Nominalist “Judaism” and the Late-Ancient Invention of Religion
DANIEL BOYARIN
In the scholarly literature as it stands today, there are two almost directly opposed stances to the category “religion,” especially religion as a form of identity. On the one hand, we find still many—if not most—writers writing as if “religion” were an essential category, something as essentially human as language or walking, and the only relevant question being what kind of religion any given human group or human individual “had” or adhered to, usually expressed in terms of his (sic) beliefs or faith (or, occasionally, whether it is possible for human beings to survive without any religion at all). On the other hand, we find too, although not quite as frequently, a recognition that religion is a historical and thus historicizable category, not an essence but a particular kind of institution, something like literature or science and thus no more universally applicable to all human societies than they are. Typically, for scholars holding this view, religion is not found before the Enlightenment. Thus Jonathan Z. Smith wrote famously that religion “is solely the creation of the scholar’s study” and hence it “has no existence apart from the academy,”1 while Talal Asad declared that there can be no “universal definition of religion … because the definition is itself the historical product of discursive processes.”2 Another slightly earlier Smith, Wilfred Cantwell, wrote famously, “I have not found any formulation of a named religion earlier than the nineteenth century.”3 I would suggest that the subject in question in these authors is not religion so much as the concept of religions, the thesis that there are different species of a genus, religion, disembedded from other practices and discourses that make up cultures (whether that genus is held to be a universal possession of humans or not), different answers to a question phrased: What religion are you; or, to what religion do you adhere?4 This genus is usually defined as having gods, beliefs (= faith), and various practices of service to those gods. The claim made by Cantwell Smith is that this genus itself is only a product of modern analysis, whether ecclesiological or anthropological. My own researches and reflections have led me to a third position, one like the Smiths and Asad in that I, too, regard “religion” as a culturally specific, nonuniversal, historicizable category, but unlike them in that I consider “religion” to be an essential category of Christian self-understanding and of the constitution of Christianity itself and thus a product of late antiquity and not of modernity or the Enlightenment. I wish to propose here that while, indeed, “religion” may be the product of the scholar’s study, it was a late-ancient scholar (or rather group of scholars) who created it and that consequently studying the genealogy of the historical product that is the definition of “religion” is, contra Asad, a study of late-antique (and not only modern) discursive processes. That episcopal scholar’s study, then, had enormous impact on discursive processes that made a significant difference (for good or ill) in the lives of millions of human beings.
IOUDAISMOS AND CHRISTIANISMOS BEFORE RELIGION
One of the first questions to be asked in claiming that there was no religion before Christianity is the question of Judaism, since it is typically reckoned, both by lay folk and by scholars, that the religion, Christianity, grew out of and away from a prior religion, Judaism. Indeed, it almost seems counterintuitive to even think of denying such a proposition. Yet, to start thinking of such a denial, it seems highly significant that there is no word in premodern Jewish parlance that means “Judaism.” When the term Ioudaismos appears in non-Christian Jewish writing—to my knowledge only in 2 Maccabees—it doesn’t mean Judaism the religion but, as Steve Mason has demonstrated definitively, loyalty to Judaic practices in response to Hellenizing, behaving as a Greek.5 After that, it is used as the name of the ascribed Jewish religion only by writers who do not identify themselves with and by that name at all, until, it would seem, well into the nineteenth century.6 It might seem, then, that “Judaism” has not, until some time in modernity, existed at all, that whatever moderns might be tempted to abstract out, to disembed from the culture of Jews and call their religion, was not so disembedded and did not ascribe particular status by Jews until very recently.7
In order to even make sense of this claim, I need first to articulate the meaning or rather the possible meanings of the counterclaim, that there was such a thing as Judaism in the first century. I can think almost of no better place to start than with the realist/nominalist divide. This, as is well known, consists of an argument as to whether or not universals are “real” objects or only collections of particulars to which our minds/languages ascribe a name (hence nominalism). The argument in favor of the existence of Judaism would have to be a realist one, asserting that the practices, corporeal, verbal, and textual, that Jews engaged in were instances of a real universal, Judaism, and indeed that some of those practices were legitimately so and others not. I would assert that notwithstanding the position taken about such universals in the nonhuman world, such as “green,” for human products or categories such as “religion” or “Judaism” only a nominalist position is tenable. Since there is no way of even characterizing the particulars that would enter into the category “Judaism,” there is no possible universal in the world to which it could apply. Unless we seriously entertain the notion of ideas in the mind of God, we must recognize that there can be no thing “Judaism” until it is, in fact, a social fact and thus named as such by humans.8 It is the case, moreover, that it only exists for those humans for which it is named. Universals of this sort only exist insofar as they are named as such by a group of humans within a human language. If “Judaism” had no name for ancient, late-ancient, and medieval Jews, there was no such object in the world for Jews. If others, but not Jews, have a name for it, it is an object in their sociocultural world, not that of Jews.
The lack of existence of the word (for Jews) is a more or less established fact. If there was no word for it in the ancient Jewish conceptual world, without the assumptions of a realist ontoepistemology there can have been no object, for the only conceptual objects that exist are names. On the other hand, as we shall see below, “Judaism” comes to be named by Christians as such very early on to serve Christian discursive needs. Quite paradoxically, “Judaism” thus exists for Christians but not for Jews until modernity. This point can be documented from the third century on, as we shall observe in the research of Mason.9 There remains, to be sure, another possible, perhaps even plausible—maybe even unavoidable—use for the term “Judaism,” namely, to name an object in our conceptual world that helps us make sense of an ancient world that is not ours and does not know of this object.10 There just seem to be times when scholars (including me) need to abstract from the practices of the ethnic group, Jews, certain particular practices and group them as “religious” in order to make assertions about what was or was not the case, what was or was not possible within that abstraction, or to compare them to this or that set of practices by others. In this sense, “Judaism” is produced over and over in the scholars’ studies, and at the very least must be defined by each scholar who is using the term, or else he or she is likely to simply fall into whatever the nineteenth century defined as “Judaism” and assume it as a “real” category that is certainly an anachronism for Jews of the first, second, third, or fourteenth century.
