewish Alexandrian exegesis. Paz suggested that Alexandrian Jewish exegetes were the ones who adapted Homeric interpretation to biblical exegesis, which were then taken up by the rabbinic academies.9
Maurycy Gottlieb, Jews Praying in the Synagogue on Yom Kippur, oil on canvas, 96.4 x 75.5” (1878). Wikimedia Commons. Public Domain
My own intervention in the field is neither historical nor comparative nor hermeneutic in the theoretical sense (although I try to make use of all these tools) but mainly in analyzing midrashic terminology. Systematic analysis of midrashic terminology yields, time and again, a different result than individual treatments of selected drashot.
From a close reading distance, these selected drashot seem like taking an unpredictable, freewheeling flight path through the Torah. Closer up, one notices clear patterns, a limited number of exegetical maneuvers that repeat in variable sequences. In the same way that microhistory tells us a story different than those painted by the large canvas of social and political historiographies, terminological inquiry can put midrashic hermeneutics in an utterly different light.
Tannaitic Midrash stands out for its utilization of prescribed terminology organized within rigid configurations, reiterated repeatedly. This sharply contrasts with the landscape of pre-rabbinic biblical interpretation, such as the Dead Sea Scrolls, where scholars toiled hard to assemble only a handful of midrashic-like terms.
These midrashic frameworks establish a series of constraints, within which diverse permutations come to light. Our focus in reading midrashic terms should thus encompass both commonalities and deviations, involving an exploration of the structure as a vessel of domestication (everything sheltered beneath it emerges as a recognizable midrashic maneuver) as well as a generative force (enabling the continual origination of fresh variations). I will flesh this out with a few examples.
Take the term, “it is not needed,” or “it is unnecessary” (eino tzarikh). Tannaitic Midrash tends to present multiple interpretations of a single verse or phrase, which is in itself an unprecedented phenomenon in the history of biblical interpretation, especially since it makes no effort to decide or even create hierarchy between them. The Midrash usually separates these interpretations with attributions to named Sages (Rabbi X says . . . Rabbi Y says) or the separator davar aḥer (another matter), without stating any preference.
But 55 times in Tannaitic Midrash the term eino tzarikh, “it is unnecessary,” appears, mostly in the school of Rabbi Yishmael, which is known for its more developed technical terminology. The term is placed after a drasha to convey that it is unnecessary, because a better way to learn the same halakhah10 exists.
Sometimes, eino tzarikh means that no exegesis is necessary at all. For example, when the Midrash tries to establish that priests must have their genitalia covered based on the prohibition against having a stepped altar (Ex. 20:26), Rabbi Ishmael objects on the grounds that it is explicit in another verse, which states that the pants worn by the priests cover their private parts (Ex. 28:42). Why work at extrapolating something that is stated plain as day?
Quite often, though, eino tzarikh introduces another, ostensibly better drasha. Here is an example discussing the requirement for kohanim (priests) to stand while reciting the priestly blessing:
“Thus shall you bless the Children of Israel” (Num. 6:23)—while standing. Do you say while standing, or is it instead while standing or not standing? It therefore teaches, “These shall stand to bless the people on Mount Gerizim” (Deut. 27:12). It says “blessing” here and it says “blessing” there. Just as the blessing said there is while standing, so the blessing said here is while standing.
Rabbi Natan says, It is unnecessary (eino tzarikh). For it already says, “and the kohanim, the sons of Levi, shall come forward etc. in the name of the Lord” (Deut. 21:5), juxtaposing blessing to service. Just as service is while standing, so blessing is while standing. (Sifrei Numbers 68)11
The first drasha is anonymous, but then comes a named rabbi who declares it unnecessary and provides an alternative derivation of the same law. In what way is the alternative superior to the original? The Midrash never says. The answer must be extrapolated inductively from the individual drashot.
A clear distinction is noticeable here. The original drasha is narrow and limited, while the improved one is conceptually broader. Instead of a specific analogy, Rabbi Natan offers a general inference that blessings are to be said while standing. An analysis of all the occurrences of eino tzarikh discloses that the superior drasha is based on a more direct verse, has a wider compass, or entails fewer midrashic moves. The rejected dashah exhibits a tendency towards complex structures, replaced by a simpler exegesis.
