This review essay is part of the 2023 Society of Biblical Literature's review panel for Yael Fisch, Written for Us: Paul’s Interpretation of Scripture and the History of Midrash. Find the full panel here.
In this fascinating book, Yael Fisch proposes to find a way to think comparatively about Paul and midrash without assuming a direct dependence between Paul’s letters and early rabbinic (tannaitic) interpretation or assuming a radical divorce between them. She is not the first comparativist to buck against the constraints of the genealogy vs. analogy paradigm; this problem is a mainstay of comparative studies in all fields.[1] Fisch herself acknowledges that scholars comparing Paul and Qumran have transcended this paradigm when investigating terminology, eschatology, and law (23). Her contribution is (1) to establish a three-way comparison – Paul, Qumran, and tannaitic midrash – and (2) to do so on topics pertaining to hermeneutics. She is interested in showing how Qumran, Paul, and the rabbis use common methods of scriptural interpretation for disparate communal and ideological ends (26). Fisch argues that thinking about Paul and midrash together can advance and refine our understanding of the history of midrash (14); expose specific reading and rhetorical strategies as employed by each of these bodies of literature; and test and refine our definitions of various interpretative methods -- specifically pesher, allegory, and midrash.
The book is organized as a series of three case studies that are chock-full of terrific insights. Chapter 1 focuses on what Fisch refers to as midrash-pesher, a shared method of interpretation deployed to different ends in Qumran, in tannaitic literature, and on one occasion in the letters of Paul, in Romans 10. The chapter shows – brilliantly and convincingly -- how Paul uses the rhetorical form of midrash-pesher to construct righteousness by pistis (faith or trust) as a teaching grounded in Scripture, right alongside righteousness by works of the law, rather than over-against it.
Chapter 2 examines the allegory of Hagar and Sarah in Galatians 4, the only place in Paul’s letters where the term allegory appears. Fisch argues that Paul uses allegory in Galatians 4 to re-interpret genealogy in a way that makes gentiles-in-Christ the descendants and heirs of Abraham. Fisch also uses Galatians 4 as an occasion to critique the view of midrash and allegory as diametrically opposed modes of interpretation and to rethink the history of Jewish allegory, points to which I shall return.
Having considered the interpretative methods of pesher in chapter 1, and allegory and midrash in chapter 2, Fisch turns in the 3rd chapter to consider the different – indeed opposing -- hermeneutical postures of Paul and Qumran on the one hand, and the early rabbis on the other. In 2 Corinthians 3 Paul represents the veil placed over Moses’ radiant face in Exodus 34 as interposing between revelation and the Israelites. According to Paul, Moses’ veil is the historical explanation for the inability of Israel to read Scripture properly. The remedy is faith in Christ which “unveils the gaze and the heart of the reader” (149) allowing a true understanding of Scripture. In contrast to Paul, whose hermeneutic of unveiling is a hermeneutic of the end time in which final meanings are revealed, the early rabbis did not see themselves as the final generation. They espoused a hermeneutic of modesty that resisted final meanings. Indeed, they expressly disapproved of hamegaleh panim batorah, the one who unveils the face in connection with Torah, i.e., who offers unrestrained and heretical interpretations.
Despite her suspicion of “genealogical accounts,” Fisch suggests the following dialectical interaction between Paul’s ideal of unveiled reading and the rabbis’ ideal of interpretive restraint. Paul likely transformed a negatively charged Hebrew idiom -- megaleh panim batorah -- into a positively charged idiom connoting not a heretical reading but a full and final understanding of Scripture. The rabbis rejected the Pauline revision of the idiom and the unrestrained hermeneutic it supported, and (re-)asserted the idiom’s negative valence and the ideal of interpretative modesty. This is an example of what Shaye Cohen has described as an “antipodal” situation. In an antipodal situation, two authors or texts agree on the facts but have opposite evaluations of those facts as good or bad.[2] So here, Paul and the rabbis share a view of Jewish interpretation of Scripture as restrained. For Paul, however, this is a failing while for the rabbis, it is a virtue: antipodal assessments of the selfsame phenomenon.