Reiterating my point above, I would insist that a category cannot exist in a culture unless it is named as such, since categories have no existence other than as a human practice of category making. To designate that object and make it clear that it is only a product of my study in my work, I will always enclose the word “Judaism” in quotation marks until I can find a better neologism, one that does not imply in any possible way a “real” object, an object that is historical or ontological, before the term comes to be used by Jews as a name for their own “religion.” Taking my nominalism seriously, I insist that there is no category, “Judaism,” in the world but that it is a name we give to an aggregate of objects that we sometimes find analytically convenient to group together. One frequently encounters at this point an alleged counterexample, namely, the ancient economy. Even though, it is suggested, ancient folks did not have a word that is equivalent in extension to our economy, we nevertheless study economies in antiquity. This alleged counterexample, however, only strengthens my claim insofar as it is clear, even now, that an economy is a purely discursive object and thus also the product of economics as a discourse, a product of the scholars’ study just as much as I am claiming for “religion.” Ancient economies are thus the product of economic histories. This does not devalue such discourses, but they must be recognized for what they are, analytical tools and not entities in the world. This has, perhaps, more consequences for the study of “religion” than for the study of economics.
In my usage, moreover, when I choose to use it, the scholarly term “Judaism” will designate all the expressions of that which we now name “religion”: dealings with gods, sacralization of objects and people, beliefs and practices taken to be enjoined by sacralized people in the name of gods that are attested for the ethnic group called Jews in antiquity and the Middle Ages. Until and only when and where there is another name for people of the Jews who believe in Jesus, such people are also designated Jews in my work and their “religion” is, then, “Judaism,” which only means that insofar as it makes sense at all to speak of Jews having a religion, there is no reason for Jewish followers of Jesus to be left out of it. Again, the caveat: “Judaism,” when it refers to a time before it was named at all, is only a product of the modern scholar’s study; after it was named as such by Christian writers, it is an element within a Christian system of culture and language, not within the Jews’ at all; only when Jews start naming their own “religion” as “Judaism” could one say that a religion called “Judaism” exists for the Jews.
THE THIRD-CENTURY INVENTION OF “JUDAISM” BY TERTULLIAN
In his recent article, Steve Mason has decisively demonstrated that which other scholars (including me) have been bruiting about in the last few years, namely, that there is no “native” term that means “Judaism” in any language used by Jews of themselves until modernity, and, moreover, that the term Ioudaioi is rarely if ever used by people to refer to themselves as “Jews.”11 In a fascinating demonstration, Mason shows that the term Ioudaismos/Iudaismus only comes to mean “Judaism” in the mid-third century (with the Latin actually preceding the Greek), when the practices and beliefs of the Jews are separated polemically by Tertullian from their landedness, their history, “all that had made it compelling to Judaizers,” and Iudaismus means now “an ossified system flash-frozen with the arrival of Jesus.”12 This is not, of course, a historically accurate representation of the state of the Jewish people at the time (after all, a certain heyday of Palestinian Jewish life, the time of the Mishna), as Mason shows eloquently. His explanation for Tertullian’s new usage is equally convincing:
By about 200 C.E. the Church was making headway as a popular movement, or a constellation of loosely related movements. In that atmosphere, in which internal and external self-definition remained a paramount concern, Tertullian and others felt strong enough to jettison earlier attempts at accommodating their faith to existing categories, especially efforts to portray themselves as Judaeans, and to see commitment to Christ as sui generis. Rather than admitting the definitive status of the established forms and responding defensively, they began to project the hybrid form of Christanismus on the other groups to facilitate polemical contrast (σύνκρισις). The most important group for Christian self-definition had always been the Ioudaioi, and so they were the groups most conspicuously reduced to such treatment, which generated a static and systemic abstraction called Ἰουδαϊσμός/Iudaismus.13
The clear and critical conclusion to be drawn from this argument is consonant with my thesis in Border Lines that “Judaism” as the name of a “religion” is a product of Christianity in its attempts to establish a separate identity from something else that they call “Judaism,” a project that begins no earlier than the mid-second century and only in certain quarters (notably Asia Minor), gathers strength in the third century, and comes to fruition in the processes before and following the Council of Nicaea.
Thus, bearing out in much greater detail some of the philological points that I made with reference to the semantics of Ioudaismos in context with Hellenismos and Medismos, Mason entirely supports the conclusion drawn there (without, however, mentioning it) that there is no “religion” (or even way of life) called Ioudaismos in antiquity.
Where I disagree with Mason is in his acceptance of Wilfred Cantwell Smith’s conclusion that “early western civilization was on the verge, at the time of Lactantius [d. ca. 325 CE], of taking a decisive step in the formulation of an elaborate, comprehensive, philosophic concept of religio. However, it did not take it. The matter was virtually dropped, to lie dormant for a thousand years,”14 to which Mason comments decisively: “It is only western modernity that knows this category of religion.”15 In the next section of my argument, I will present evidence that Smith (and thus Mason) is wrong on precisely this point, that a robust notion of “religion” both existed and was necessary for the existence of a transethnic Christendom. Mason himself has given us the material for a hypothesis. First of all, to sum up, he has shown how by the third century Christian writers are using both Ioudaismos/Iudaismus and Christianismos/us to refer to belief systems abstractable from cultural systems, kinship networks, and coterritoriality. Second, he has argued that the later meanings of “religion”—the allegedly modern ones—are prepared for in antiquity by the concept of a “philosophy” as a system of beliefs and practices “voluntarily” adopted and maintained (which is not to imply that not adopting or maintaining such a system didn’t have negative consequences; “voluntary” here is not necessarily voluntary). These two elements—the latter one ignored by me till now to the detriment of my argument—I now strongly suggest led to a late-ancient development of something quite close to our modern notion of religion.
