Abstract
This essay provides a summary and critical assessment of scholarship on sacrifice in the ancient Mediterranean over the last two decades. It focuses on Greek, Roman, Judean and Christian evidence from approximately the eighth century
Keywords
Introduction
I start with an appropriate note of trepidation at the prospect of encapsulating scholarship on the sprawling field of sacrifice over the last few decades. Sacrifice has, for over a century, held an almost mesmeric power over the fields of Classics and Religious Studies. It has served as a locus for theorizing, theologizing and polemicizing from the emergence of the fields to the present day. The purpose of this essay is, humbly, to lay out some of the recent arguments as well as some trending future prospects. I hope this essay will aid in orienting readers to the topic and pave the way for continued advancement in our thinking about this important practice and the complex discourse surrounding it.
I will not fully address the voluminous and very important work on this topic prior to the rather arbitrary cut-off point of the mid 1990s. Interested readers can find excellent summaries of this earlier work in Bell’s Ritual: Perspectives and Dimensions (1997), the early chapters of which are nearly an essay on early interpretations of sacrifice. The entry on sacrifice in Brill’s New Pauly (Cancik et al. 2002–10) also provides a generally excellent introduction, particularly the discussion of theory by Becker. Carter’s edited volume, Understanding Religious Sacrifice: A Reader (2003), is also highly valuable as it brings together succinct extracts from many of the most important theorists from the 1800s to the 1990s. Although Carter’s excerpts are necessarily short, they provide a great range of influential approaches to the topic. A quick foray into just this volume is evidence of how enormously varied approaches to sacrifice have been. Further condensation of earlier scholarship can found in Ekroth (2007a); Faraone and Naiden (2012: 1-10); Graf (2002); Knust and Várhelyi (2011: 3-31); Mack (1987); and Reed (2013). Scholarly interest on this area is now quite old and recent years show little evidence of it slackening, though it is certainly changing.
In the last two decades, there have been a significant number of academic conferences devoted to sacrifice, several of which have produced volumes of essays: Baumgarten (2002); Bremmer (2007); Faraone and Naiden (2012); Georgoudi et al. (2005); Knust and Várhelyi (2011); and Schweidler (2009). There is also a Society of Biblical Literature (SBL) section devoted to the topic ‘Sacrifice, Cult, and Atonement’, which produced a volume edited by Eberhart (2011). Sacrifice also appears as a frequent theme of other SBL groups, such as ‘Greco-Roman Religion’, ‘Meals in the Greco-Roman World’ and ‘Redescribing Christian Origins’.
The variety—and at times disparity—of the current work defies any attempt to summarize or ascertain a single direction. If any overarching theme may be said to be emerging it is that scholars are less sure of how coherent or important sacrifice, as a category and concept, may be in the first place. As Rives observes, ‘the monolithic place of animal sacrifice in accounts of ancient religion is now starting to be called into question’ (2011: 187). A 2008 conference at the University of Chicago broached this issue directly with the provocative title ‘The Centrality of Animal Sacrifice in Ancient Religion: Ancient Reality or Modern Construct’? (essays are collected in Faraone and Naiden 2012). Other scholars have raised the concern that the term sacrifice itself is problematically vague and theological. They call for more critical reflexivity in our scholarly use of this term (Frankfurter 2011; McClymond 2008, 2011a; McGowan 2014b; Modéus 2005; Rives 2011).
In place of monolithic views of sacrifice as a stand-alone cultural phenomenon, or a universally human je ne sais quoi, many current studies work hard to integrate analysis of sacrifice into specific historical, social and ritual networks—what Frankfurter usefully terms ‘life contexts’ (1998: 33). Recent work has tended to question the special role specifically animal sacrifice held in earlier scholarship. For a range of reasons, earlier scholarship tended to separate ‘blood sacrifice’ from other types of offerings (grain, fruits, vegetables, wine, oil, etc.). One reason for this privileging of animal offerings is the special symbolic role blood came to serve in ancient Judean religion and, more importantly, early Christianity (Biale 2007; Gilders 2004; Stowers 1998). Several recent scholars have called into question the separation and privileging of animal offerings (McClymond 2008; Bruit Zaidman 2005). Thus, the two main current trends present in many of the works discussed below are a reintegration of animal sacrifices with other forms of offerings, and a recontexualization of all sacrifices, animal and non-animal, within broader historical, economic and social contexts. These advances in scholarly discussion of sacrifice are part of similar theoretical trends in the study of ancient Mediterranean religion generally. For a summary of recent trends in the study of Greek and Roman religion, see Naiden (2013a) and Rives (2010).
The major challenge of an essay of this type is to find a way to conceptualize and organize the disparate strands of scholarly interest on the subject. Instead of the more traditional breakdown of Graeco-Roman, Judean and Christian, I have tried to integrate different scholarly discussions by focusing on broad topics of analysis and scholarly dispute (though there are two short sections on specific issues in Judean religion and Christianity below). I begin with a discussion of theory, since this has played a major role, for better and worse, in the study of sacrifice. The majority of the essay is then devoted to analyzing specific areas of scholarly analysis and contestation. I end with a brief discussion of the broader implications of current scholarship.
Ritualization, Theory and Sacrifice
Any discussion of sacrifice must start with a discussion of ritual and ritual theory, which has changed radically over the last century with particular insights coming from scholars such as Bell (1992, 1997) and new areas of study such as cognitive theory (Whitehouse 2004; Lawson and McCauley 1990; McCauley and Lawson 2002) and sociology (Bourdieu 1977, 1984; Schatzki 2002). The most important currents in this field show a clear move away from seeing ritual as a form of enacted myth or as actions that encode universal essences or meanings.
It was a principle of many early approaches that rituals enacted societal values and myths, thus symbolically communicating the deep truths, or reflecting the deep structures, of a culture. This view is still prominent in some circles; for example the International Encyclopedia of the Social Sciences defines ritual as ‘a patterned, repetitive, and symbolic enactment of cultural belief or value’ (Davis-Floyd 2008: 259). More up-to-date work argues that rituals are, generally, not a medium for symbolic communication of complex ideas. (The common notion of symbolism itself has been critiqued and transformed by Sperber 1975; these points are raised in Gilders 2011: 99-101; see also Struck 2004. For a critique of this position, see Klawans 2006 and 2011, discussed below.)
By their very nature, ritualized behaviors are underdetermined; they exponentially multiply potential interpretations, thus confounding communication. As Boyer points out, ‘ritualization reduces rather than increases the amount of information potentially conveyed’ (2010: 64). There is thus a great divide between discursive practices such as speech and writing, and ritualized practices, which create a much greater degree of ambiguity.
It may be useful here to make a distinction in the use of the term ‘communication’, specifically between the signaling of simple ideas and the complex discursive concepts that the word ‘communication’ often implies. For example, witnessing someone sacrificing might signal assumptions about the existence of gods or other ‘non-obvious beings’ (on this term see Stowers 2011b; Arnal and McCutcheon 2013: 151-52). But these assumptions are deeply ambiguous (what exactly does the sacrificer believe about the meaning and purpose of his or her action?), and the act is also open to deception. Clarifying these ambiguities would require discursive practices such as speech or writing. Critically, the motivation to effect such clarification is not universal.
Recent work has built on earlier ethnographic studies to clarify what ritualization can and cannot communicate. This work is only starting to influence the study of sacrifice, but it shows clearly that many scholarly interpretations of sacrifice have forced the ritual to bear far more discursive meaning than is possible (Staal 1979 remains the classic critique; for more recent work see Whitehouse 2004). The ‘meaning’, as this word is commonly used, is not present in the ritual; rather ‘meaning’ is always open to interpretation, dispute and competition through discursive practices, which are secondary to the ritual itself (Jay 1992; Stowers 2011b; Ullucci 2012). Attempts to get at meaning via linguistic analysis of sacrificial terminology are equally problematic, as they locate discursive meaning in a linguistic tradition that is often just as ambiguous (see, for example, Carter 2003: 2-3; Eberhart 2011; Meshel 2013, 2014). A better theoretical framework for both sacrifice and ritual in general is clearly needed.
Criticism of the Terms Ritual and Sacrifice (and Religion)
Definitions of ritual are often far too vague to be useful and, worse, are liable to mobilize a range of a priori judgments about what counts as ritual, as well as what counts as religion (Arnal and McCutcheon 2013; Nongbri 2013). Was the act of 9/11 a ritual of self-sacrifice or mass-murder (McClymond 2008: 160-64)? Was the death of Saddam Hussein an execution or a heroic act of ritual self-sacrifice (McClymond 2009)? Current ritual theory attempts to address this problem by avoiding the vague term ‘ritual’ as a universal category and speaking instead of ‘ritualized behavior’ (see Boyer 2010: 63-69 and Grimes 1995, 2014). This is not just a semantic or pedantic shift. Ritualized behavior can be seen in many situations that would not normally be called ritual, such as children’s bedtime behaviors and the actions of those suffering from obsessive-compulsive disorder (OCD) (Boyer 2010: 63-69). On the other hand, some things commonly labeled ritual do not display any ritualized behavior. Ritualized behavior is not the same as routinized behavior; in fact, they are opposites. Whereas routinized behavior is mindless and automatic (like brushing one’s teeth or driving to work), ritualized behavior is intentional and strategic (like a handshake at a job interview or a communal moment of silence) (Boyer 2010: 68). There is great clarity to be gained from seeing sacrifice as a form of ritualized behavior rather than as an example of some vaguely defined category ‘ritual’. The main benefit is that this shift immediately makes a division between the action itself and interpretations of the action. This division plays a key role in many of the works discussed below.