This finding adds a deep wrinkle to the conventional account of the Midrash’s anchoring of extra-biblical law in the Torah. The typical directionality proposed in scholarship is towards increasing complexity and sophistication. While it is true that over time darshanim forged more effective tools that allowed them to craft more sophisticated midrashic arguments, they concurrently honed their skills to the point that they could eliminate waste and bloat. Even if we should be cautious about drawing paradigm-shaking conclusions from snatches of data, we can state that optimization for simplicity did exist in the Tannaitic bet-midrash,12 quite contrary to the common scholarly image of midrash.
Here is another example for the power and promise of examining terminology. Take ella, which literally means “but” or “rather.” My interest here is not with the thousands of occurrences of ella as a word in the Tannaitic Midrash, but rather with a specific structure in which ella appears after a question and before a drasha. Roughly 380 such drashot appear within the Tannaitic Midrash. Within this distinct structure, the term ella serves as a pivotal juncture, facilitating the shift from a typical straightforward and minimalist interpretation to a more enriched, homiletical, and moral perspective.
The term ella finds its origins in Aramaic (‘en la, אן לא; if not), but its adoption to convey the meaning of “rather” likely bears the imprint of Greek influence. The usage of ἀλλά in a comparable manner can be traced back to the Homeric Scholia, forming a question-and-answer structure, as highlighted by Yakir Paz. However, neither the Aramaic lineage nor the Greek parallel fully elucidates the consistent and widespread application of this structure within the Tannaitic Midrash.
The salient hallmark of this configuration lies in its dialectical nature, wherein the interplay between the question and its response cultivates a transformation in the comprehension of the biblical text. Deciphering this transformation, explicitly marked by ella, can teach us a lot about what the Midrash is up to.
Ella appears mostly in aggadic context. In many of these aggadic homilies the plain reading isn’t merely a starting point, but rather stands as a genuine alternative, which is rejected due to interpretive or ideological challenges. Consider the following two examples:
“You are standing this day all of you” etc. “your little ones,” etc. [Deut. 29, the passage of hakhel, the periodical reenactment of the covenant]. But what do the little ones know about distinguishing good from evil? Rather, to give the parents reward for their children . . . (Mekhilta RI, Pascha 16)
“The children of Jacob were twelve” (Gen. 35:22). Is it not known that they were twelve? Rather, He was announced by the Holy One, blessed be He, that Reuben had repented. (Sifre Deuteronomy 31)
Why is it necessary to bring young children to a covenant-making assembly? Why does Scripture interrupt the story of Reuben’s sin with the seemingly disjointed phrase, “The children of Jacob were twelve”? The first question pertains to the event represented while the second centers on the biblical text itself. However, in both instances, and in many similar ones, the questions seem genuine, and are intriguing modern commentators as well. Note that while this is a matter of gradation and not dichotomy, we do have marks of genuine questions in Midrash, especially when they give rise to multiple answers, while artificial queries are consistently followed by a single response, to which they function as but an introduction.
A distinct phenomenon becomes evident in these two midrashic units: the reinterpretations stemming from the term ella introduce an additional layer of theology and morality to the verse, a layer that is absent in a plain reading. The purpose of including the children is to “give a reward,” while the enumeration of Jacob’s sons alludes to a comprehensive narrative of repentance and forgiveness.
In most of the occurrences of the term in aggadic context, the ella drashot indeed end with an additional religious-moral message. Here are two examples:
Being Bound up in Their Clothes upon Their Shoulders (Ex. 12:34). R. Nathan says: And were there no animal there [to carry the dough]? Has it not been said: “And a multitude went up with them [and flocks and herds even very much cattle]”? (Ex. 12:38) What does “being bound up in their clothes upon their shoulders” teach? Rather, the Israelites cherished the commandments [and so wanted to carry their Matzot themselves]. (Mekhilta RI, Pascha 13)
“A Psalm, A Song at the Dedication of the House of David” (Ps. 30:1). Was it David who built it? Did not Solomon build it, as it is said: “So Solomon built the house, and finished it” (I Kings 6:14)? What does “A Psalm, A Song at the Dedication of the House of David” teach? Rather, since David gave his whole soul to it, it is named after him. (Mekhilta RI, Shira 1)
In the first case the question is realistic, while in the second it is an intertextual contradiction. The next move is similar in both homilies, namely a seemingly inconsequential detail in the verse evolves into a profound teaching: reverence for the commandments and dedication to the temple.