“NO RELIGION IN A SWAMP,” OR, RELIGION IN A FIFTH-CENTURY SCHOLAR’S STUDY
In early Latin, religio certainly did not mean anything like what we mean when we say “religion.” Carlin Barton writes:
Religio is, like pudor, a whole system of emotional, psychological, and behavioral responses to bonds and obligations and their transgressions. It is part of the internalized homeostatic systems by which the Romans governed themselves until the civil war period. It is in that period that religio becomes, in the works of Cicero, a disciplinary system, imposed by the state magistrates, authorized and reinforced by the gods. For Cicero, the threat to the state from the violence of the civil wars caused him to look for ways to reinforce the authority structures of the state. For him, more fear, more inhibition, more anxiety with regard to the gods and their spokesmen, the magistrates and priests, were necessary. So, while Cicero uses religio in all of its ancient usages, he often explicitly calls it the “cult of the [state authorized] gods” and makes the priests and magistrates to oversee and enforce that cult of which they are the most important part.
It was Cicero, and especially the book of Cicero’s widely-read by Christians, De natura deorum,16 that gave Minucius Felix, Lactantius and Augustine—and probably Tertullian and Ambrose as well, the word that would become “religion.”17
In other words, some (at least) of the important semantic features of our usage of “religion” were already in place in Cicero’s post–Civil War writing, and these were picked up and developed further by Latin Christian writers of the second to the fifth centuries. I, accordingly, find inconclusive Asad’s declaration that “When the fifth-century bishop of Javols spread Christianity into the Auvergne, he found the peasants ‘celebrating a three-day festival with offerings on the edge of a marsh.’ … ‘Nulla est religio in stagno,’ he said: ‘There can be no religion in a swamp.’ For medieval Christians, religion was not a universal phenomenon.”18 It may be that there can be no religion in a swamp, but this does imply that there can be religion other than Christianity, whether true or false. If he meant that there cannot be any Christianity in a swamp, he could have said that; I would suggest that he is implying rather that the practices of these swamp-dwellers don’t even rise to the level of “paganism.” In other words, I am suggesting that there are other criteria for the existence of “religion” as a concept than universality, criteria sufficient to mark the usage as very close, if not identical, to modern usage of the term. As David Chidester has amply demonstrated, Christians who explicitly did recognize the existence of other religions, “Judaism, Islam, and Paganism,” nevertheless were quite capable as late as the beginning of the nineteenth century of denying that the indigenes of southern Africa had any religion at all, even an idolatrous one.19 As he shows there, these early “ethnographers” would observe various ceremonies but insist that they were not “religion,” and by this they did not mean that they were not Christian.20 The concept of “religion” is not dependent, as is sometimes claimed, on the Enlightenment assumption that “religion” is simply a natural faculty of all human groups, that all humans have “religion.” Some humans may have “religion,” some may not, but “religion” in its modern sense of an organized and authorized system of beliefs and practices about gods not essentially tied to a particular ethnos or place already existed in Christian late antiquity.
By the end of the fourth century and in the first quarter of the fifth century, we can find several texts attesting how Christianity’s new notion of self-definition via “religious” alliance was gradually replacing self-definition via kinship and land.21 These texts, belonging to very different genres, indeed to entirely different spheres of discourse—heresiology, historiography, and law—can nevertheless be read as symptoms of an epistemic shift of great importance. As Andrew Jacobs describes the discourse of the late fourth and early fifth centuries, “Certainly this universe of discourses engendered different means of establishing normativity: the disciplinary practices of Roman law, for instance, operated in a manner quite distinct from the intellectual inculcation of historiography or the ritualized enactment of orthodoxy. Nevertheless, the common goal of this discursive universe was the reorganization of significant aspects of life under a single, totalized, imperial Christian rubric.”22
This construction of “Christianness” primarily involved the invention of Christianity as a “religion,” disembedded, in Seth Schwartz’s words, from other cultural practices and identifying markers.23 Susanna Elm shows that late-fourth-century Christians were already committed to the idea of “religions” in something very much like the modern sense and even understood quite well the difference between religious definition and other modes of identity formation. She finds evidence for this claim as early as Julian, “the Apostate” who formed his “religion,” Hellenism, in the 360s on the model of Christianity,24 but there is evidence for this concept going back at least as far as Eusebius in the first half of the century.25 Julian insists that only one who believes in “Hellenism” can understand it and teach it, as justification for his denial of the right to teach philosophy to Christian teachers. Vasiliki Limberis also emphasizes how, for all Julian’s hatred of Christianity, his notion of religion and religious identity has been deeply structured by the model of Christianity.26 As Limberis puts it: “Christians had never been barred from letters. Not only was this an effective political tool to stymie Christians, it had the remarkable effect of inventing a new religion and religious identity for people in the Roman empire.”27 I would slightly modify Limberis’s formulation by noting that Julian did not so much invent a new religion as participate in the invention of a new notion of “religion” as a category, as a modality of identity formation, and as a regime of power/knowledge. She writes: “In particular, Julian echoes Christianity’s modus operandi by turning pagan practices into a formal institution that one must join.”28
Although Julian seems never to have used any word that parallels our “religion,” his usage of Hellenismos, in its contexts, certainly seems to add up to that meaning. At any rate, the great fourth-century Cappadocian theologian Gregory Nazianzen so understood him and retorted to Julian:
But I am obliged to speak again about the word … Hellenism[:] to what does the word apply, what does one mean by it? … Do you want to pretend that Hellenism means a threskeia, or, and the evidence seems to point that way, does it mean a people, and the language invented by this nation.… If Hellenism is a threskeia, show us from which place and what priests it has received its rules.… Because the fact that the same people use the Greek language who also profess Greek religion does not mean that the words belong therefore to the threskeia, and that we therefore are naturally excluded from using them. This is not a logical conclusion, and does not agree with your own logicians. Simply because two realities encounter each other does not mean that they are confluent, i.e. identical.29
I have modified Elm’s translations here, substituting threskeia for the printed “religion,” so as not to prejudice the case, but it seems clearly correct to translate threskeia as “religion” in something quite close to its modern meanings. Gregory has some sort of definition of an object that comes very close to the modern usage of the concept, very like the modern concept of “religion” in mind here, distinct from and in binary semiotic opposition to ethnos, contra the commonplace that such definitions are an early modern product.30 In other words, it seems fair enough to consider his term threskeia as the semantic equivalent, or very close to that, of “religion” as it functions in our semantic systems. “Religion,” that nominalistic category, named in both Greek and Latin, had been invented by the fourth century at the latest by Christians, enabling them to see and identify a “religion” called “Judaism” within their semantic system as well, but not yet within the semantic system of the Jews themselves. Since the term threskeia is quite rare in Greek prior to this time,31 I suggest tentatively that it is a calque on the Latin religio in its latter meaning. Whether or not this is the case, I suggest that the catachresis in the language marks a catachresis in the (conceptual) world, an event that we might be tempted to regard as the invention of “religion.”