A major revelation from current ritual theory is the placing of ritualized behavior into the broader context of human behavior generally, as well as into the context of evolutionary biology. No longer are ritualized behaviors confined to religion or even to humans. The work of E.O. Wilson, the father of sociobiology, has shown the presence of ritualized behavior not merely among advanced primates, but also among eusocial insects such as ants and bees (Wilson 1980, 2004; Hölldobler and Wilson 2009; for a comparison of animal and human ritualization, see Driver 1998: 12-31). Psychologists have shown that spontaneous ritualized behavior is common in children cross-culturally. Such behavior usually appears around the age of two, peaks at age five, and declines by age seven. In some individuals this behavior does not subside, leading to the pathology known as OCD. New work in this area, by cognitive scientists, sociobiologists and psychologists, promises to elucidate some of the recurring elements of ritualized behavior, such as preoccupation with order, routine, repetition and cleanliness (Boyer 2010: 63-69). This is macro-level analysis, which may not yield direct revelations about specific rituals like sacrifice, but analysis of sacrifice must position itself in relation to these findings. Ritualization is a normal and inherent human behavior; sacrifice is just one instance of this larger phenomenon.
Scholars have accounted for the pervasiveness of ritualized behavior as an offshoot of cognitive faculties, specifically precaution systems (Boyer 2010: 66-67; Atran 2002: 149-74). By appealing to these evolutionarily shaped cognitive faculties, this new strand of ritual theory can account for the pervasiveness of ritualized behavior without appealing to old categories that have proven so problematic: meaning, essence, and to a lesser extent community. It was once common to assume that people found rituals appealing because the meaning of the ritual persuaded them, or because the ritual enacted some deep truth for them. This focus on meaning has finally been laid to rest by a whole range of ethnographic work which has conclusively shown that ritual participants often have little conception of the ‘official’ meaning of a ritual, if one even exists, and possibly no conception of a meaning at all (Whitehouse 2004: 95-96). Their ritualized behavior is pre-reflective or, better, a-reflective. (On this issue, however, see the contrary position of Klawans 2006 and 2011, also discussed below.)
Models that understand ritualized behavior from a cognitive perspective can account for why rituals are compelling without claiming that the rituals encode some deep meaning that is important to—and shared by, all the doers of a ritual (Boyer 2010; Modéus 2005; Whitehouse 2004). Focusing on creative and strategic ritualized behavior moves us usefully away from predetermined scholarly schema and towards the creative, contradictory and contested world of ritual actors themselves, as Latour urges (2005).
The question of a definition or definitions of sacrifice also needs to be reframed in light of new criticism about the validity of our categories (Nongbri 2013; Latour 2005). Making normative claims about sacrifice is the activity of those engaged in sacrificial discourse; it is not the job of the scholar to say what counts as sacrifice in a normative sense, but rather to report what counts as sacrifice for people being studied. The meaning of the sacrifice (is it gift or substitution or atonement?), the things offered, what they might symbolize, how they might be transformed, and how the gods might manifest a response, are all the purview of the persons involved. Their meanings and interpretations—however confusing, contradictory, incomplete, inarticulate or incoherent—form our object of study, a point well made by Sheehan (2009). Defining, organizing or limiting these things a priori would put us in the position of some grand religious authority passing down dogma, rather than the much more humble position of a scholar (Latour 2005).
This does not mean, of course, that we surrender our right as analysts to delimit our object of study. A range of concerns, including personal interest, expertise, current events, and the practicalities of the academic tenure and promotion process, effects this delimiting (Latour 2005). We may, for various reasons, focus only on animals, or vegetables, or civic sacrifice. The litmus test for such scholarly creations is how useful they are in helping us redescribe the phenomenon in illuminating ways. They are only dangerous when they are presented not as scholarly creations but as fundamental, universal or natural categories, as many earlier discussions of sacrifice presumed (for example, the assumption that blood sacrifice was categorically different from other types of offerings). The issue of definition is broached directly by a number of recent works.
Watts (2011) addresses this issue in an essay on the rhetoric of sacrifice. He argues that the meaning of the term sacrifice frequently comes from narrative analogies with stories of sacrifice rather than comparison with actual ritual practices. In other words, ‘sacrifice’ is an evaluative not a descriptive term. This is a problem when the term is used in academic theorizing where sacrifice is assumed to refer to some agreed-upon set of practices—a phenomenon particularly apparent in the way ideas about human killing have been attached to sacrifice (discussed more below). This creates confusion because different scholars are theorizing different things. When disparate scholarly work is grouped together under a single category ‘sacrifice’ that is assumed to be universal and self-evident, confusion and a host of other problems ensue. This tendency is apparent in many of the most influential theories of sacrifice. (For further discussion along these lines see Eberhart 2006.)
McClymond responds to this problem by suggesting a polythetic definition involving a number of recombinable ‘building blocks’, none of which are essential. (For a critique of this approach to definition, however, see Boyer 2010: 63.) She argues that the regularly occurring practices of sacrifice are: selection, association, identification, killing, heating, apportionment and consumption (2008: 27-34). With this definition, McClymond highlights her argument that killing is not an essential or necessary component of sacrifice, and certainly not the central action as has often been assumed. Frankfurter builds on this work in his own analysis of Egyptian sacrifice (2011).
Sheehan (2009) argues that the label ‘sacrifice’ is not a scholarly category at all, but an ideological argument. Calling something a sacrifice is an attempt to argue for some transcendent value for that thing. Thus, he argues what we need is not a theory of sacrifice, but a history of sacrificial claims. A theory of sacrifice would attempt to unify all so-called sacrifice under some theoretical umbrella. A history, on the other hand, would disintegrate sacrifice into a series of competitive and discursive claims. He calls for a ‘dispersion of the sacrificial claim into specific cultural, political, ethical, and religious moments’ (Sheehan 2009: 99). The reflective and reflexive work of these scholars and others (see also McGowan 2014b; Modéus 2005: 28-35; J.Z. Smith 1987) has pushed the field away from the kind of grand overarching theories of sacrifice that predominated in previous scholarship.
A Move Away from Grand Theories
Sacrifice is an area of scholarship where grand theories have historically abounded. The data, perhaps, breed this tendency. The very fact that sacrificial rituals appear in so many diverse historical and cultural settings, throughout time and across the globe, might suggest that there is something fundamental about this type of ritualized behavior for humans. Articulating what that fundamental something is without essentializing and reducing either the data or human nature itself has proved much more challenging (for a particularly successful attempt see Atran 2002: 114-46). As a result, sacrifice has become a Rorschach test in which theorists project their macro assumptions about human nature, religion and society.
Almost all of the classical theorists of religion (Tylor, Frazer, Durkheim, Freud, Marx, Eliade, Turner et al.) discussed sacrifice as a way to demonstrate their theories. For each of them, the ritual illustrated the fundamental ways in which religion supposedly worked. Carter’s (2003) collection illustrates this very well (see also Bell 1997: 1-22). The fact that so many theorists chose sacrifice as their test case shows just how diverse the concept ‘sacrifice’ is. A less complex concept could never have been rearranged and transposed into ‘proof’ for so many divergent and contradictory theories of religion.
Historically, these theories have provided distorting lenses through which all the data on sacrifice were filtered. The legacy of these lenses is still very much with us despite how frequently they have been demonstrated to be problematic. A good example is Girard’s theory that sacrifice is about redirecting innate human violence. Theories like Girard’s are compelling on their surface because they promise to explain a huge swath of data with a single stroke, but they do so only by doing violence to the evidence itself. In Girard’s case, his focus of killing and violence as the essence of sacrifice simply does not fit the ancient evidence (for a critique of Girard, see Klawans 2006; McClymond 2011a, 2011b; J.Z. Smith 1987). Despite this criticism, Girard’s theory still remains highly influential (see for example the volume of Astell and Goodhart 2011).
As McClymond (2011a) has argued, the data for sacrifice are far more complex and varied than the grand theories allow. Grand theories are also far less objective than they seem. This issue has been addressed by Lincoln and Graf in two essays that focus on the enduring influence of four key theories, those of: Hubert and Mauss (1964); Burkert (1972); Girard (1972); and Detienne and Vernant (1979). I will discuss these two essays at length since moving beyond the ‘grand theories’ is, I believe, a major current trend in the study of sacrifice.
Lincoln and Graf on Detienne, Vernant, Burkert and Girard
Lincoln (2012) points out that Burkert’s Homo Necans (1972) and Detienne and Vernant’s La cuisine du sacrifice (1979) have dominated scholarly discussions of sacrifice over the last thirty years. He shows that each of these works were deeply affected by previous theorists who were themselves influenced by the historical, political and social circumstances of their times. These grand theories are based on flawed lenses—lenses ground askew by the particular historical contexts of their creators. For example, Hubert and Mauss were affected by the ‘French crisis of the 1890s’ as Strenski previously noted (2002). This period bred an influential discourse of self-sacrifice among the right wing in France, built on heroes such as Jesus, Louis XVI, Joan of Arc and Captain Dreyfus. Hubert and Mauss, students of Durkheim, responded negatively to this ideology. They saw sacrifice as a communal act that brought contact with the sacred (understood, following Durkheim, as the community) and opposed the profane (imagined as all non-communal self-interests). This approach decentralized the sacrificial victim and deflated notions of heroic self-sacrifice. Lincoln concludes that their theory reflects ‘the Durkheimians’ concern to generate theories consistent with an ethic of inclusivity and tolerance, in pointed opposition to the views advanced by anti-Semites, anti-Dreyfusards, reactionary Catholics, and enemies of the Republic’ (2012: 18). Hubert and Mauss’s theory stands behind the work of Detienne and Vernant (1979), which argues that sacrifice is essentially a practice of community indexing and reaffirmation.