Even though her goal is the comparison of Paul and the rabbis, Fisch wisely realizes in these comparative analyses that the two are not best studied as an isolated pair. Paul and the rabbis are but two dots on a larger map that includes many mutually illuminating dots. The more dots we juxtapose in our comparison the more the contours and significance of any individual dot becomes apparent.
But I’d like to suggest a slight variation on this technique of comparison that can shed even more light on the topics investigated by Fisch. Scholars as diverse as Katell Berthelot, Simcha Gross, Richard Hidary, and Matt Thiessen have proposed seeing each “dot” on our map as responding to some problematic arising from the dominant Hellenistic or Roman imperial context. For example, Katell Berthelot compares and contrasts late antique Jewish responses to Roman imperial ideology which positioned Rome, not Israel, as a divinely chosen and virtuous people destined for a universal and eternal rule through a superior legal order to the world. This ideology resembled and threatened Jewish self-conception, and evoked an array of Jewish responses, adaptations, and countermodels. My own study of ancient conceptions of divine law considers the rabbis as one group among many groups who were all responding to conceptions of divine law that permeated the dominant culture. Rather than responding only or even primarily to one another, these groups can be seen as working out varying responses to a shared set of Greco-Roman discourses about divine and human law ubiquitous in late antiquity and incompatible with the biblical representation of divine law.
I refer to this comparative technique as “triangulation” as if the broader cultural problematic – Roman imperial ideology, Stoic conceptions of divine law – sits at the apex of a triangle, stimulating and conditioning the responses of various sub-groups arrayed along the base of the triangle. As these sub-groups formulated their responses to the shared cultural problematic at the apex of the triangle, they were also often attuned to and seeking to differentiate themselves from one another.
Thus, Fisch traces several variations on the notion of a double Torah in late antique Jewish sources (166). Now if we were to adopt the model of triangulation, we would see that Paul, the rabbis, as well as the other Jewish writers and groups reviewed by Fisch (Jub, 4 Ezra, Philo) were all primarily responding to Greco-Roman discourses of a double law stretching back to Plato, even as they also kept an eye on one another’s response.
Ambivalence about written law can be found in Plato’s dialogue Protagorus (320c-324c): written human law rescues humans from the chaotic state of nature but it cannot eliminate secret wickedness and produce true virtue because virtue is based not on coercion and fear of punishment but on love of the good. Elsewhere Plato says that written laws cannot inculcate virtue without a savior (a soter) which is logos (reason).
Plato’s ambivalent assessment of even the best written law is a mainstay of classical thought into the Roman period, with philosophers like Xenophon, the Pythagoreans, the Stoics, Plutarch contrasting the lifeless and enslaving law that is written on wooden tablets and that fails to bring true virtue, with the superior living law (the nomos empsuchos) of the ideal king or sage who is himself an embodiment of the unwritten higher law or “living reason”[3]
This Hellenistic denigration of written Law as incapable of bringing righteousness posed a serious challenge to Jews whose divine law took the form of written legislation. They responded to this challenge by formulating their own diverse versions of a double nomos, most of which defied the Hellenistic conception by privileging writtenness (e.g., Jubilees and 4 Ezra). Philo, however, followed the Hellenistic model by downplaying the writtenness of Israel’s divine nomos, casting it as a mere copy of an unwritten archetype embodied in the patriarchs. He describes Abraham as an animate law and an unwritten statute, recalling the Pythagorean concept of nomos empsuchos, the living law superior to the law written on wooden tablets.[4]
The rabbis also engaged this larger cultural context (as attested by terminological and conceptual echoes), when developing their distinctive idea of a double law, oral and written, even as they differentiated themselves from earlier Jewish responses to the same cultural problematic.