Gregory knew precisely “what kinds of affirmation, of meaning, must be identified with practice in order for it to qualify as religion”:32 it must have received its rules from some place (that is, some book?—Gregory surely doesn’t mean geographical locations, for then he would be playing into Julian’s hands) and some priests. He separates the language, Greek, from the religion of the Greeks and claims that it, as well as many cultural practices associated with it, including philosophy, now effectively secularized, is not tied to the threskeia itself. In my view, this explicit definition makes it eminently defensible to translate the Greek term by our modern “religion,” remembering that, for instance, there is not as yet, and won’t be for over a millennium, a word in Hebrew with any such meaning.33 While Gregory’s definition of “religion” is, of course, somewhat different from the Enlightenment Protestant one (a difference oddly homologous to the difference between Catholicism and Protestantism between rules of priests and a “faith”), he nevertheless clearly has a notion of religion as an idea that can be abstracted from any particular manifestation of it; for Gregory, different peoples have different “religions” (Christianity and Hellenism), and some folks have none. Once again, we see his Greek approximating the later meanings of the Latin religio.
Gregory affords a definition of “religion” as clear as that of later comparatists (although quite different from them). A “religion” is something that has priests, rites, rules, and sacrifices. It is absolutely clear, moreover, from Gregory’s discourse that, for this Christian, “the emergence of religion as a discrete category of human experience—religion’s disembedding,” in Seth Schwartz’s terms34—has taken place fully and finally, as he explicitly separates religion from ethnicity and language. In Julian, in contrast, on the one hand, we find “Hellenism” described as something very like a modern religion, but, on the other hand, his tight association of his “Hellenism” with all of Hellenic practice is more akin to ancient national cults than to latter-day “religion.” As Schwartz explicitly writes, “religion” is not a dependent variable of ethnos; indeed, almost the opposite is the case.35 A corollary of this is that language itself shifted its function as identity marker.
As Claudine Dauphin has argued, by the fifth century linguistic identity was tied to religious affiliation and identity, and not to geographic or genealogical identification.36 The fullest expression of this conceptual shift may be located in the heresiology of Epiphanius (fl. early fifth century), although his terminology is not entirely clear. For him, not only “Hellenism” and “Judaism” but also “Scythianism” and even “Barbarianism” are no longer the names of ethnic entities37 but of “heresies,” that is, “religions” other than orthodox Christianity.38 Since Epiphanius’s usage of the term “heresy” is quite idiosyncratic—the term is never applied to “Hellenism” elsewhere to the best of my knowledge—apparently what he means by “heresies” is often what other Greek writers of his time, such as Eusebius,39 call threskeiai (and even Epiphanius uses this term elsewhere too): “<Hellenism originated with Egyptians, Babylonians and Phrygians>, and it now confused <men’s> ways.”40 It is important to see that Epiphanius’s comment is a transformation of a verse from the Pauline literature, as he himself informs us.41 In Colossians 3:11 we find, “Here there cannot be Greek and Jew, circumcised and uncircumcised, barbarian, Scythian, slave, free man, but Christ is all, and in all.” This is a lovely index of the semantic shift. For pseudo-Paul, these designations are obviously the names not of “religious” formations but of various ethnic and cultural and even social groupings,42 whereas for Epiphanius they are the names of “heresies,” by which he means groups divided and constituted by “religious”—in the sense that I have defined that word above—differences fully disembedded from ethnicities: How, otherwise, could the “heresy” called “Hellenism” have originated with the Egyptians?43 Astonishingly, Epiphanius’s “Hellenismos” seems to have nothing to do with the Greeks; it is Epiphanius’s name for what other writers would call “paganism.” Epiphanius, not surprisingly, defines “the topic of the Jews’ religion [threskeian]” as “the subject of their beliefs.”44 For Epiphanius, as for Gregory, a major category (if not the only one) for dividing human beings into groups is “the subject of their beliefs,” hence the power/knowledge regime of what we call, following Latin Christian usage, “religion.”
Whatever the terminological issues involved, Epiphanius’s meaning seems quite clear from context. The system of identities had been completely transformed during the period extending from the first to the fifth centuries. The systemic change resulting in “religious” difference as a modality of identity that began, I would suggest, with the heresiological work of Christians such as Justin Martyr works itself out through the fourth century and is closely intertwined with the triumph of orthodoxy. Orthodoxy thus not only is a discourse for the production of difference within, but functions as a category to make and mark the border between Christianity and its proximate, invented, other “religions,” particularly a “Judaism” that it is too inventing.