Lincoln (2012) also contextualizes the work of Karl Mueli (1946) who influenced Burkert (1972). Mueli’s work focused on the emotional response to death and killing. His essay ‘Griechische Opferbräuche’ (1946) analyzed ways in which Greek sacrificers and Central Asian hunters purportedly attempted to mitigate the emotional guilt of killing animals. Lincoln argues that this focus is not hard to understand given Meuli’s historical situation. He was a Swiss-German writing just after the end of World War II, when the full scale and horror of Nazi atrocities were being revealed. Meuli’s hunters and sacrifices, who perform a ‘comedy of innocence’ (Unschuldskömodie) but still know and feel their inescapable, unquenchable guilt, appear quite conspicuous in light of Meuli’s own historical situation. Beyond this, his theory is also entwined in a response to the racial claims of the Third Reich (Lincoln 2012).
By showing ways in which grand theories often reflect the historical situatedness of their creators, Lincoln illustrates keenly the danger addressed above, that the concept of sacrifice has become so broad and multifaceted that it invites a Rorschach-effect among theorists—the data could be made to fit almost any theory—thus casting doubt upon the usefulness of the category as it is normally used.
Graf (2012) continues Lincoln’s analysis of major theories in the same volume from the Chicago conference, focusing on Girard (1972); Burkert (1972); and Detienne and Vernant (1979). He points out that while Girard and Burkert differ greatly in their analysis of the origins of sacrifice, both of their theories converge on a basically negative anthropology. Girard’s theory is based on the idea that sacrifice derives from murder and the need to redirect the inherent violence of human communities; it is easily traced to the influence of Freud. Burkert’s focus on inherent guilt can be traced to the influence of the biologist Konrad Lorenz (in addition to Meuli, as discussed by Lincoln). This negative anthropology, Graf argues, can be seen as a reaction to both men’s personal experience of World War II, influenced as well by the philosophies of Schopenhauer and Nietzsche.
Vernant and Detienne’s less pessimistic and more matter-of-fact approach to sacrificial killing is also perhaps traceable to their experience of the war, but more so to the influence of French intellectual trends of the 1960s. Graf points out that while each of these theories remains influential in various places, they have been soundly critiqued and rejected, not least by J.Z. Smith, who appeared as a skeptical final voice in a volume edited by Hamerton-Kelly (1987). In this volume, entitled Violent Origins: Walter Burkert, René Girard, and Jonathan Z. Smith on Ritual Killing and Cultural Formation, Smith’s piece rejects the quest for origins inherent in Girard and Burkert’s work, as well as their focus on guilt and violence. Graf reinforces Smith’s rejection of origins and emphasis on killing; however, he also questions Vernant’s ‘matter-of-factness’, suggesting that some ancient writers do show ‘squeamishness’ over the death of the animal.
Overall, Graf shows interest in Smith’s ‘highly tentative’ suggestion in Violent Origins that sacrifice is a ritual response to domestication of food animals, although he admits this observation has not led to fruitful theorization. Graf (2012: 51) calls for more theoretical work that is cross-disciplinary and cross-cultural, pointing to a significant expansion in our archeological evidence from the Neolithic period unavailable to Burkert or Meuli.
Alternative views: continuance of large theories
The move away from grand theories is not universal, of course. Numerous recent works continue older theories, mostly Burkert or Girard, or offer other unifying (critics would say essentializing) theories of sacrifice (see, for example, Hedley 2009 and 2011; Lang 2002). A more complex example of the continuance of unifying theories is Gruenwald (2003). In a book that self-identifies as the first of its kind, Gruenwald builds on work in anthropology to offer a new theorization of ritual in general and animal sacrifice in ancient Judean religion specifically. He argues that we must first separate discursive meanings and interpretations from ritual practices themselves. These meanings and interpretations are secondary to the practice and are not somehow encoded in the practice. In making this first move, Gruenwald is very much in agreement with Staal, Jay, Whitehouse and several of the authors discussed below who critique grand theories of sacrifice. However, he ultimately differs from them radically in his second theoretical move.
Gruenwald argues that rituals are the spontaneous and direct expressions of the human mind, unmediated by formalized theology, symbols or ideas. The human mind expresses itself most directly, he claims, in structured actions, which we call rituals. The mind expresses these actions to ‘define and maintain’ existence by creating transformations that maintain the existence of the individual and group (2003: 24-26). Rituals are, he claims, ‘purposely structured, transformative actions’, bearing their own ‘inner logic’ (2003: 14). Ritual actors are, however, often unaware of this inner logic because the ritual actions are the direct expression of the mind; actions are ‘purposely structured’ but the structuring is unconscious (2003: 11-14). Gruenwald thus imagines that rituals, in their raw, pre-theologized state, give us access to fundamental concerns of an individual or group. Chief among these are existential concerns of life and death. Sacrifice deals directly with this concern and is therefore the ‘heart of rituals’ because it works to sustain life at moments of crisis (2003: 180). He argues, ‘a major factor in the shaping of sacrifices is the intimation that a certain reality, or existence, is either under threat or actually undergoing disintegration. The sacrificial ritual is done to prevent this from happening or to repair the damage that has already been done. In the first place, the sacrificial act enacts or repeats an act of breaking: it mimetically repeats the essence of a disastrous event’ (2003: 185).
Gruenwald’s model is thus similar to the theories of Burkert, Girard (whom he cites positively [2003: 259]), and others in that it assumes a single essential purpose behind all sacrifice. Gruenwald is careful to separate what he considers secondary theologizing from what he takes to be the essence of sacrifice, that is, the direct expression of the human mind, but it is hard to see how this can be done since all of the written material he relies on should be classed as secondary theologizing. Indeed, the main tension in the book is the theoretical insistence on ritual as the direct expression of the human mind over and against a complete reliance on secondary written texts to access that expression. This is particularly apparent in the discussion of Halakhic texts, which Gruenwald takes to be excellent sources for the study of ritual (2003: 33). Here rituals are proscribed in handbook fashion. He argues that such manuals prescribe rituals for specific purposes and that people deeply trust the efficacy of the rituals prescribed in their sacred texts. This theory of ritual puts great interpretive weight on texts that prescribe or at least appear to prescribe ritual practices, such as the Hebrew Bible and Rabbinic texts, which are precisely the texts Gruenwald is interested in. However, such texts are neither universally common (there are few analogs in Greek, Roman or early Christian tradition), nor are they explicit enough to actually serve as the blueprint for real rituals, as J.Z. Smith argued (1987: 210). It is not even clear that these texts are broadly descriptive of real rituals in the details they do provide (see Cohn 2013). Gruenwald’s definition of religion itself also overemphasizes texts as the sine qua non of ‘established’ religion (2003: 49).
Gruenwald’s goal of getting behind secondary interpretations of rituals towards the practical goals of the ritual actors themselves is critical, as is his assertion that people performing rituals are often not immediately reflective about the discursive meanings of their actions. However, his final conclusions do not differ greatly from earlier models. He has simply pushed secondary reflective meaning and essence into a preconscious realm. The study offers a grand theory of sacrifice in which the essence of all sacrifice is taken to be rectification of disastrous events. Other models, such as reciprocity, are never entertained. Here still, a small body of religious texts is being read to reveal the essence of sacrifice in all times and places.
The Future of Grand Theories
These grand theories of sacrifice pay insufficient attention to the complexity of the data and to basic problems of definition. They often focus on very different aspects of the ritual to the extent that they almost appear to be speaking about different things. This is particularly apparent in the divide between Burkert and Girard on the one hand and Detienne and Vernant on the other. Girard and Burkert focus on killing as the key aspect of sacrifice, paying little attention to the later eating of the animal. Detienne and Vernant focus on eating, paying little attention to the killing and actions preliminary to the killing. Their theories are hardly talking about the same thing. This suggests that ‘sacrifice’ is simply too broad, disparate and synthetic a category to be the subject of a single grand theory, as McClymond has argued (2008). Theorizing specific aspects of the synthetic and amorphous category ‘sacrifice’ promises to be much more fruitful for scholars. For example, we might put forward theoretical analysis of reciprocity, or prayer, or the imagining of non-obvious beings.
The problem ultimately comes down to one of definition. Theorizing sacrifice as a single category or phenomenon means (1) defining what is in that whole and what is not; and (2) determining what aspects of the whole are most salient. Both of these are truly not the purview of scholars but of participants in sacrifice. Perhaps the best way to move forward is to dispense with grand theories and heed Latour’s (2005) advice to ‘follow the actors’. Participants in ancient (and modern) sacrifice are constantly making determinations about what matters in sacrifice, what counts as a sacrifice, and what sacrifice does (Sheehan 2009). Theories that take the category sacrifice as universal or predetermined are more participations in sacrificial discourse than they are true theorization of ritualization, in that they are making a normative claim about what ought to count as sacrifice.
The current move away from grand theories in the study of sacrifice parallels a larger discussion in the critical study of religion. Once it is recognized that religion (as the term is normatively used) is not a natural category—an ontological ‘thing’—theories that assume that it is (as most of the classical theories in religious studies do) no longer make sense. The same can be said for sacrifice. Once sacrifice is seen as an ever changeable and contestable series of discursive claims and/or ritualized behaviors, theories that see it as a single phenomenon with a single meaning or essence cannot stand. Scholarship is just now entering a phase without grand theories. This will require scholars to be more reflexive about what, exactly, they are studying under the term ‘sacrifice’, and why. Moving forward without these structuring, yet flawed, theories means closer analysis of the available evidence for ritualized practices themselves, as well as more theoretically sophisticated analysis of the discursive claims made about sacrifice by various interested parties in the ancient Mediterranean.