And of course, Paul’s double nomos described so well by Fisch, can also be profitably framed first and foremost as a response to Greco-Roman ideas of a double law rather than a response to an alleged Pharisaic notion of oral and written Torah. In 2 Corinthians 3, Paul draws on a robust and centuries old tradition when he contrasts the law written on tablets of stone and associated with a lack of righteousness, slavery, death, with the law written on the heart and associated with righteousness, freedom, and life. The written Law was good insofar as it rescued the Israelites from the worst sins, he argues, but like all written law it could not bring full righteousness. For this, law needs a soter – just as Plato had claimed.
I have argued that Paul adopted the ambivalent discourse around written law in his letters to ex-pagan gentiles because the Hellenistic dichotomy between an unwritten nomos that brings perfect virtue and a written nomos that does not was familiar and meaningful to them and because it served Paul’s purpose perfectly.[5] He wanted the nations to join Israel without joining Israel in full observance of the commandments because he was not an inclusivist but an exclusivist who believed the commandments to be a privilege given to the biological seed of Abraham only. This is why he denigrates the written Law and the righteousness of works but only for ex-pagan gentiles, not for Israel in the flesh.
Fisch is eager to find a way to study Paul and midrash together without resorting to genealogy. The model of comparison she offers aligns with the technique of comparison proposed by Jonathan Z. Smith. For Smith, juxtaposition is based on a point of similarity. Similarity is the precondition that makes comparison possible, while the resulting identification of difference is what makes comparison interesting.[6] Smith also reminds us that the principle of juxtaposition, the similarity that determines what will be and will not be included in the comparison, does not inhere naturally in the items compared, but is generated by the comparer’s arbitrary and creative interests.
Fisch states her principle of juxtaposition plainly. She juxtaposes Paul’s letters, Qumran literature, and midrash because each uses Scriptural hermeneutics to build a sacred community. This principle of juxtaposition is as reasonable – but also as arbitrary -- as any other principle.[7] It enables Fisch to compare the hermeneutical methods of the three bodies of literature in order to identify their differences. She begins with a hermeneutical technique she calls midrash-pesher, a deictic technique that uses third person pronouns (zeh, zo, eleh) or specific technical terms to decode successive elements in a scriptural verse as references to some contemporary reality (A=X, B=Y). As Smith also reminds us, it is the resulting identification of difference that makes comparison interesting and Fisch uncovers the following very important differences: At Qumran and in Romans 10, this deictic technique is used to contemporize eschatological prophecies in a univocal manner. In tannaitic works, however, the technique is non-eschatological. It contemporizes in the opposite direction, linking scriptural events or realities of the past with contemporary, tannaitic reality. Moreover, it strips the pesher form of its univocity.
These significant differences led me to question the utility of the term “midrash-pesher” as an umbrella term for such radically different applications of a shared rhetorical form. Fisch lays out the differences beautifully and clearly, but then conceals her remarkable findings by applying the single label “midrash-pesher” to these three bodies of literature. Yet the fact that the three share this rhetorical form is the precondition for their comparison, not its end result. Comparison cannot reassert the sameness that justified the comparison to begin with; the point is to foreground the differences made visible by the comparison. So I would suggest that a more accurate representation of Fisch’s findings would be conveyed by abandoning the term “midrash-pesher” as a common label for such radically diverse applications of this hermeneutical form and instead referring to three types of pesher: Qumranic pesher, Paulinian pesher, and rabbinic pesher. Doing so communicates clearly that there is a point of similarity that justifies their comparison (all three use pesher, a deictic exegetical method that contemporizes Scripture in some way). At the same time, adding the qualifiers Qumranic, Philonic, and rabbinic to the word pesher communicates that there are differences in the way they deploy this shared hermeneutical technique: differences in their underlying assumptions of univocality or multivocality and differences in the direction of their contemporizing (from eschatological future to the present or from authoritative past to present). In fact, foregrounding these differences helps us see that the later rabbinic use of the pesher technique may be best understood as a counter-hermeneutic resisting earlier eschatologically driven and univocal hermeneutics. And this supports Fisch’s conclusions in chapter 3 contrasting Paul’s eschatological hermeneutic of unveiling final meanings with the rabbis’ non-eschatological counter-hermeneutic. Unfortunately, this important point is obscured if we level the differences uncovered in chapter 1 by applying a single label: midrash-pesher.