ORTHODOXY AMONG THE RABBIS
It could be maintained, along with other scholars, that Christianity developed the notion of “orthodoxy” simply as a technology for defining what it was in a world in which such a thing, disembedded religion, had not existed before. I am using “orthodoxy” in the sense referred to by Rowan Williams when he wrote, “ ‘Orthodoxy’ is a way that a ‘religion,’ separated from the locativity of ethnic or geocultural self-definition as Christianity was, asks itself: ‘[H]ow, if at all, is one to identify the ‘centre’ of [our] religious tradition? At what point and why do we start speaking about ‘a’ religion?’ ”45
Given that, as Williams argues, the notion of orthodoxy is almost exclusively significant for the development of a transethnic “religion” as a mode of identification, I argued in Border Lines that the apparent development of a similarly defined “Judaism” at the time of the Mishna, namely, rabbinic Judaism, was a product of the encounter with Christian orthodoxy. This part of my argument in that book came, rightly so, under considerable fire.46 A much more broadly based historical contextualization of the development of rabbinic orthodoxy is necessary. The evidence for the development of a virtual orthodoxy as definitional for rabbinic Judaism, that is, the representation of Judaism as a nascent orthodoxy in the Mishna, stands up in my view and I will not rehearse it here, nor yet the exclusion of the minim (believers in a binitarian godhead) from that community of the orthodox. I would no longer, however, see this as a product of the impact of Christian developments (such as those of Justin and Irenaeus). On the other hand, the contrary thesis, defended by Alain Le Boulluec, that the notion of orthodoxy developed in Christian circles owing to the impact of the rabbinic developments, seems to me (still) equally untenable and based on outmoded assumptions about the historical fealty of rabbinic accounts of events long before their times. For one thing, as pointed out correctly by my critics, the rabbinic Judaism and earliest orthodox Christianity are way too close in time (and Christianity is still so insignificant institutionally) for either to have directly caused the development in the other. I would suggest now rather that we see both in such scholarly Christian writers as Justin and Irenaeus and in the equally scholarly producers of the Mishna the impact of the philosophical schools and their own developing notions of orthodoxy and authority.47
The correct generalization seems now to me that while for the nascent Church the use of such a model and the nascent notion of “heresy” that it offered was necessary for Christian self-definition for the reasons given by both Rowan Williams and Mason, since “Judaism” was supported by a vigorous and ongoing ethnic identity, nascent notions of heresy and orthodoxy were never crucial for rabbinic self-definition and ultimately fell into desuetude, as argued in my book.
I remain committed, and find nothing to contradict, my claim that the definition of “Judaism” as a religion served ongoing Christian discursive and polemical needs that were manifested in such documents as the Theodosian Code as well as some late-ancient Christian narratives of the conversion of Jews.48 That Christian identification of Judaism as a religion has had ongoing and complex effects on Jewish self-definition then, from late antiquity until modernity. In the “definitive” formulation of rabbinic Judaism in the Babylonian Talmud, the rabbis rejected this option, proposing instead a distinct ecclesiological principle: “An Israelite, even if he [sic] sins, remains an Israelite [one remains a part of a Jewish or Israelite people whether or not one adheres to the Torah, subscribes to its major precepts, or affiliates with the community].” Whatever its original meaning, this sentence was understood throughout classical rabbinic Judaism as indicating that one cannot cease to be a Jew even via apostasy.49 The historical layering of these two ideologies and even self-definitions by the rabbis themselves of what it is that constitutes an Israel and an Israelite provide for the creative ambivalence in the status of Judaism today. This thesis should not in any way, shape, or form be construed as a claim for greater tolerance of diversity among Jews than Christians.50
THE CHRISTIAN INVENTION OF RELIGION
Hegemonic Christian discourse also produced Judaism (and paganism, for example, that of Julian) as other religions precisely in order to cordon off Christianity, in a purification and crystallization of its essence as a bounded entity. Julian cleverly reverses this procedure and turns it against Christianity. In at least one reading of Julian’s “Against the Galileans,” the point of that work is to reinstate a binary opposition between Greek and Jew, Hellenism and Judaism, by inscribing Christianity as a hybrid. Eusebius’s claim that Christianity is a religion halfway between Judaism and Hellenism now constitutes an argument that Christianity is a monstrous hybrid, a mooncalf: “For if any man should wish to examine into the truth concerning you, he will find that your impiety is compounded of the rashness of the Jews and the indifference and vulgarity of the Gentiles. For from both sides you have drawn what is by no means their best but their inferior teaching, and so have made for yourselves a border of wickedness.”51 Julian further writes: “It is worth while … to compare what is said about the divine among the Hellenes and Hebrews; and finally to enquire of those who are neither Hellenes nor Jews, but belong to the sect of the Galileans.”52 Julian, as dedicated as any Christian orthodox writer to policing borderlines, bitterly reproaches the “Galileans” for contending that they are Israelites and argues that they are no such thing—they are neither Jews nor Greeks but impure hybrids.53 Here Julian sounds very much like Jerome when the latter declares that those who think they are both Jews and Christians are neither, or Epiphanius when he refers to the Ebionites as “nothing.” This would make Julian’s project structurally identical to the projects of the Christian heresiologists who, at about the same time, were rendering Christianity and Judaism in their “orthodox” forms the pure terms of a binary opposition, with the “Judaizing” Christians, the hybrids who must be excluded from the semiotic system, being “monsters.” I suggest, then, a deeper explanation of Julian’s insistence that you cannot mix Hellenism with Christianity. It is not only that Hellenism and Christianity are separate religions that, by definition, cannot be mixed with each other, but even more that Christianity is always already (if you will) an admixture, a syncretism. Julian wants to reinstate the binary of Jew and Greek. He provides, therefore, another instance of the discursive form that I am arguing for in the Christian texts of his time, a horror of supposed hybrids. To recapitulate, in Julian’s very formation of Hellenism (or should I say Hellenicity?)54 as a religious difference, he mirrors the efforts of the orthodox churchmen. This is another instantiation of the point made above by Limberis.55 While he was protecting the borders between Hellenism and Judaism by excluding Christianity as a hybrid, Julian, it seems, was, unbeknown to himself, smuggling some Christian ideas in his very attempt to outlaw Christianity.