New and Reassessed Archeological Evidence
Fortuitously, we have a wealth of material evidence for sacrificial practices: huge temples, household altars, inscriptions, artistic depictions, and even caches of bones. In comparison to other religious practices like prayer or song, we have a lot to look at when analyzing sacrificial practices. Brief overviews of this evidence can be found in Burkert (1972) (with the caveat that his theoretical model is flawed, as discussed above); Beard, North, and Price (1998: II, 148-65); Larson (2007); and Schwartz (2012).
Historically, archeological evidence for sacrifice has played a significant role in the interpretation of the ritual, but often in problematic ways. Meuli and Burkert were deeply influenced by supposed evidence for sacrifice among Paleolithic hunter-gatherers (Burkert 1972). When this archeology was challenged, Burkert’s theory was cast into doubt (for a critique of the evidence for Paleolithic sacrifice see J.Z. Smith 1987; for a discussion of archaeological evidence for very early sacrifice see Porter and Schwartz 2012). Likewise, several scholars were influenced by artistic depictions of sacrifice, many on vases (Durand 1989; Van Straten 1995). These vases almost never show the actual moment of sacrificial killing; rather, they focus on the preparations for sacrifice, the burning of offerings, and the meal that follows. Detienne and Vernant argued that this illustrated the true importance of sacrifice—the ritual meal not the kill. However, attention must be paid to the artistic medium of the vases themselves. What are the realities of the production, purchase and use of these objects? How useful are they in telling us something about how a significant percentage of people interpreted sacrifice? These examples show the difficulties in analyzing archeological data, and more recent work has been more circumspect. Material evidence, however, remains an important class of data and recent scholars have explored it in fruitful new ways.
Naiden (2012, 2013b), building on the work of Ekroth (2002, 2007b, 2008a, 2008b), has made use of material remains to address the very basic—but critical and surprisingly neglected—question of how much meat was really produced by sacrificial practices and how this amount compares to estimates of civic populations. This is a critical issue for questions of participation, hierarchy and the ancient meat economy. It has often been assumed that all, or nearly all, of the meat in an ancient city would have come from animal sacrifices at the local temples. The communal sharing of this meat—albeit in strategically unequal ways that indexed hierarchies—is a key part of many theoretical understandings of sacrifice, including those dependent on Vernant and Detienne and, to a lesser extent, Burkert.
Naiden argues that, in light of a consideration of the material evidence, there simply was not enough meat to go around. Approaching this question, as Naiden recognizes, involves a lot of difficult estimates: average yield of meat per animal, number of animals sacrificed in any particular location, and the population of that locale. None of these are easy to determine with any accuracy. Drawing on a range of archeological and zoological studies, Naiden concludes that, in those locales where evidence allows for conclusion, such as Athens and Delos, major sacrifices yielded little meat in relation to the size of the population. Either people received only a tiny token amount or a small number received more reasonable portions and the majority received nothing. To the archeological evidence Naiden adds texts and inscriptions, which often reference sacrifices of only one animal and rarely mention the division of meat. For example, the sacrifices of the Attic Demes frequently consisted of a single animal for a group averaging 200 people, not nearly enough to give everyone a meal. Naiden thus concludes that Greek sacrifices most often provided meat for only the group of elites performing the ritual (priests, patrons and honored guests termed parasitoi); others got nothing (Naiden 2012, 2013b).
Because of the difficulty in estimating the variables involved, Naiden’s numbers can be challenged at every point, and certainly much more analysis is needed, ideally by scholars versed in the special methods of zooarchaeology and faunal analysis. Nevertheless, Naiden’s basic point may be true; we may have to modify our imagination of a grand civic sacrifice from an image of a huge communal banquet to something much more hierarchical, in which participation in sacrifices, particularly for those on the periphery, does not necessarily mean getting something to eat.
Naiden’s work challenges the centrality of specifically animal sacrifice to ancient Greek religion. This line of argument is picked up in two other pieces in the volume from the University of Chicago conference. Neer (2012) analyzes examples of Greek statuary to broach the question of the relationship between monumental art and architecture and the practice of sacrifice itself. He concludes that Greeks, particularly Athenians, spent more money on buildings like the Parthenon, than they did on sacrifice itself, but he himself points out that it is unclear what this means for the question of the centrality of animal sacrifice to Greek life. The art simply cannot answer this, nor is it clear that relative financial expenditure is directly tied to importance.
Elsner (2012) asks the same question of Roman art and Roman animal sacrifice. He concludes that a study of Roman representations of sacrifice show a move away from animal offerings towards libations and incense offerings beginning in the third century. He concludes that Roman religion was already moving toward bloodless offerings before Christianity—with its anti-animal sacrifice invective—came to prominence. Elsner’s argument faces the same challenge of how to relate a shift in iconography to a shift in practice or relative importance of practice, a problem he fully acknowledges. His study is another note of caution that scholars have overemphasized the importance of specifically animal sacrifice as opposed to other types of offerings. For further discussion of sacrifice in Roman art, see Huet (1996, 2005); on images of women’s sacrifice in Roman art, see Huet (2008).
Patton (2009) offers an interesting and thorough analysis of pictorial depictions of gods involved in sacrificial practices. These depictions of deities engaged in sacrificial acts have raised interesting artistic and ideological questions. What do these scenes depict? How can the gods perform sacrifice? To whom are they sacrificing? Patton argues that these scenes reflect a belief in the gods as sources for cultic practices, what she calls ‘divine reflexivity’ (2009: 13). Her work illustrates the challenges of interpreting artistic depictions of religious practices and the difficulty of determining what such images might have meant to those who produced and consumed them.
The growing reintegration of archeological evidence with textual evidence for rituals promises to greatly increase our understanding and overcome the problematic privileging of written texts (Stowers 2011b: 45-46). Other fruitful examples include: DesRosiers (2013) on the Athenian Anthesteria; Rosenblum (2013) on evidence for Judean household sacrifice; and Nasrallah (2011) on depictions of sacrificial blood and violence in Roman imperial iconography. Finally, a conference volume on household religion in the ancient Mediterranean edited by Bodel and Olyan (2008) includes several entries addressing our evidence for sacrifice within the ancient household and its relationship to civic sacrifices (see, especially, the essay of Faraone).
In the area of material evidence, food and meals traditions have recently become a fruitful topic of scholarly research and this has added to our understandings of the practices involved in and surrounding sacrifice. Important works include: Klinghardt and Taussig (2012); Rosenblum (2010); D.E. Smith (2003); Smith and Taussig (2012); and Stowers (2011a). Much of this work has focused on the transition between sacrificial meals and the Christian Eucharist (McGowan 1999, 2012, 2014a, 2014b).
There has also been a growing reassessment of archeological evidence for offerings among early Christian groups. It was once common to view such evidence as examples of incomplete Christianization or deviant Christian groups, but the growing weight of the archeological data demands a reintegration of sacrifices and offerings in early Christian contexts. Given how much archeological evidence remains from Christian tombs and martyr shrines, it is now impossible to deny that physical offerings (there is no reason not to call them sacrifices) continued in Christian contexts well into the fourth century. MacMullen (2009) offers a provocative exploration of this evidence (see also MacMullen 2010). Other sources for this evidence include: Jensen (2008); Goddard (2006); Belayche (2005); Corley (2010); Kovaltchuk (2008); Snyder (1985); and Trout (1995). Analysis of Christian responses to sacrifice must take this evidence seriously. There is a clear disparity between what Christian religious experts said about sacrifice, and what non-experts actually did.
Inscriptions represent another important but challenging class of data. Naiden (2013b) is a good example of the integration of this evidence into a diachronic study of sacrifice. The continued publication and digitization of documentary papyri, mostly from Egypt, promises to shed new light on the realities of local sacrificial practices. Some of this data have been collected in Llewelyn and Harrison’s New Documents Illustrating Early Christianity (1998), which now runs to ten volumes. Frankfurter (1998) also makes good use of this evidence in analyzing Egyptian sacrifice. However, no full-scale study of the evidence for sacrifice in papyrological sources exists. It would be as valuable as it would be daunting to produce.
Finally, newly emerging analytic techniques promise to draw more information out of existing and future archeological finds (for an introduction to these scientific techniques see Brown and Brown 2011). One such technique, strontium isotope (87Sr/86Sr) analysis, provides the ability to track the geographic movement of humans and animals from their calcium based remains (bones and teeth). It can determine, for example, whether sacrificial remains consist of local animals or animals that were brought in from other areas, possibly for the sole purpose of sacrifice. In addition to providing insights into the movement of domestic animals for sacrifice, strontium isotope analysis may provide new evidence for how far the sphere of influence of specific cult sites extended in different time periods. On this technique, see Glascock et al. (2007).
Discourse on Sacrifice
Aside from this rich archeological data, the main evidence for sacrifice, and understandings of sacrifice in the ancient Mediterranean, comes, of course, from written texts. With the exception of documentary papyri, discussed above, the discovery of new literary texts that discuss sacrifice is rare, but not unheard of. For example, the publication of the Gospel of Judas in 2006 by the National Geographic Society added a new voice to early Christian discourse on sacrifice (for analysis, see King 2013).
For the most part, however, our ancient textual evidence is well known and has been discussed and analyzed repeatedly. This has led to a variety of scholarly debates over how to theorize and interpret this evidence. In order to organize this diverse body of scholarship, I have chosen to focus on five topics where important disagreements are being worked out and fruitful new work is being done. They are: (1) the role and reach of experts in our sources and the extent to which we can make claims about what sacrifice meant to large portions of the ancient population; (2) the question of sacrifice as a form of communication and the question of failed sacrifice; (3) the interplay between real physical offerings and metaphorical offerings and the usefulness of the concept of spiritualization; (4) the ways in which sacrificial terminology is used metaphorically; and (5) the use of sacrifice in discourses and practices of identity formation. I see these areas of scholarly contention as the most important and promising areas of current research. The final two sections address sacrifice in specific contexts of ancient Judean religion and early Christianity.