Comparable shifts are a recurring theme within ella drashot. Here’s another one:
R. Eliezer says: But was there not water underneath the feet of the Israelites, since the earth is floating upon nothing but the water, as it is said: “To Him that spread forth the earth above the waters” (Ps. 136:6)? Therefore, what does it teach by saying “And found no water” (Ex. 15:22 [at the beginning of the Journey of the Israelite from Egypt into the desert])? Rather, to tire them out. (Mekhilta RI, VaYasa 1)
R. Eliezer’s question sounds weird. Does he really assume that there is water everywhere in the desert? His actual argument, though, is not realistic but textual. The verse does not say that there was no water, but rather that the Israelites did not find it. God, deduces the darshan, concealed the water purposely, to test the resolve of the Israelites. Thus, the phrase “and found no water” evolves from a mere factual observation into a tale of divine providence, holding within it an explicit directive for the ancient Israelites and an implicit one for the homilist’s audience.
On some occasions, this shift from information to lesson is explicitly marked:
Now, why was the account whether Amram took a wife or did not take one, needed? Rather, to make known to all the inhabitants of the world the merit of Amram the Righteous! (Mekhilta RSBI 6:2)
The verse’s goal is not to convey facts, but to “make known,” to teach a lesson on righteousness.
The supplementary moral layer often introduces a figurative-like interpretation, dislocating the verse from its straightforward, literal reading. Thus, as we have seen, David is narrated as if he built the temple. This is indicative. The majority of aggadic drashot with an added dimension incorporates an “as if” maneuver, with or without the explicit term, which transforms a word or phrase into a non-literal rendition. Thus, a clear pattern is revealed: an interpretive question which leads the darshan to move from one exegetical level to another.
Scholars commonly perceive midrash as the antithesis of allegory due to its refusal to recognize distinct exegetical strata—surface versus depth—presenting instead all interpretations on an equal plane. According to this prevalent perception, in the midrashic approach, a non-hierarchical intertextuality supplants the vertical allegorical logic, and metonymy takes precedence over metaphor. This tendency is frequently linked to the rabbinic disavowal of the Platonic partition between language and reality, rhetoric and (hidden) truth. The metaphysical quest for veiled truths is replaced by an endless intertextual play.
Here are two classic scholarly examples: “[N]or does the midrashic meaning take any precedence over the plain, simple meaning . . . There is no hierarchical scheme in midrash; no interpretation has more authority than any other”;13 “Ben Azzai does not speak of having achieved the original meaning or inner meaning or hidden meaning of Torah . . . He did what he did not by linking texts with their meanings but by linking texts with texts.”14
The dismantling of the common midrash/allegory dichotomy might be another reward of the painstaking terminological search.
Given this classic contextualization of Midrash, which I basically share, it is remarkable to encounter a structure which is fundamentally akin to that employed by Alexandrian allegorists, as delineated by David Dawson in his Allegorical Readers and Cultural Revision in Ancient Alexandria.15 Dawson contends that allegory should not be defined by its thematic components (abstraction, spiritualization, transformation, etc.), as these descriptions are laden with theological implications, but rather through its formal characteristics: two coexisting levels of interpretation—literal and non-literal—held in a dynamic tension. This mirrors precisely the situation at hand: a paraphrase or basic inference supplanted by a creative reinterpretation that reveals deeper moral or theological insights. The homily is explicitly presented as a second tier of interpretation, emerging when the straightforward meaning faces complications or is revealed as inadequate. It is this second level that bestows the text with its full significance. The dismantling of the common midrash/allegory dichotomy might thus be another reward of the painstaking terminological search.