This interpretation adds something to that of Jacobs, who writes that
the Christian reader could come to understand through the representation of the known and dominated figure of the Jew that Christians have power. The production of naturalized and totalized forms of Christian knowledge, significantly oriented around the paradigmatically known figure of the Palestinian Jew, delineated the first crude outlines of imperial Christian power.… The Jew emerges throughout the fourth century, the “Age of Constantine,” as a malleable “shadow-other” of the Christian-in-power; in the totality of the new Christian worldview, power from the Christian is articulated though imperial power over the Jew of the “frontier zone” of the holy land.56
I am suggesting that the heretic can also be read as a proximate other, producing a hierarchical space between the Christian and the Jew. This point is at least partially anticipated by Jacobs himself when he writes that “Jews exist as the paradigmatic ‘to-be-known’ in the overwhelming project of conceptualizing the ‘all in all’ of orthodoxy. This comes out most clearly in the [Epiphanian] accounts of ‘Jewish-Christian’ heresies.”57 One way of spinning this would be to see heresiology as central to the production of Judaism as the “pure other” of Christian orthodoxy, while the other way of interpreting it would be to see Judaism as essential to the production of orthodoxy over against heresy. My point is that both of these moments in an oscillating analysis are equally important and valid. Seen in this light, orthodoxy itself, orthodoxy as an idea, as a regime (as opposed to any particular orthodox position), is not only or even primarily a mode for the oppression and exclusion of people called “heretics”—particularly as so many of these latter seem to be fictive—but essential to the very formation of Christianity as religion, qua religion, as an entity and form of identification that is autonomous of particular ethnicities, histories, genealogy, and language. In this sense, perhaps, “Judaism” is never—until modern times—a religion, except perhaps as the “other” of Christianity.
It may then be suggested, in conclusion, that Cantwell Smith was correct in not discerning any named “religion” before the nineteenth century, with the proviso that this is only true of groupings other than Christianity.58 The Christian invention of Judaism in the third century and the confident talk of “religion” in both Latin and its hypothesized Greek equivalents in the fourth century by ecclesiastical writers would suggest that the notion itself, however, has much more ancient historical sources within Christian writing than are acknowledged by Cantwell Smith, Talal Asad, or J. Z. Smith.
NOTES
This paper was originally written and submitted in 2011. To be sure in 2014, I made certain revisions but in some ways it reflects my thinking of five years ago. I mention this simply because it will appear, insha’allah, at about the same time as work that goes in a somewhat different (albeit not totally incompatible) direction. The critical aspect about “Judaism” is now being developed into a book—Judaism: A Genealogy—that I am currently writing for Rutgers University Press for the Keywords in Jewish Studies series. The argument about the fifth century and Christianity will need the most rethinking. On the one hand, it is clear to me that something very important happened at that time; it is less clear to me now that it ought to be called the invention of religion. Much of my new thought is the product of the joint work undertaken in the last five years on Imagine No Religion with Carlin A. Barton (2016) for Fordham University Press. Jonathan Boyarin made some critical interventions as well, for which I thank him as always. It is clear to me that there remain unsolved, and as yet unsolvable (by me), terminological and conceptual problems, and I invite critical and suggestive response. Finally, I would like to thank Ruth Haber, who has been an indispensable research assistant for the last five years, as she moves on now to bigger and better things.
1. Jonathan Z. Smith, Imagining Religion: From Babylon to Jonestown (Chicago: University of Chicago Press, 1982), 11.
2. Talal Asad, Genealogies of Religion: Discipline and Reasons of Power in Christianity and Islam (Baltimore: Johns Hopkins University Press, 1993), 29. For discussion of these matters (and the placement of these two quotations side by side), see Elliot R. Wolfson, “Structure, Innovation, and Diremptive Temporality: The Use of Models to Study Continuity and Discontinuity in Kabbalistic Tradition,” Journal for the Study of Religions and Ideologies 6, no. 18 (2007): 148.
3. Wilfred Cantwell Smith, The Meaning and End of Religion (London: SPCK, 1978), 61.
4. Cf. Steve Mason, “Jews, Judaeans, Judaizing, Judaism: Problems of Categorization in Ancient History,” Journal for the Study of Judaism 38, no. 4–5 (2007): 482.
5. Ibid., 465–68.
6. John J. Collins, “Cult and Culture: The Limits of Hellenization in Judea,” in Hellenism in the Land of Israel, ed. John J. Collins and Gregory Sterling (Notre Dame: University of Notre Dame, 2000), 39. The medieval Yahadut does not mean “Judaism” but the state of being Jewish, as I shall show, Deo volonte, in the book.
7. The reason that the term “Jews” is not scare-quoted in this project at any rate is that it certainly is an ancient category defined then by a group in which there was mutual recognition of identity. Thus one could be a “bad” Jew in the eyes of another without ceasing to be recognized as a Jew.
8. Note that this does not imply that particulars do not or did not exist. People obviously had intercourse before there was sexuality, and there is no reason to believe that people who did not have the same names for emotions did not feel them substantially as we do.
9. Mason, “Jews.”
10. I write this paragraph in the wake of Seth Schwartz, as mentioned earlier.
11. Mason, “Jews.”
12. Ibid., 473.
13. Ibid., 476.
14. Smith, The Meaning and End of Religion, 488.
15. Mason, “Jews,” 489.
16. For the popularity of De natura deorum among Christians such as Tertullian, Minucius Felix, Lactantius, Arnobius, and Augustine, see Arthur Stanley Pease, introduction to M. Tulli Ciceronis de natura deorum, by Cicero (New York: Harvard University Press, 1979), 1:52–57; H. Bouillard, “La formation du concept de religion en Occident,” in Humanisme et Foi Chretienne: Melange scientifiques de Centenaire de l’Institut Catholique de Paris, ed. C. Kannengiesser and Y. Marchasson (Paris: Editions Beauchsene, 1976), 453; Andrew Dyck, introduction to De natura deorum: Book I, by Cicero (Cambridge: Cambridge University Press, 2003), 14–15.