Symbolic Systems and the Reach of Experts
Although the search for singular cross-cultural meanings, which characterized much of twentieth-century scholarship, has fallen from favor, the idea that sacrifice constitutes a coherent symbolic system is supported by several current scholars. Few, or perhaps none, would argue that sacrifice never holds symbolic meaning; the question is how broadly such meanings are spread. This question has led to a current debate over the universality of any particular symbolic interpretation. The debate can best be seen as a disagreement over the role and reach of religious experts. One side of this debate asserts that ancient general populations often shared common symbolic understandings of sacrifice. The other side argues that such symbolic interpretations of sacrifice are the products of particular religious experts and that the general population was unlikely to be familiar with—or particularly interested in—the idiosyncratic formulations of such experts. This debate is significant because the vast majority of our evidence for discourse on the meaning and purpose of sacrifice comes from written texts, which are undeniably the products of experts.
Representative supporters of the side that argues for broadly held symbolic meanings of sacrifice include Scheid (2012); Eberhart (2011); and Klawans (2006, 2011). Scheid (2012) builds on his extensive earlier work on Roman religion to conclude that Romans understood sacrifice as establishing social hierarchies in the human world and the superiority of the divine world in relation to the human. Thus, for Scheid, similar to Vernant and Detienne, a Roman practicing sacrifice was made aware of his or her place in an elaborate cosmic hierarchy. This symbolic meaning, for Scheid, is found in Roman texts. While he is careful to stress that he is not offering a grand theory of sacrifice, he does conclude that sources like the epigraphic records of the Arvals and the story of the Vinalia festival provide a way to ‘explain what Roman celebrants understood and meant when they were sacrificing in a traditional public context’ (2012: 86). Scheid’s operating assumption is that these written records are representative of what all or at least most Romans understood sacrifice to be about (see also Scheid 1990 and 2005). (This basic approach and assumption are also reflected in Prescendi 2007.)
Similarly, Eberhart seeks to recover the original ‘“native” interpretation of sacrificial rituals’ for ancient Israelites by a close analysis of sacrificial terms in the Hebrew Bible (2011: 17). He concludes that the preponderance of terms for sacrifice in the text relate not to the act of killing, but to the act of burning. Thus, he concludes that the ‘original communities and tradents’ of the text saw cultic sacrifice as primarily an act of burning things for God. The assumption, again, is that the technical terminology of the Hebrew Bible reflects the basic sacrificial understandings of a community, not simply the persons responsible for the text. The assumption of unified meaning is also seen in Eberhart (2011).
Klawans enters this debate from a slightly different angle. He argues that symbolic interpretations are commonly assumed for many religious practices, but that they are often denied to animal sacrifice, creating the assumption that sacrifice is an act devoid of meaning or spiritual significance. This devaluing of sacrifice as meaningless or empty ritual is part of the evolutionary and supersessionist trends that have plagued religious studies generally (Klawans 2006: 3-5). Klawans advocates for an analysis of the ways in which sacrifice served as a symbolic system in ancient Judean religion, in line with his earlier work on purity (2000). He advances the argument that the ancient Judean sacrificial cult was a rich symbolic system based on a model of imitaio dei: ‘as Israel is to Israel’s herds and flocks, so too is God to Israel’ (Klawans 2006: 61). The symbolic system of sacrifice thus echoed ancient Judean conceptions of the relationship between the community and Yahweh (2006: 49-73).
In opposition to these views, another group of scholars has argued that the symbolic interpretations of sacrifice preserved in written texts derive from a particular class of religious experts in the ancient world and cannot be taken as representative of a broad or general understanding of sacrifice in the ancient Mediterranean. Gilders (2011), Stowers (2011b) and Ullucci (2011, 2012, 2013) are representative of this position.
Stowers (2011b) attempts to theorize broadly about religion in the ancient Mediterranean. Building on insights from cognitive theory and practice theory, Stowers argues that we must distinguish between different practices and the forms of capital involved in each. Making this distinction, he argues, helps us understand the different practices and players involved in sacrifice. At the heart of Stowers’s theory is a distinction between performing animal sacrifice and creating complex interpretations of sacrifice. He argues that these two very different practices are part of what he terms two different modes of religion. The first he calls ‘the religion of everyday social exchange’ (2011b: 36). This mode of religion focuses on the assumption that the gods are powerful figures with whom humans can cultivate relationships of reciprocity. (On reciprocity in sacrifice see Atran 2002: 114-46; Gill et al. 1998; Høgh-Olesen 2006; Linders and Nordquist 1987; Satlow 2013; Seaford 1994; Parker 1998; Ullucci 2012: 24-35.) Building on cognitive theory, Stowers show how the practices of ‘the religion of everyday social exchange’ derive intuitively from the assumption that the gods are powerful agents interested in human affairs. Sacrifices, offerings and prayers are intuitive means of cultivating relationships of reciprocity; they do not require symbolic meanings, reflexive interpretations or complex theologies (2011b: 37-41; on practical concerns in sacrifice, see also Boyer 2010: 18).
Distinct from the religion of everyday social exchange is another mode, the ‘religion of the literate cultural producer’ (2011b: 41). The religion of the literate cultural producer involves different practices, such as parsing of rituals and the creation and interpretation of written texts. The religious experts who make up this mode are very few, accounting for, Stowers estimates, two percent or less of the population. These experts are ‘parasitic’ upon the religion of everyday social exchange in that they claim to represent true knowledge of the gods and correct understandings of rituals like sacrifice (Boyer uses the metaphor of ‘veneer’ for the same phenomenon [2010: 36]). They also attack the logic of reciprocity, which governs the religion of everyday social exchange by claiming that, without their complex symbolic interpretations, rituals like sacrifice are simply base attempts to bribe the gods (Stowers 2011b: 41-46).
Stowers argues that making the distinction between these two modes is critical because religious studies in general has historically privileged the voices of the small number of religious experts and allowed their understandings of sacrifice to speak for the whole population. Undoing this means de-privileging written texts as avenues for understating what sacrifice meant to the majority of ancient practitioners (2011b: 45-46).
Gilders (2011) takes up this approach in his analysis of Philo of Alexandria. By arguing that Philo was a literate cultural producer of the type Stowers outlines, Gilders places Philo in a coherent field of philosophical production and competition, without assuming that he is representative of first-century Judean understandings of sacrifice (2011: 102-103). My own work (Ullucci 2011, 2012, 2013, 2014) has built on Stowers’s model by further explicating the ways in which competition worked within the religion of the literate cultural producer. Ullucci (2011 and 2013) take up the idea of critique of sacrifice (long a part of the discussion of the notion of spiritual sacrifice, discussed below) and reframe it as the process of cultural producers in competition. Ullucci (2013) proposes a typology of religious experts in the ancient Mediterranean and their competition over sacrifice based on Stowers’s model.
Critics of Stowers’s position claim that it is elitist in that it denies theological sophistication to non-experts. Stowers never makes such a claim, through others have. Gruenwald, for example, makes the alarming claim that average ancient Mediterraneans were intellectually incapable of advanced theological reflection (2003: 4-5, 56). Stowers’s position is not that non-experts are incapable, but that they are uninterested. This debate plays out in the pages of the conference volume from the 2008 Boston University conference on sacrifice. The essay by Klawans (2011) is in large part a response to Stowers’s paper in the same volume. For supporters of Stowers, the purpose is not to deny symbolism and theological sophistication to non-experts, but to show that such second-order meanings and interpretations are not required for the performance of the practice. Not every ancient person was invested in reflecting on the meaning and purpose of sacrifice (Gilders 2011: 103). More importantly, we have little or no access to their reflections if they did make them. Assuming that their interpretations match those of experts gives ancient experts exactly what they wanted but never had—a privileged place from which to dictate what sacrifice should mean for everyone, and the power to make them listen (Stowers 2011b: 37).
Gruenwald (2003) represents an unusual entry into the debate I have just outlined. His theoretical framework would suggest that he is on Stowers’s side in arguing for the secondary nature of theological explications of sacrifice. However, his final formulation places him on the side of those who argue for broad universal meanings for sacrifice. For Gruenwald, sacrifice is the spontaneous, pre-theological expression of the human mind. Written texts, like the Rabbinic writings, may record ritual procedures, but for Gruenwald, rituals themselves always start with the individual’s attempt to cope with existential crises. Gruenwald argues strongly throughout his book that theological ‘meanings’ and interpretations are secondary to the practice of sacrifice itself. Participants in sacrifice do not act based on these secondary interpretations, though they may provide a long-term support structure for practices. Given this position, Gruenwald appears to be on the side of those who say all interpretations are secondary. However, this impression is incorrect for two reasons. First, the universal ‘pre-theological’ meaning that Gruenwald ascribes to sacrifice is so full, complex and explicit that it amounts to a universal symbolic interpretation of sacrifice. Second, Gruenwald’s distinction between primary, ‘pre-theological’ interpretations of sacrifice and secondary discursive interpretations of sacrifice is arbitrary. His theory shifts what others would call a secondary interpretation into the pre-reflexive mind of every sacrificer, but there is no way to empirically substantiate this move, nor is there any attempt by Gruenwald to do so. All of Gruenwald’s evidence consists of secondary texts written by experts. Ultimately, his position is that, for ancient practitioners, sacrifice was about rectifying existential crisis, whether or not they were consciously or reflexively aware of this. This single meaning model puts Gruenwald closer to the ‘broad symbolic meaning’ side of the current debate.