At the same time, the conventional hermeneutical analysis of Midrash is not discredited altogether. For the shared formal trait also underscores a crucial differentiating aspect of ella drashot: their steadfast commitment to the realm of actuality. The “as if” lesson doesn’t elevate these midrashic units to a loftier, spiritual dimension. Rather, the figurative twist allows the darshan to anchor the message in the real world: thus the “soul that [Abraham and Sara] did” (Gen. 12:5) is reread as conversion (real conversion of real people!) just as “a horse and its rider” (Ex. 15:1) is reread with the “as if” technique but remains in the tangible realm of actual horses and riders while also becoming a sign of divine providence. Rather than spiritualization, we have a moral turn. Tannaitic midrash seeks to have its cake and eat it too, and this is exactly what this structure enables.
A complementary line of research is to look for terms that function in both halakhic and aggadic contexts and compare their function. Here is an example (forgive me for the abundance of examples, it is not only because I had a hard time leaving out good stuff; it is because studying Midrash without constant examples is the closest thing to learning to swim on the sand): the term “was it not already said” (והלא כבר נאמר; ve-halo kvar ne’emar) appears a hundred times in the Mekhilta RI, conveniently split to half halakhic and half aggadic appearances. But it functions in these two contexts differently: always marking contradictions in narrative contexts and redundancies in legal contexts! Why? Because legal contradictions are usually solved without any explicit notification, through a dialectical move in which one verse “corrects” what can be learned from the other. As in the following case:
“And he shall serve him forever” (Ex. 21:6 [the pierced Hebrew slave])—until the Jubilee year. [. . .] Or is “and he shall serve him forever” as it sounds? It therefore teaches, “[It shall be a Jubilee for you,] and each man shall go back to his ancestral holding” (Lev. 25:10). (Mekhilta RI, Nezikin 2)
Instead of marking the verses from Exodus 21 and Leviticus 25 as contradictory, and looking for a compromise, one verse is used to “correct” the reading of the other. Explicit presentation of contradictions is far more prevalent in aggadic contexts, apparently because there was less sensitivity outside of halakhah to admit that contradictions exist. We (this is not a royal we, but a work done with my brother, Assaf, who is also a Talmudist)16 thus suggest that the darshanim code switched when they moved from halakhah to aggadah, and that terminology can show that.
But terminology is not the whole story. Another line of study in uncovering midrashic structure is the analysis of its fundamental narratological concepts, especially that of the addressees. Narratologists distinguish between three types (and two subtypes) of rhetorical addressees (and here I am mixing the terminologies of two generations, the classic narratological chart of Chatman and the more recent rhetorical poetics of Phelan): 1) concrete internal addressees engaged in character-character dialogue within the narrative world of the plot (i.e., God, Moses and Aaron, or the Israelites in Egypt); 2) concrete narratee—a specified figure addressed, either explicitly or implicitly, by the narrator; 3) abstract internal narratees whom the narrator addresses implicitly (aka implied/ideal reader or narrative audience); 4) authorial audience, the interpretative community the author(s) have in mind while writing and to which the text is intended; and 5) concrete “flesh and blood” external audiences (for example, the darshanim, who are by definition non-ideal, and cannot be identical with the authorial audience).17
The Midrash, we (again, an actual we, this work was done with Yaakov Kroizer, a student of literary theory)18 argue, systematically blurs the distinction between the characters, the narratee, the implied reader, and the actual reader. It assumes that the biblical text is addressed to it, just as it addresses the inner characters (like the Israelites in Egypt) and the original—imagined—authorial audience (conceived by the Midrash as those who received the Torah in Sinai).
Scripture is relevant for them, meant to teach and instruct them, should be fully understood by them, and legally binds them, just as for the ancient Israelites. Except for unique matters that the Torah explicitly describes as time-dependent (such as the Passover in Egypt), the assumption is that they are the direct recipients of the biblical commandments. The darshanim are unwilling to accept any distance between them and Scripture and seek to bridge this distance whenever it arises. They are unwilling to accept any detail that is known to the characters but not to them. To bridge these gaps is the mission of Midrash. What the Midrash misses due to its distance from the original addressees, it compensates through intertextual tools.
Here is an example:
As He swore to you. (Ex. 13:11)—And where did He swear? It is said, “And I will bring you into the land concerning which I lifted up my hand,” etc. (Ex. 6:8).