17. Personal communication, May 25, 2011.
18. Asad, Genealogies, 45.
19. David Chidester, Savage Systems: Colonialism and Comparative Religion in Southern Africa (Charlottesville: University Press of Virginia, 1996).
20. This point contradicts one observation frequently made of the supposed Enlightenment notion of “religion,” to wit, that it assumes religion as a universal.
21. Hal A. Drake, “Lambs Into Lions: Explaining Early Christian Intolerance,” Past and Present 153 (1996): 25. Drake’s theory is germane to the hypothesis of this article. Limberis argues that for second-generation Christians this process was reversed in Vasiliki Limberis, “ ‘Religion’ as the Cipher for Identity: The Cases of Emperor Julian, Libanius, and Gregory Nazianzus,” Harvard Theological Review 93, no. 4 (2000): 377. I am not entirely persuaded by her argument on this point but do not wish to entirely disallow it, either. One way of thinking about it would be to see who is left out of “us.” In both the earlier rabbinic and orthodox Christian formations, exemplified by Nazianzen below, there are those tied to us by tradition, kinship, and land who are, nevertheless, not us; they are heretics. See also Rosemary Radford Ruether, “Judaism and Christianity: Two Fourth-Century Religions,” Sciences Religieuses/Studies in Religion 2 (1972): 1–10; and Jacob Neusner, Judaism and Christianity in the Age of Constantine: History, Messiah, Israel, and the Initial Confrontation (Chicago: University of Chicago Press, 1987), who take related positions.
22. Andrew S. Jacobs, “The Imperial Construction of the Jew in the Early Christian Holy Land” (PhD diss., Duke University, 2001), 28–29. See also Jonathan Boyarin, The Unconverted Self: Jews, Indians, and the Identity of Christian Europe (Chicago: University of Chicago Press, 2009), 59.
23. Seth Schwartz, Imperialism and Jewish Society from 200 B.C.E. to 640 C.E. (Princeton: Princeton University Press, 2001), 179.
24. Susanna Elm, “Orthodoxy and the True Philosophical Life: Julian and Gregory of Nazianzus,” Studia Patristica 37 (2001): 79–80. See also Limberis, “Cipher,” 383.
25. The Eusebian material is complex and will require a separate treatment of its own in the monograph for which this is an exploratory drill.
26. Limberis, “Cipher,” 378, 382, and throughout. This is true, as well, for those fourth-century Roman aristocrats who designed themselves “pagans.”
27. Ibid., 386.
28. Ibid., 399. I accept Limberis’s assent to Asad’s critique of Geertz, but nevertheless see much more continuity and a shift toward something that could be called “religion” in the modern sense taking place precisely in these fourth-century echoes of Christianity.
29. Oration 4.5 and 96–109, quoted in Elm, “Orthodoxy and the True Philosophical Life.” See also Limberis, “Cipher,” 395.
30. Cf., for example, Asad, Genealogies, 40–41.
31. According to Liddell-Scott, less than one hundred total occurrences, of which about 90 percent are in one writer, Josephus. A full study of the semantics of this term in Josephus is being carried out for the larger project.
32. Asad, Genealogies, 45.
33. In future work I will be treating the semantics of Hebrew dat and yahadut and showing, Deo volente, that the first does not mean “religion,” nor the second “Judaism.”
34. Schwartz, Jewish Society, 179.
35. This point is not contradicted in any way by Denise Kimber Buell, “Race and Universalism in Early Christianity,” Journal of Early Christian Studies 10, no. 4 (Winter 2002): 429–68. Buell’s compelling analysis of second- and third-century texts indicates early Christianity’s struggle to find a mode of identity, with notions of Christianness as a new ethnos/genos being very prevalent indeed. However, Buell herself marks a shift that takes place in the fourth century: “Beginning in the fourth century, ethnic reasoning serves to naturalize the equation of Christianness with gentileness, or Romanness, in part through the oppositional construction of non-Jewish non-Christians as ‘pagans’ ” (Buell, “Race,” 465). I would argue, however, that such a classification marks the undoing of an “ethno/racial” definition of Christianness, insofar as in general throughout the fourth century “pagans” were understood to be just as Roman as Christians. “Pagan” surely did not constitute an ethnic or racial designation but a religious one. Even in the earlier writings considered by Buell, where Christianity is defined as an ethnos or a genos, these terms are the dependent variable of “faith.” This is decidedly not the case for Jews much before the Christian era or for Judaism since the early middle ages. Buell argues elegantly that Christian universalism should not be seen in opposition to or against the background of a putative Jewish particularism: “Seeing that early Christians defined themselves in and through race requires us to dismantle an oppositional definition of Christianness and Jewishness on the basis of race or ethnicity. Doing so may also contribute to resisting periodizations that mark an early and decisive split between Christianities and Judaisms. Not only do many early Christians define themselves as a people, even competing for the same name—Israel—but early Christians adapt and appropriate existing forms of Jewish universalism in formulating their own universalizing strategies in the Roman period.… Since ethnic reasoning also resonates with non-Jewish cultural practices of self-definition, it offers an analytic point of entry that treats both Jewish and non-Jewish frames of reference as integrally part of Christian self-definition, not as its ‘background’ ” (Buell, “Race,” 467). At the same time, notwithstanding Buell’s reference to Isaiah as “emphasizing attachment to Yahweh as defining membership in Israel,” I would suggest that the notion of “orthodoxy” as defining membership in the Christian community and the feints in that direction in rabbinic literature that define orthodoxy as the criterion for membership in Israel represent a “new thing.” That new thing would ultimately be called “religion.”