Turning now to Fisch’s comparative analysis of allegory and midrash: Fisch argues against two claims advanced by Daniel Boyarin. First, that allegory and midrash are not merely different but diametrically opposed hermeneutical techniques, allegory connecting textual signifiers to abstract signifieds in a univocal way and midrash linking textual signifiers to other textual signifiers and, through the multiplication of intertextual links, producing polysemy. Second, she objects to Boyarin’s characterization of Paul’s exegetical method as allegorical and thus diametrically opposed to midrash.
Against the first claim, Fisch argues that midrash and allegory cannot be thought of as binary opposites since we find allegory in tannaitic midrash. But I’m not persuaded that the six passages designated by Fisch as tannaitic allegories are in fact allegories. They do not move from a concrete, historically embedded plane of meaning to an abstract, ahistorical plane of meaning. Two passages decode wood and water as references to the Torah of Israel to say that the historical people of Israel gathered at Marah failed to observe Torah. This is a concrete, historically-embedded decoding not an abstract decoding. A passage that presents the exertion of collecting manna as fulfilling the curse that man shall earn his bread by the sweat of his brow is also not an allegorical decoding, but a classic intertextual reading of one text through the lens of another. A passage identifying various priestly gifts as rewards earned by the zealous deeds of Pinhas is also not an allegorical abstraction, but a classic midrashic use of the measure for measure principle to explain why the priestly class deserved certain perquisites – they were rewards for Pinhas’s actions.[8]
But even if we grant Fisch’s claim that these few texts are allegorical, it’s not clear that we can refer to them as tannaitic allegories because as Fisch herself acknowledges they are attributed to a sub-group known as dorshe rashumot (“expounders of hidden things”), distinct from the tannaim themselves. So can we refer to these six homilies as “tannaitic” or “rabbinic” allegories” if they do not move from concrete signifiers to abstract signifieds and, even if they did, they are only preserved by the tannaim rather than produced by them. Similarly, I wasn’t persuaded that the three examples from Qumran blend elements of allegory and midrash. To take just the first example, a passage in Nahum Pesher decodes the sea and certain geographical places not as abstractions but as symbols for concrete historical enemies of God (the Kittim, various haughty leaders of other nations). Not every symbolic reading is an allegorical reading; not unless it leaves the plane of the concrete.
But I am convinced by Fisch’s claim that Paul’s allegory in Galatians 4 is a genuinely hybrid text that challenges the dichotomy between allegory and midrash. Paul invokes the Greek term for allegory and the content of his interpretation bears some relation to Philo’s allegorical reading of the Hagar and Sarah narrative. At the same time, Paul argues by intertextual inference from the prophets to the Pentateuch in a manner typical of later rabbinic midrash. So here I agree, we do have an interpretation that blends key elements of allegory and midrash.
And now an important question arises: what do we do when a text defies our genre categories by combining elements of more than one genre? Do we dismantle our categories despite their utility in literally hundreds of other cases, or do we recognize that our categories are not natural kinds but heuristic devices that will occasionally be violated in real life? I prefer the latter path: well-drawn, empirically supported categories are useful if we do not reify them. They help us sort our data; they help us identify the few odd cases that straddle the boundaries (so we can explore the reasons for those few violations).
Elsewhere, Fisch herself seems to recognize the utility of these categories. Her careful account of the history of allegory on p. 128 highlights the important differences between allegory and midrash. As she explains: allegorical interpretation of scripture was born in 2nd c B.C.E. Alexandria, to solve textual problems by introducing Greek philosophical ideas into Scripture. Jewish literature in 2nd Temple era Palestine contains no allegorical interpretation, with the possible exception of three examples from Qumran pesharim and six examples preserved in tannaitic literature as marginal homilies attributed to a closed group of non-tannaim, but even these are never employed toward a philosophical end. Allegory – to the extent that it existed in Palestine -- was abandoned as a living exegetical practice in the late 1st c C.E. Allegory, Fisch notes, is simply not produced by tannaim. (And I’ll just add that I was confused by the term “midrashic allegory.” It was the label for an allegory that employs intertexts. But all the examples are pre-rabbinic. The problem is that many people -- right or wrong -- associate the word midrash not just with intertextual reading but with rabbinic intertextual reading. Given that the rabbis abandon allegory, it might be best to refer to these pre-rabbinic cases as “intertextual allegories” rather than midrashic-allegory, to prevent confusion.)