And to your fathers.—Where did He swear to your fathers?
On Abraham, what does it say? “In that day the Lord made a covenant with Abraham,” etc. (Gen. 15:18). On Isaac, what does it say? “Sojourn in this land,” etc. (Gen. 26:3). On Jacob, what does it say? “The land whereon you lie,” etc. (Gen. 28:13).
And shall give it to you.—You should not look at it as an inheritance from your fathers but as if it was given to you on that day. (Mekhilta RI, Pascha 18)
The drasha reads “As He swore to you” as a textual reference to something that has already happened and looks for the relevant verses that describe it. This is not a self-evident reading; one could assume that the oath is not recorded in Scripture, and we only hear about it from here as a flashback, analepsis. But that would compromise the midrashic assumption of the correlation between text and reality. What was said to our fathers (the oath) is also told in the stories about them in the book of Genesis. If it were not so, the verse would contain a reference only for the concrete internal addressees but not for the external ones, and that would have made “us” (namely, the midrashic sages) deficient addressees, which is inconceivable. The profit from finding the internal textual reference is thus twofold: theologically—we see both the promise and its fulfillment; and textually—we validate the Torah as a self-sufficient whole whose parts complete each other.
But to whom is the speech directed? Who are its addressees? Ostensibly, they are the concrete internal ones, the Israelites in Egypt. The Midrash reads Genesis against Exodus, a promise against its fulfillment. But in the last drasha, “that You should not look at it as an inheritance from your fathers but as if it was given to you on that day,” the distance between the internal addressees and external readers is blurred, and the verse is read as directly addressing the actual audience of the darshan. After all, there is no meaning to the expression “but [you should look upon it] as if it was given to you on that day” (the parallel in Mekhilta RSBI is even more explicit “but as if it is now given to you as a gift”) if it refers to the Israelites in Egypt who did not yet inherit the land. Only for the darshan and his audience is there a meaning to the historical imagination that recreates the entry into the land. You, says the darshan, and not only the Israelites in Egypt, are the descendants of the Patriarchs, and the promise addresses you as well.
Far from unintentional blurring of direct and indirect addresses, this seems like a deliberate shift from the historical Israelites to “us.”
Here is another example: Exodus 24:7 tells of “the book of the covenant” Moses read to the people at Sinai. What is this book? In an excerpt of rewritten Pentateuch from Qumran (4Q158, 4:6–8) we find an elaboration of this book of covenant referring to the covenant with the patriarchs. In contrast, Mekhilta RI (Bahodesh 3) assumes that the book of covenant is a part of our own Pentateuch. Menahem Kister notes the resemblance of the exegetical approach of the two works.19 But we should highlight the difference as well. There is a disagreement in the Mekhilta as to what the book of covenant is (“From the beginning of Genesis up to here” versus independent “commandments” etc.), but there is a consensus that it consists of parts of the Pentateuch itself.
This rabbinic insistence on using the Torah as the definitive source represents a fundamentally new approach, arising from a different conception of the biblical text.
According to the rabbis, the reference must point to texts available to us, and it cannot be supplemented creatively, as done in the rewritten Pentateuch from Qumran. Therefore, while the passage from Qumran creates a new text forged from different biblical passages, the drasha cites specific texts from the Torah to which our verse supposedly refers. This rabbinic insistence on using the Torah as the definitive source represents a fundamentally new approach, arising from a different conception of the biblical text. It positions the Torah as a definite, holistic work containing all answers, with midrashic techniques serving to uncover them.
This is even more radical in legal contexts. According to the midrashic view, when it comes to the legal parts of Scripture, there are no intermediary factors like the characters in the narrative parts. Although the commandments and laws are given within the narrative framework of the Torah, the darshan reads these laws independently, stripped from any context. If the halakhic message was seen as similar to the narrative one, the Midrash would have said something like: In the desert, the Israelites were commanded to observe the Sabbath, and we should learn from them, or: we are required to imitate them. But this is not the case. The halakhic strand of Midrash assumes that we are the direct addressees of the halakhic obligations, a concept which creates a hermeneutic divide between the midrashic attitudes to biblical law and biblical narrative.