What is really at stake in this scholarly debate is the representative value of our entire body of evidence. To put it in the starkest terms: do our texts from the ancient world give us a representative window into what most ancient people believed about sacrifice or do they represent only the idiosyncratic speculations of a very small portion of the ancient population—speculations that the majority of ancient people would have found quite possibly uncompelling and quite probably uninteresting? Where this debate goes in coming decades will deeply impact the shape of future scholarship on sacrifice.
Communication, Rejection of Sacrifice and Determining the Will of the Gods
If sacrifice is ritualized reciprocity, as several of the above scholars have argued (Stowers 2011b; Ullucci 2012), it is also ritualized communication, since reciprocity is intimately tied to the maintenance of relationships and the communication required to do this (on ritual communication in general, see Stavrianopoulou 2006). Naiden (2013b) analyzes sacrifice as ritualized communication in which humans send signals to the gods and seek signals in return. The book expands on an earlier article focusing on instances of rejected sacrifice in Greek literature (Naiden 2006), arguing that it is often in instances of failure that the nature of sacrifice as a system of communication is most apparent. His book contains an appendix of all such rejected sacrifice from Homer to Sopater (fourth century
Naiden analyzes many instances of approved and rejected sacrifice as attempts at theophilia. Many of these sources are literary texts that describe positive or negative signs from the gods indicating approval or disapproval. This is the critical point for understanding sacrifice as a form of communication. Naiden refers to sacrifice as a performance with an audience of one (the god), but this is perhaps misleading, as overemphasis on the ‘god’s response’ can mask the real practices at work. Whom is the performance for? What is being communicated and to whom? Seeing sacrifice as an act of communication is complicated by the reality that it is communication with beings that do not exist. It is imagined communication. The gods are certainly real to the people making sacrifice, but they do not give empirically verifiable replies. Who gets to say what constitutes a ‘reply’ from the gods? Who gets to say when they are happy and when they are not?
The role of ‘interpreting the gods’ will’, or more precisely speaking for them, is filled by a whole range of religious experts. Some imagined signals are, of course, so obvious that little interpretation is needed. In a cultural context in which the gods are assumed to control the weather a devastating flood or drought is obviously a signal of divine displeasure and, by extension, a signal that previous sacrifices did not achieve their goal. But just as in human communication, most signals are far more subtle and ambiguous. Even seemingly obvious signals like the one just mentioned require parsing to be intelligible (which god sent the flood? Why? What sacrifices, specifically, were not well received? What will appease the god and repair the relationship?).
Naiden argues that the Greeks ‘delegated’ the responsibility for interpreting the gods’ will to persons who had the time, money and expertise to do the job. Naiden’s delegation model assumes that the community willingly cedes authority to a class of religious experts who are recognized as capable of mediating between the gods and community. This model, however, ignores the role of religious experts themselves in coopting power and authority by claiming to speak for the gods. This opportunity to speak for the gods is a byproduct of the practices of sacrifice itself. The very fact that sacrifice is reciprocal communication with imagined beings leaves an opening for a class of experts who can claim to speak for those beings. The more directly the actions of these beings are tied to everyday life events, the more opportunity arises for experts to assert authority by claiming to know the gods’ will. Sacrifices cannot always be effective at gaining the gods’ goodwill because things do not always go well in the real world. This is a problem of ‘conceptual control’ (Barrett 2004); claims about the activities and powers of non-obvious beings must be couched in terms that are not easily or at least directly falsifiable. Discussion of failed sacrifice plays a role in this by outlining various ways in which sacrifice could fail to achieve its desired effect—it is a strategy for maintaining conceptual control by explaining why sacrifice does not guarantee unmitigated good fortune.
As Naiden points out, there were both perks and dangers in being a religious expert in the position of speaking for the gods. There was significant authority and power associated with the job, but also serious liability. Catastrophes, both human and natural, could always be blamed on improper sacrifice or other aberrant behavior of religious experts. Nevertheless, in most ancient Mediterranean contexts, the people responsible for making sacrifices were the elite men of the society, suggesting that the perks far outweighed the hazards. But commenting on the efficacy of sacrifice is not the sole prerogative of the people actually performing the sacrifice. Many of Naiden’s examples are from poets and tragedians. This suggests that the group of experts who claimed to speak for the gods was neither homogeneous, nor univocal. Different interested parties were involved in competition to divine the will of the gods.
McClymond (2011a) has analyzed ritual failures in a way that focuses attention on this strategic coopting of power by different religious experts. Whereas Naiden focuses on the Greek tradition, particularly Greek tragedy where the majority of failed sacrifices are recorded, McClymond focuses on the Mishnah, another text that frequently discusses ritual errors and their outcomes (see also Cohn 2013).
McClymond shows that the Rabbinic authors of the Mishnah were claiming the authority to determine what was and what was not correct ritual—they were claiming to speak for God. The discussions of failed ritual in these texts serve as a medium for the authors to display their textual and exegetical expertise. The text ‘establishes a new intellectual system, one that replaces priestly authority and ritual practice with rabbinic authority and ritual argument’ (2011a: 244). Although the cultural and literary context of the Mishnah is quite different from the texts Naiden addresses, McClymond’s arguments have significance for understanding all debates over proper and improper sacrifice. McClymond’s analysis reveals not a delegation of authority, but a competition for it. She also shows how far removed competition over interpretations of sacrifice can be from sacrifice itself. The Rabbis are debating the efficacy of rituals that have not taken place in generations (since the Temple’s destruction in 70
The fact that determining what is and is not acceptable sacrifice goes on even when the actual practices have ceased is good evidence that such systems of evaluation were always a game played by experts. Greek and Roman thinkers parsing sacrifice are involved in the same game as the Rabbis. Ultimately, it is always humans who get to say what constitutes evidence for approved or rejected sacrifice. The privileged position of speaking for the gods was always contested.
Spiritualization or Not?
The question of what really matters in sacrifice, physical things offered or metaphysical dispositions and beliefs is as old as Plato, and it occurs again and again in classical literature and on into late antiquity (for examples, see Attridge 1978; Ullucci 2012: 31-64). This question also played a major role in early Christian theology and apologetics, in which early Christian writers argued that Christianity was superior to traditional Mediterranean religions, Judean religion included, because it offered pure, spiritual worship, which is what God really wanted, not empty and meaningless sacrifice. The question also played a role in later Catholic/Protestant disputes over correct Christian ritual, the aesthetic of Christian art, architecture and worship, and the essence of Christianity.
Given how historically prominent this discourse over ‘spiritual sacrifice’ has been within Christianity and Judaism, it is not surprising that scholarship on ‘real’ (physical) sacrifice has been deeply affected by this debate. Two main trends are observable. The first sees a historical process of ‘spiritualization’ of sacrifice beginning with Greek philosophy and ending with Christianity. I have called this the ‘critique of sacrifice’ model, as it posits that various ancient voices critiqued what they saw as theological flaws in the practice, leading to its eventual rejection under Christianity (Ullucci 2012: 4-14). Those who reject the critique model (Klawans, Ullucci et al.) argue that it is evolutionary and supersessionist—claims to ‘spiritual’ sacrifice need to be recast as a type of competitive discourse.
The basic outlines of the critique model are to be found in the ancient texts themselves. Greek philosophical texts often attack the stereotyped ‘ignorant masses’ who purportedly believe that sacrifice can buy the gods’ favor. Hebrew Bible texts also claim that Yahweh really wants obedience and devotion, not dead animals. The theme is played out in various genres, from Platonic dialogues to the satires of Menander and Lucian. It is made even more explicit in Christian texts, which claim that sacrifice is rejected, replaced and superseded by the death of Jesus and Christian practices such as the Eucharist.
Scholarship following this model traces a progression from physical offerings with no spiritual meaning, to physical offerings with a deeper spiritual intent, to a final phase where the physical offerings were discarded leaving only spiritual worship. Ferguson (1980) is one example of this approach. More recently, versions of this position have been argued by Daly (2009; this work builds on extensive earlier work that traced the development of Christian positions before Origen, Daly 1978a and 1978b), Finlan (2011) and Stroumsa (2005). (For Stroumsa, unlike the other authors just mentioned, the pinnacle of spiritual sacrifice is reached in Rabbinic Judaism, not Christianity.) Each of these works takes the basic approach that previous critiques of sacrifice find their fulfillment in a rejection of the practice.
In opposition to this line of argument, Klawans (2006) argues that models that invoke an opposition between spiritual sacrifice and non-spiritual sacrifice most often mask evolutionary and supersessionist views of religion, which derive largely from Christian apologetics (though modern Jewish apologetics plays a role here as well, as Klawans points out [2006: 8-9, 31-31]). Such models view physical offerings as belonging to a primitive phase of human religious development, which was superseded when humanity advanced towards a more ‘enlightened’ age. Klawans shows clearly how these theological models (largely Protestant) have skewed scholarly understanding of ancient religion, particularly animal sacrifice. There is nothing inherently ‘unspiritual’ about animal sacrifice. Such claims are simply thinly veiled religious polemic. Klawans argues that in the ancient Judean and early Christian contexts, the importance of sacrifice and the Temple are illustrated by numerous practices that ‘channel the sanctity of the temple into other realms of daily life’ (2006: 106). Rather than critiques of the Temple cult, or spiritualization of sacrifice, these extended sacrificial metaphors and symbols simply illustrate how important the Temple cult was for many ancient Judeans, even those who lived far from the Temple (2006: 103-144).
My own work builds on Klawans’s critique of evolutionary and supersessionist models in the study of animal sacrifice (Ullucci 2012). I attempt to show that the practice of sacrifice has long served as a locus for religious experts in competition. Most of these authors did not want to actually end sacrifice; they simply wanted sacrifice interpreted their way. Rather than a linear supersessionist model, I attempt to illustrate a field of intellectual production and competition. Philosophers, the authors of some of the books of the Hebrew Bible and some early Christian writers were all part of this field of competition. Within this model, claims to spiritual or pure sacrifice are competitive positions being carved out by cultural producers in a variety of complex historical and cultural situations. Terms such as ‘spiritual sacrifice’ vel sim. are strategic and competitive terms, not descriptive ones.