After reading Fisch’s book I am convinced that Paul’s general hermeneutic should not be identified as a radicalization of Alexandrian allegory, or as allegory at all. And I can accept, based on Paul’s blend of the intertextual method featured in later rabbinic midrash with the terminology and content of allegory in Gal 4, that allegory and midrash are not always diametrically opposed, at least for Paul. Nevertheless, as Fisch herself recognizes and details, allegory and midrash differ in numerous ways. Moreover, they are not blended in the vast majority of works of ancient Jewish interpretation or in rabbinic literature, which suggests that their distinction as hermeneutical systems has heuristic value.
I conclude with Fisch’s affirmation of the importance of difference. She writes, “When allegory is defined too broadly--like in Quintillian’s definition, as saying of one thing and meaning another--differences between distinct systems of hermeneutics (such as hermeneutics of the pesharim, the allegorical interpretations of Philo, Paul’s interpretation of scripture and midrashic hermeneutics) collapse” (116). And, I would add, when differences collapse, then comparison – as Smith reminds us -- has neither motivation nor interest.
Dr. Christine Hayes is Sterling Professor of Religious Studies Emeritus in Classical Judaica from Yale University.
[1] For example, in the study of parallels in the various works of the rabbinic corpus, the study of parallels in Jewish and Christian sources, or the study of biblical reception generally.
[2] Shaye J. D. Cohen, “Antipodal Texts: B. Eruvin 21b–22a and Mark 7:1–23 on the Tradition of the Elders and the Commandment of God,” Near Eastern Languages and Civilizations, Harvard University, preprint, 2013.
[3] Diotogenes (1st c B.C.E.-1st c C.E.), On Kingship; Plutarch (1st-2nd c C.E.), To an uneducated ruler, 780c.
[4] The unwritten archetypes are clearly superior to their written copy. The unwritten commands that Abraham followed, Philo says, were “manifest by nature with clearer signs, and apprehended by the sense which is the most truthful of all and superior to hearing, on which no certain reliance can be placed.” Philo follows Greek precedent in ranking the written law lower than the unwritten law (pace Fisch).
[5] As the apostle to the Gentiles, Paul addresses his ex-pagan audience on their own terms. Given that Paul states explicitly that he adapts his behavior to the behavior of those around him, behaving like a Gentile when with Gentiles (1 Cor 9:19-23), it is not difficult to imagine that he adapted his rhetoric to those around him as well. We need not go as far as Nanos (2009:17) and argue that this is the intended meaning of 1 Cor 9 – that Paul argued like one outside the Law when with those outside the Law. Nevertheless, the adaptation of behavior to conform to context renders more plausible the suggestion that Paul adapted his rhetoric and conceptual framework to conform to his context and audience.
[6] This is the approach taken by Shinaan and Zakowitch in their comparative study of biblical interpretation, as well as other scholars of biblical reception. Scholars differ in the range of works they choose to include in their study.
[7] The principle serves to exclude Philo and Josephus, for example, who while engaging in Scriptural interpretation do not aim to build a sacred community.
[8] Similarly, the technical exegetical term nimshalu be- used in some tannaitic texts does not introduce an allegorical reading, but a simple analogy between two entirely concrete entities (such as fire and Torah) based on perceived empirical similarities. The translation of nimshalu be- as “symbolized by” is untenable: fire is not presented in this derashah as a stand-in for Torah. The more literal translation of nimshalu be- as “is likened to” is to be preferred, because the figure is a simple simile likening Torah to fire in order to elaborate on its qualities: like fire, Torah descends from heaven, is capable of bringing benefit or harm, and so on.