Lastly, we get to politics (you did wait patiently for that). Scholars look for the rabbis’ constructions of distinctions, us versus them, their alterology. I would argue that here too tracing midrashic technique may help present a clearer picture. Adi Ophir and I dedicated a whole chapter to Tannaitic Midrash in our quest for the birth of the Goy,20 since we found there radically new structures.21 Form analysis shows that in contrast to the general tendency of Midrash to break ontological dichotomies (playfully remodeling and recreating collective and individual figures according to local textual needs), the binary relation between Jew and gentile is not shaken in Tannaitic Midrash. In fact, this dichotomy is presupposed and enacted throughout the midrashic text with neither irony nor problematization.
Here is an example: in the Tannaitic exegesis on the biblical narrative of the voyage from Egypt to Sinai (Ex. 13–19), found in both Mekhilta RI and RSBI, goyim are present mostly in their political and collective manifestation, as “the nations of the world.” But in most of these midrashic contexts, the “nations” are a unified entity, the eternal enemy of Israel.
This is nowhere clearer than in the “index drashot,” where the story of God’s war against Pharaoh and Egypt is attached to additional scriptural passages about other nations and heroes and concludes with a generalized lesson about the sins and punishments of the “nations.” Egypt becomes in these cases a synecdoche for all nations. These are a single and unified entity opposed both to Israel and to God. Here is how some of these indexes are summarized:
For with the very thing with which the nations of the world pride themselves before Him, He punished them. (Mekhilta RI, Shirata 2)
And woe unto the Nations of the World, what do they hear with their own ears, Behold, He by whose word the world came into being will fight against them. (Mekhilta RI, Shirata 4)
God showed the nations of the world how dear the children of Israel were to Him . . . But it is not enough that they do not treat them with respect, they even put them to death in all sorts of cruel and strange ways. (Mekhilta RI, VaYehi 1)
Egypt is just a model. The nations, all nations, are responsible for Israel’s dire status, and in the future, God will fight them all collectively.
Since the God of Israel is also the creator of the world, there is always the possibility of treating Israel and the nations in the same manner, as subdivisions of God’s single created universe. This obvious potential is mostly absent from the Tannaitic texts. The Nations are presented as a multiplicity that is quickly reduced to an indiscriminate totality. Any differences between them are negligible. The midrash can thus emphasize distinctions between different nations, according to local ideological and exegetical needs, but then “forget” them and return to speaking of a basically homogenous group. So, too, Israel is described as a diverse group of peoples within a basic unity.
Look at this drasha:
“And Egypt Pursued [in the singular!] after them” (Ex. 14:9)—teaching that none of them [the Egyptians] has stumbled [on the way], so that they could not augur [i.e. see it as a divine sign] and turn back [to Egypt, and so cannot be defeated by God]. And thus we find everywhere [in the Bible] that the gentiles practice augury, for it is written “Those nations that you are about to dispossess do indeed resort to sooth-sayers and augurs” (Deut. 18:14). “The elders of Moab and the elders of Midian set out with divinations in their hands” (Num. 22:7), “and Balaam son of Beor, the augur” (Josh. 13:22). (Mekhilta RI, VaYehi 3)
The Mekhilta bases its interpretation on the well-known stereotype of the Egyptians as magicians. But, before we can start speculating about the rabbinic conceptualization of Egypt, the drasha adds: “And thus we find everywhere that the gentiles [the goyim] practice augury.” And indeed, the verses cited there are about Canaanites, Moabites, and Midianites, but not Egyptians. Goyim are goyim.
Recent studies justly nuanced previous notions about the lack of rabbinic interest in history.22 But the rabbinic conception of a unity behind the multiplicity of the “nations” puts a structural limit on their historical sensitivity. The amalgamation of all nations into one abstract category of “nations of the world” erases historical particularity. The unfolding of the world in time becomes a scene for reiterations of various phases of the same formulaic relationship between God, the Jew, and the Goy.
So, what is midrash? we have identified two major characteristics that distinguish Tannaitic Midrash from other ancient modes of exegesis.