The two sides of this debate are clear, but there is, as yet, no indication of a winner. Finlan, for example, has argued against the ‘anti-spiritualization’ trend in scholarship, specifically responding to Klawans and citing what he sees as spiritualization of sacrifice in Paul and Hebrews (2011). Finlan argues that the term ‘spiritualization’ can have a broader meaning than its critics allow. He uses the term to refer to sacrificial metaphors which, he claims, undermine the ideology of physical offerings (2011: 90). His argument still, however, bears the basic value judgment common to all supporters of the critique model, that not sacrificing is better than sacrificing. It is this position—fundamentally theological, not scholarly—that opponents of the critique model reject.
Even on the side of those who reject the notion of spiritualization, there is disagreement on the shape of the debate over proper sacrifice, as discussed above. Klawans argues for a broad symbolic interpretation of sacrifice among the general public (2006, particularly 2011). Stowers and myself have argued that such complex discursive interpretations of sacrifice are the products of religious experts in competition (Stowers 2011b; Ullucci 2011, 2012). Naiden’s (2013b) argument for delegation of interpretive authority is also part of this debate.
Sacrifice as Metaphor and Evaluative Term
It makes good sense that ancient Mediterraneans would use sacrifice, a prominent and important practice of their world, to symbolically interpret other things in their lives. This metaphorical interpretation can be based on homologies with the physical practices of animal sacrifice, such as giving, killing, blood, fire, eating, reciprocity, and so on. Alternatively, symbolic interpretations can expand to the point that the metaphorical ‘sacrifice’ has no connection at all with offering practices. As we saw above, ‘sacrifice’ is not a natural category with ontological status; it is a contested nexus of ritualized practices and/or discursive interpretations (McGowan 2014b; Sheehan 2009). Given this, the symbolic and metaphorical possibilities for the word sacrifice are endless. The best example of this is the Christian interpretation of the crucifixion of Jesus—an event that has almost no similarity with sacrifice whatsoever—as not simply a sacrifice but the paradigmatic sacrifice. Like all metaphors and symbols, calling Jesus’ crucifixion a sacrifice not only altered the meaning of that event, it widened the symbolic range of the word sacrifice itself, pushing it in new and different directions and opening the door to further metaphoric and symbolic expansion (to the Eucharist, for example).
The interpretation of Jesus’ death as a sacrifice—and the spread of this interpretation via Christianity’s success in the fourth century—may ultimately be the most historically significant ancient symbolic interpretation of sacrifice. By choosing to elaborate this metaphor, early Christianity set the stage for a complete reinterpretation of what is and what is not sacrifice. This process began towards the end of the first century, but continued well into the fourth and fifth centuries and beyond, ultimately changing the discourse and ritual of the ancient Mediterranean. Christians, of course, were not the only ancient Mediterraneans who employed symbolic interpretations of sacrifice; they built on a rich legacy that has been analyzed by several scholars.
Henrichs (2012) analyzes the use of sacrificial metaphors for human death in Greek tragedy. He argues that Greek tragedians were often preoccupied with depicting sacrifice gone wrong (on this evidence see also Naiden 2013b). The tragedians also commonly use sacrificial terminology for human death. This begins with Aeschylus and reaches a peak with Euripides. These texts explore and expand a metaphorical field based on homologies between the intentional killing of humans in cases of murder and the killing of animals in sacrifice. Henrichs argues that while tragedy blurs the line between sacrifice and murder, ultimately the two categories are maintained as distinct: ‘murder is not sacrifice’ (2012: 194), at least not in these texts.
Várhelyi (2011) addresses metaphorical uses of sacrifice to refer to human killing during the transition from the Roman Republic to the Empire. She argues that, beginning with Tiberius Gracchus, sacrificial images and themes became attached to political murders in Rome. Calling a murder a sacrifice became increasingly common. This association of political execution with sacrifice spurred an ongoing debate between different factions in the Roman political arena over whether a specific person’s death constituted a sacrifice or not. The claim that a particular killing was a sacrifice could be used by interpreters as a strategy to glorify the dead victim and enhance the significance of their death. Denying a sacrificial interpretation did the opposite. Várhelyi thus shows an excellent example of the way in which definitions of sacrifice could be contested and expanded metaphorically for strategic purposes. (For a similar discussion of sacrifice in Roman poetry, see Feeney 2004.)
Boustan (2011) addresses a similar situation of competition over sacrificial interpretation in Judean sources. He shows that Judean authors in Late Antiquity were sometimes interested in deploying sacrificial metaphors for human death, but sometimes rejected such symbolism. His examples are the positive framing of death as sacrifice in The Story of the Ten Martyrs and the explicit rejection of sacrificial symbolism in Rabbinic accounts of the death of Zechariah. Boustan thus illustrates another example of the ways in which sacrificial symbolism—or the denial of sacrificial symbolism—could be used strategically in discourse of power and legitimacy (see also Boustan and Reed 2008).
Many early Christian texts are involved in the same kind of metaphorical evaluation of death-as-sacrifice seen in Várhelyi’s and Boustan’s examples. There is an enormous amount of scholarship on early Christian martyrdom texts; however, this scholarship has rarely addressed, in a comprehensive manner, the way in which sacrificial symbolism is deployed competitively in these texts. Too often, claims that the death of martyrs constitutes true sacrifice are taken at face value, or are incorporated into the supersessionist or ‘spiritualization’ models discussed above. An exception is Heyman (2007), who analyses the phenomenon of early Christian martyrdom as a competitive arena where sacrificial discourse served as a medium for competition over legitimate power. Heyman argues that Christian martyrs appropriated a sacrificial discourse to argue that they, not their executioners, possessed true power. Even Heyman’s study does not go far enough in redescribing early Christian uses of sacrificial symbolism. Heyman often places the locus of discursive competition with the martyrs themselves, in the Roman arena, proclaiming themselves sacrifices before Roman crowds. This approach puts too much stock in the historicity of the martyrdom texts. Moss’s (2013) work has finally put to rest the notion that early Christian martyrdom texts contain much historical fact (on Judean martyrdom literature, see Rajak 2001). This means that the sacrificial symbolism, which is played out elaborately in early Christian martyrdom texts, is very likely a literary creation, not a reflection of the way actual victims conceptualized or advertised their own deaths. Thus, analyzing all martyrdom texts as examples of competitive discourse creation, along the lines Várhelyi sets out for Roman sources, seem the most fruitful. Further discussion of the symbolic interpretation of human death as sacrifice can be found in Nasrallah (2011) and Castelli (2004: 50-52).
Beyond martyrdom discourse, several other studies have illustrated ways in which sacrificial images and terminology were expanded symbolically. Nasrallah (2011) illustrates the use of sacrificial symbolism to positively value correct human behavior, as understood by particular ancient authors. Jacobs (2011) addresses ways in which the incarnation of Jesus was framed as a self-sacrifice by some early Christian writers eager to defend the reality of the Christian idea of the incarnation.
The concept of ‘human-sacrifice’
This work on the competitive use of the label ‘sacrifice’ raises a serious concern for the way in which the term ‘human sacrifice’ is often used, even in scholarly contexts. If, as many of the above studies have shown, the term sacrifice is part of emic discourses of power and legitimacy, then the term ‘human sacrifice’ is seriously problematic as a term of scholarly analysis, since it already implies an evaluation of certain instances of killing as sacrifice. The current scholarship on the ritual killing of humans in various contexts often employs the term ‘human sacrifice’ quite freely as a descriptive term (Bremmer 2007; Porter and Schwartz 2012; Ruane 2013). This is particularly true of work on ‘non-European’ religions, such as those of Africa, the Pacific and the ancient Americas (Carrasco 1999; Aldhouse-Green 2002; Dodds Pennock 2008). Much of the scholarship in the previous section, particularly the work of McClymond (2011b: 326-28), shows how problematic this use of ‘human sacrifice’ is.
The ritual killing of humans is a very different practice from the killing of animals for the purpose of honoring the gods, enacting reciprocity and acquiring food. Grouping these practices together under the term sacrifice certainly happened in the ancient world (it still happens today), but these are first-order emic descriptions that require analysis and redescription by scholars. The term ‘human sacrifice’ is at the heart of the problem because it elides exactly the questions at hand: why is some specific act of killing being called a sacrifice? By whom? For what purpose? Did a young royal retainer, ritually killed and placed in the tomb of her king at his death, conceptualize her own death as sacrifice, murder or apotheosis? What did her mother think? These are the kinds of redescriptive questions that never seem to get asked of ‘human sacrifice’. I suspect that this is because the Christian idea of Jesus’ death as the paragon of sacrifice has made the equation of human killing and sacrifice seem natural and obvious rather than strategic and deliberate.
Perhaps the best redescription and analysis of the ways in which charges of ‘human sacrifice’ were deployed as part of a competitive discourse in the ancient Mediterranean remains Rives’s (1995) analysis of Greek, Roman and early Christian uses of the charge. In this essay Rives shows that charges of human sacrifice in the Roman world were part of a discourse of othering. For Romans, human sacrifice was the work of barbarous foreigners such as Scythians and Carthaginians, or suspicious internal enemies such as Bacchics and Christians (the Christians, for their part, bounced the accusation right back). As Rives shows, claims about human sacrifice are most often fictive. However, as McClymond (2011b), Várhelyi (2011), Sheehan (2009) and others have shown, even when real human death is involved, the term ‘human sacrifice’ is still a strategic and competitive label, not a descriptor. For these reasons, I would suggest that ‘human sacrifice’ is simply not a useful scholarly category.