36. Claudine Dauphin, La Palestine byzantine: Peuplement et populations (Oxford: Archaeopress, 1998), 133–55.
37. Which is not, of course, to claim that the notion of ethnic identity is a stable and fixed one, either. See Jonathan M. Hall, Ethnic Identity in Greek Antiquity (Cambridge: Cambridge University Press, 1997).
38. Epiphanius of Salamis, The Panarion of Epiphanius of Salamis, Book I, Sections 1–46, trans. Frank Williams (Leiden: Brill, 1987), 16–50. Cf., however, Eusebius, Demonstratio evangelica 1.2.1; Eusebius, The Proof of the Gospel, ed. and trans. W. J. Ferrar (London: SPCK, 1920), 9.
39. I will treat Eusebius at length, Deo volente, in the fuller work for which this essay is a preliminary study.
40. Epiphanius, Panarion, 17–18; David Frankfurter, Religion in Roman Egypt: Assimilation and Resistance (Princeton: Princeton University Press, 1998), 79. In another part of the Christian world, Frankfurter points out, for the fifth-century Coptic abbot Shenoute “Hēllēne did not carry the sense of ethnically ‘Greek’ and therefore different from ‘Egyptian,’ but simply ‘pagan’—‘not Christian.’ ” In Syriac, as well, “Aramean” is no longer the designation of an ethnic or linguistic group but simply means what others call “pagan.”
41. Epiphanius, Panarion, 9.
42. For a highly salient and crystal clear delineation of these terms, “ethnic” and “cultural,” see Jonathan M. Hall, Hellenicity: Between Ethnicity and Culture (Chicago: University of Chicago Press, 2002), 9–19.
43. See discussion in Andrew S. Jacobs, Remains of the Jews: The Holy Land and Christian Empire in Late Antiquity (Stanford: Stanford University Press, 2004), 45–46; Aline Pourkier, L’hérésiologie chez Épiphane de Salamine (Paris: Editions Beauchesne, 1992), 85–87; and Frank Young, “Did Epiphanius Know What He Meant by ‘Heresy?,’ ” Studia Patristica 17, no. 1 (1982): 199–205. As has been noted by previous scholars, for Epiphanius “heresy” is a much more capacious and even baggy-monster category than for most writers.
44. Epiphanius, Panarion, 24. “But to keep from getting side-tracked, bypassing the topic of the Jews’ religion, and failing to touch on the subject of their beliefs, I shall give a few examples of them.”
45. Rowan Williams, “Does It Make Sense to Speak of Pre-Nicene Orthodoxy?,” in The Making of Orthodoxy: Essays in Honour of Henry Chadwick, ed. Rowan Williams (Cambridge: Cambridge University Press, 1989), 3.
46. Daniel Boyarin, Border Lines: The Partition of Judaeo-Christianity (Philadelphia: University of Pennsylvania Press, 2004). The following paragraphs represent a revision of my thesis in Border Lines, a revision occasioned by some very sharp and pointed critique by the respondents to the book in a symposium at the AAR, later published as a special forum in the journal Henoch: Virginia Burrus et al., “Boyarin’s Work: A Critical Assessment,” Henoch 28 (2006): 7–30. As a result of this criticism, as well as some other, I came to realize two great shortcomings in the argument of that book: (1) I had not in any way properly assessed or even really thought about the non-Christian, non-Jewish environment of the invention of “religion”; (2) insofar as I had done so, my statements about it were merely contradictory to my general thesis. In these few paragraphs here, I hope to sketch out a more satisfactory outline for a synthesis.
47. Élie Bikerman, “La chaîne de la tradition pharisienne,” Revue biblique 59, no. 1 (1952): 44–54; John Glucker, Antiochus and the Late Academy (Göttingen: Vandenhoeck und Ruprecht, 1978); Albert I. Baumgarten, “The Pharisaic Paradosis,” Harvard Theological Review 80 (1987): 63–77; all remain crucial for this point.
48. Daniel Boyarin, “The Christian Invention of Judaism: The Theodosian Empire and the Rabbinic Refusal of Religion,” Representations 85 (2004): 21–57.
49. For instance, a Jew who “converts” to another religion does not have to convert back but only repent his or her sins in order to be accepted in the community again.
50. The claim that orthodoxy (as the necessary concomitant of “religion”) was never meant, and should not be read, as a disparagement of Christianity. (As Joel Marcus most starkly represented it in an oral presentation: the Christians developed a disease, the Jews caught it for a while, then shook it off.) “Orthodoxy” is to be taken as neutral a term as “church” or “bishop” or “Jewish People,” no more, no less and not the name for an intrinsically evil institution.
51. Julian, “Against the Galileans,” in The Works of the Emperor Julian, trans. W. C. F. Wright (London: Heinemann Macmillan, 1913), 389.
52. Ibid., 319–21.
53. Ibid., 393–95. Fascinatingly, this perspective gives us another way of understanding Julian’s intention to allow the Temple in Jerusalem to be rebuilt. A large part of his polemic consists, as we have seen, of charges that Christians are nothing, since they have abandoned Hellenism but not become Jews, given that they do not follow the Torah. He imagines a Christian answering him that the Jews, too, do not sacrifice as they are enjoined (ibid., 405–7). What better way to refute this Christian counterclaim and demonstrate that the only reason that Jews do not sacrifice is that they have no Temple than to help them rebuild their Temple and reinstitute the sacrifices?
54. Hall, Hellenicity, xix.
55. Julian, “Against the Galileans,” 315. Wright, the translator, points out that Julian has Christlike figures in his own theology.
56. Jacobs, Remains, 53–54.
57. Jacobs, “Imperial Construction,” 57.
58. Another term that needs careful semantic historicization in this context.