All ancient Jewish and Christian readers of the Bible sought instruction within its texts. The key question, however, is: instruction for what? Unlike Philo, who looked for guidance on the soul’s spiritual journey, or the authors of the Dead Sea Scrolls, who sought direction for the end of days, the early rabbis turned to the Torah for practical, mundane instruction. Their focus was on halakhah, daily teachings, and the aggadic lessons embedded in the everyday life of piety.
But the novelty of Midrash lies as much in its form as in its content. Its explicit use of set terminology . . . sets it apart as an unprecedented mode of study.
But the novelty of Midrash lies as much in its form as in its content. Its explicit use of set terminology—much like the use of Lemma, which distinguishes between the quoted verse and its interpretation—sets it apart as an unprecedented mode of study. No prior tradition systematically laid out both the procedure and technique of interpretation as the midrashic method did. Students were not merely given fish; they were taught how to fish for themselves.
More specifically, the study of terminology yielded a series of transformations from the common conception of Midrash. First, the Midrash proves to be more structured and less chaotic than it initially seems. It follows a relatively limited syntax in its core structures, though it is rich in variations. Its rules must be carefully studied, and we cannot be satisfied with broad generalizations about hermeneutic freedom. Second, the perceived tension between tradition and Midrash—often framed in scholarship as a dichotomy between midrashic formation of halakhah versus preservation of existing traditions—is misleading. In reality, there is a dynamic interplay in which both Midrash and halakhah evolve in response to one another. Third, halakhah and aggadah each operate with distinct hermeneutic assumptions, making the midrashic movement between law and narrative in the Pentateuch a kind of code-switching. Fourth, certain conventional dichotomies—such as that between Midrash and Hellenistic allegory—are found to be overly simplistic. Fifth, intertextual tools are the rabbinic way of “being there,” midrashic textuality compensating for the actual distance between the implied and the actual addressees. Sixth and lastly, midrashic intertextuality tends to blur distinctions among biblical political entities, treating them as examples within a single midrashic framework.
Many of these techniques have precedents in pre-rabbinic biblical exegesis, but their synthesis is novel, as is the explicit use of terminology to articulate it.
Now, all these points differ, to a greater or lesser extent, from traditional views of Midrash, often found in introductory works—and from my own earlier perspective.
What I couldn’t discuss here—and will only hint at—is that my study of terminology has led me to reconsider the significance of the distinction between R. Ishmael and R. Akiva in their basic hermeneutic assumptions. I have found that the differences are smaller than scholars often assume, manifesting more in scale and intensity than in fundamental principles.
For me, all this mainly means that we need to continue the search; after all, the verb “darash” means nothing but “to seek.”
Notes:
- Midrash (lit. = study, from the root darash) is the collective name for the body of rabbinic exegetical works from late antiquity (plural form = midrashim). It is used also in reference to the approach and worldview reflected in the different midrashim. These works are the first running commentaries on the Bible. Each exegetical unit is called drasha (singular, plural = drashot), the sage whose words are cited, whether by his name or anonymously, is called darshan (singular, plural = darshanim). The Hebrew transliterations are used here because commonly used translations—such as interpretation, homily, and comment—do not carry the full meaning of the textual phenomenon of Midrash.
- Abraham Geiger, “Einiges über Plan und Anordung der Mischnah,” Wissenschaftliche Zeitschrift für jüdische Theologie 2 (1836): 474–92, 481. See also: Jay M. Harris, How Do We Know This? Midrash and the Fragmentation of Modern Judaism (State University of New York Press, 2012).
- Isaac Heinemann, Darkhei ha-Aggadah (Magnes Press, 1949).
- Susan A. Handelman, Slayers of Moses: The Emergence of Rabbinic Interpretation in Modern Literary Theory (State University of New York Press, 1983).
- David Stern, Midrash and Theory: Ancient Jewish Exegesis and Contemporary Literary Studies (Northwestern University Press, 1996).
- James L. Kugel, Early Biblical Interpretation (Westminster Press, 1986); and Traditions of the Bible: A Guide to the Bible as It Was at the Start of the Common Era (Harvard University Press, 2009).
- Daniel Boyarin, Intertextuality and the Reading of Midrash (Indiana University Press, 1995).
- Azzan Yadin-Israel, Scripture as Logos: Rabbi Ishmael and the Origins of Midrash (University of Pennsylvania Press, 2004); and Scripture and Tradition: Rabbi Akiva and the Triumph of Midrash (University of Pennsylvania Press, 2015).
- Maren Niehoff, Jewish Exegesis and Homeric Scholarship in Alexandria (Cambridge University Press, 2011); Yakir Paz, From Scribes to Scholars: Rabbinic Biblical Exegesis in Light of the Homeric Commentaries (Mohr Siebeck, 2022).
- I.e., law (n., adj. = halakhic), in contrast to aggadah (n., adj. = aggadic), which is the term for the non-legal, homiletic parts of Midrash, which correspond in many cases to the narrative parts of the Pentateuch. For a discussion of halakhic reasoning and moral considerations in Tannaitic Midrash, see Moshe Halbertal, Interpretive Revolutions in the Making (Jerusalem: Magnes Press, 1997).
- Midrashic sources are cited according to “Maagarim”—the historical dictionary project of the academy of the Hebrew Language. All translations are mine. All emphases in bold and explanations in square brackets in the sources are added by me. In the case of the Mekhilta RI, I have consulted the translation of Jacob Zallel Lauterbach, Mekilta De-Rabbi Ishmael (Jewish Publication Society, 2004). In the case of the Mekhilta RSBI, I have consulted W. David Nelson, Mekhilta De-Rabbi Shimon Bar Yohai (Jewish Publication Society, 2006l).
- Lit.=house of study, the place where midrashic activity took place, not to be confused with bet-kneset, i.e., the synagogue.
- Handelman, Slayers of Moses, 75.
- Boyarin, Intertextuality, 110.
- David Dawson, Allegorical Readers and Cultural Revision in Ancient Alexandria (University of California Press, 1991).
- Assaf Rosen-Zvi and Ishay Rosen-Zvi, “Tannaitic Halakhic and Aggadic Methodology,” Tarbiẓ 86 (2019): 203–232, and “The Hermeneutics of Aggadic Exegesis in Tannaitic Midrashim: A Terminological Survey” [Heb.], Talmudic Studies, vol. 4, ed. Yoav Rosenthal and Shlomo Naeh (Mandel Institute of Jewish Studies, 2024), 765–816 .
- Seymore Chatman, Story and Discourse: Narrative Structure in Fiction and Film (Cornell University Press, 1978), 147–151; James Phelan, Somebody Telling Somebody Else: A Rhetorical Poetics of Narrative (Ohio State University Press, 2017) 7–8.
- Ishay Rosen-Zvi and Yaakov Kroizer, “Throughout your Generations”: Toward Poetics of Tannaitic Midrash” (forthcoming).
- Menahem Kister, Dynamics of Midrashic Traditions in Second Temple and Rabbinic Literature (Magnes, 2024) [Heb.], 70.
- I.e., gentile, non-Jew; plural = goyim.
- Adi Ophir and Ishay Rosen-Zvi, Goy: Israel’s Multiple Others and the Birth of the Gentile (Oxford University Press, 2018).
- Amos Funkenstein’s critique of Y. H. Yerushalmi is one example of this. See Amos Funkenstein, Perceptions of Jewish History (University of California Press, 1993): 1–21; and Yosef Hayim Yerushalmi, Zakhor Jewish History and Jewish Memory (University of Washington Press, 1996).
Ishay Rosen-Zvi teaches rabbinic literature at the Department of Jewish Philosophy and Talmud at Tel-Aviv University and is a research fellow at the Shalom Hartman Institute in Jerusalem. He has written on issues of hermeneutics, self-formation and collective identity in Second-Temple Judaism and rabbinic literature. Among his publications are: Goy: Israel’s Others and the Birth of the Gentile (with Adi Ophir) (Oxford University Press, 2018) and Between Mishnah and Midrash: The Birth of Rabbinic Literature (Open University, 2019). This year he is the Gerard Weinstock Visiting Lecturer of Jewish Studies and a Harry Starr Fellow in Judaica at the Center for Jewish Studies, Harvard University.
Please follow our Commentary Guidelines when engaging in discussion on this site.