Sacrifice and identity
Several scholars have explored the ways in which sacrifice served to index group identifications and kinship in the ancient Mediterranean. This line of analysis goes back to the theories of Burkert, Girard, Vernant and others, but has now been taken up without the problematic theoretical frameworks from the past. The classic work analyzing sacrifice and kinship is Jay (1992), who built on the work of Valeri (1985) (see also Johnson Hodge 2007; Boustan and Reed 2008). Current work examines the ways in which sacrificial participation or non-participation were used strategically in discourses of identity. Rives (1999) addresses the Decian decrees on sacrifice, showing how Decius’s administration used sacrifice in an attempt to articulate a broad Roman identity in the sprawling empire (see also Hargis 1999). (For a discussion of sacrificial practices and identity in the imperial Roman military, see Herz 2002.)
The imperial cult itself is an example of the use of sacrifice in discourses of identity. At first, as Price (1984) shows, it was part of a strategic attempt by elites in the Greek east to articulate their alignment with Rome and their relationship to Roman power; later it was part of a Roman imperial strategy to articulate and disseminate a trans-regional and trans-ethnic ‘Roman’ identity (Rives 1999; see also Gradel 2002).
Townsend (2011) shows the ways in which anti-sacrificial discourse was used to challenge some of the very claims to identity just mentioned. She focuses on the philosopher Porphyry and his critique of meat eating, animal sacrifice and the forms of fictive kinship these articulate. Heyman (2007) also points out ways in which refusal to sacrifice—particularly to the emperor—could be framed, in some Christian texts, as a rejection of Roman identity. The link between sacrifice and Roman identity, in all its complexity, is well illustrated by Tertullian’s strained attempts to argue that Christians are the best Romans, precisely because they do not sacrifice (Apology 30.1-6). (On the salience of meat, in particular, as a focus of social prohibitions and identity formation, see Fessler and Navarrete 2003.)
Sacrifice was also a locus of competition for ancient cultural producers seeking to separate themselves from the general population, whom they portrayed as ignorant, provincial and unsophisticated. Martin (2004) focuses on the ancient discourse surrounding ‘superstition’ (Greek deisidaimonia, Latin superstitio). He shows that the meaning of these terms changed from the classical Greek to the Roman period as different cultural producers employed the terms in their own attempts to articulate a superior identity. Interpretation of sacrifice was often a key part of this articulation. For example, Greek philosophers from the classical period often defined deisidaimonia as the false belief that the gods were dangerous and had to be placated by lavish sacrifices. Much later, Christian authors appropriated deisidaimonia and superstitio to refer to belief in the traditional gods and the whole range of sacrificial practices that accompanied it (an ironic shift since superstitio was exactly the word Roman authors had first used to describe Christianity in the early second century). Martin shows how cultural producers like philosophers and Christian authors used sacrifice in their own identity claims to distinguish themselves from ‘common people’ who, they asserted, did not understand the nature of the gods or the purpose of sacrifice. Nasrallah (2011: 145-53) expands on this point in the context of first- and second-century
This same phenomenon can be seen in several comedic texts, particularly those of Menander and Lucian (see Ullucci 2012). These authors often create caricatures of people who hold what the authors see as incorrect views of sacrifice; these views are then held up to comedic ridicule. Graf (2011) provides an analysis of this phenomenon in Lucian of Samosata’s diatribe On Sacrifices. He shows how Lucian’s depiction of an absurdly over-zealous sacrificer works to condemn what Lucian sees as the excesses and ignorance of the non-philosophically minded public. Lucian’s goal is not to reject sacrifice eo ipso, but only to reject what he sees as incorrectly performed or interpreted sacrifice (Graf 2011: 210; see also Ullucci 2012). This is a strategic attempt to build an elite identity, presumably one that would appeal to Lucian’s audience, who share a laugh at the ‘uneducated’ person’s expense. For an alternate view that comedic representations of the gods, particularly in Greek comedy, represent a comedic lowering of the gods to human levels, see Redfield (2012).
Perhaps the most extreme way sacrifice played a role in contesting identity was in the case of groups that refused participation in various forms of offering. In the classical world, refusal to sacrifice was often associated with Pythagoras and his followers. The stories of Pythagoras’s own practices vary; some sources say he abstained from sacrifice of animals, some say he participated. Figures such as Apollonius of Tyana and Porphyry rejected the sacrifice of animals but not non-animal offerings. As discussed above, these practices—perhaps it is better to call them strategic non-practices—were part of attempts to claim superior knowledge of the gods and an elite identity. Abstention on the part of some philosophers, especially Porphyry, and its implications for the wider competition over the meaning of sacrifice, is well analyzed by Rives (2011) and Townsend (2011).
Non-participation could also derive from specific historical situations, absent any actual rejection of sacrifice. The key examples of this were Judeans who lived in the Diaspora. These people did not reject sacrifice, but they believed that Yahweh was the only appropriate recipient of sacrifice and that Jerusalem was the only appropriate place to perform sacrifice. This put many Diaspora Judeans in the position of de facto non-participation. Various rules and exceptions for the slaughter of animals were included in Judean tradition to deal with this unusual situation. The social situation of these people has been examined by Rosenblum (2013).
The earliest Christian groups, such as those around Paul, were in the same position, though it was complicated by the fact that Paul’s audience was made up of gentiles, not Judeans who at least had an established ethnic identity to support their non-participation. 1 Corinthians, for example, negotiates this problem. From the first century through late antiquity, various Christian authors articulated differing views on the relationship between Greek and Roman sacrifices, Judean sacrifices and Christian practices. Non-participation became a serious issue at various times when Roman leaders attempted to use sacrifice as part of their own strategies for creating an imperial Roman identity. Examples include Decius (see Rives 1999) and Julian (see Belayche 2002; Bradbury 1995; Finkelstein 2014; Ullucci 2012: 137-49). Johnson Hodge (2013) explores ways in which sacrificial practices in the household could pose a challenge for early Christians.
Stroumsa (2005) argues that changes in the discourse on sacrifice in Late Antiquity are part of a broader shift in the very concept of identity itself from a more communal identity model to a more individualistic model. He traces the rise of Rabbinic Judaism, Christianity, the new notion of religion as about personal belief, and ultimately the cessation of sacrifice to this transformation.
Discourses of identity based on sacrifice were always subject to flux as political and economic realities changed. Salzman (2011) shows the complex ways in which interpretations of sacrifice, and judgment of the efficacy of sacrifice, intertwined with political, economic and ritual change in the fourth century. She focuses on the writings of the Roman senator Quintus Aurelius Symmachus (c. 345–402). Symmachus wrote at a time when religious allegiance and practice were shifting away from traditional sacrifice and towards Christianity, with the emperor leading the way. Civic sacrifices, which had been funded by imperial and civic leaders, were defunded when these people shifted allegiance to Christianity. The result was a centrifugal dispersal of power away from officials and onto private citizens. Frankfurter (1998) traces a parallel centrifugal dispersal in Egypt, when the Roman emperors defunded civic cults established under Ptolemaic rule. Symmachus was in a parallel situation: if public sacrifices were to continue, they had to be funded by private, non-Christian individuals.
Salzman shows clearly how Symmachus reinterpreted elements of sacrificial procedure and funding to privatize civic cult, and in particular to claim that sacrifices would be efficacious in winning the gods’ favor for the whole community even if they were funded by an individual: ‘according to Symmachus, all that was necessary to satisfy the gods was for powerful private citizens to perform the rite for the benefit of the state’ (2011: 172). The crucial point, as Salzman shows, is that Symmachus’s problem was not simply economic but ideological. His solution required reworking the earlier imperial ideology of sacrifice, which envisioned the emperor as the key mediator between gods and empire. (See Salzman 2002 for a broader discussion of this issue.)
Sacrifice in Ancient Judean Religion
A strong argument can be made for not separating ancient Judean sacrifice from the wider ancient Mediterranean context, as the basic practices are no different. The realities of academic organization have, however, often produced this separation, alternately privileging or ignoring this data. Judean sacrificial practices have also been the focus of some of the strongest evolutionarily and supersessionist interpretations, both from those who see Christianity as a replacement for Judaism and from some within Jewish tradition who see sacrifice as part of Judaism’s past, but not its present or future (on the latter, see Klawans 2006: 8-9, 30-31, 203-209). These theological trends continue to skew scholarly analysis of ancient Judean sacrifice, as Klawans has shown (2006). In this section, I focus on two areas that do make ancient Judean sacrifice unique in the ancient Mediterranean: (1) the ancient Judean textual tradition, which has no exact parallel; and (2) the consequences of the belief—held by many, but not all ancient Judeans—that the only appropriate place to sacrifice was Jerusalem.
The Hebrew Bible, along with the Dead Sea Scrolls, Rabbinic texts, and the broad array of other ancient Judean writings (often collected under the scholarly category Pseudepigrapha) provide an enormous amount of written evidence for the ancient Judean sacrificial tradition, far more than we have for any other group in the ancient Mediterranean. This complex textual tradition, however, creates challenges. As J.Z. Smith points out, despite its appearance, the Hebrew Bible does not contain as much direct evidence for the cult of Yahweh as it may seem. None of the rituals mentioned in the Hebrew Bible could be easily performed solely from the descriptions given in the text (J.Z. Smith 1987: 210). Nevertheless, several scholars have focused on this collection of texts for what it can tell us about ancient Judean sacrifice and how various individuals and groups interpreted this practice.
Large-scale studies include Gilders (2004) and Klawans (2006). Terminology has also been a key focus (see Eberhart 2011; Watts 2011). Modéus (2005) provides an excellent analysis of the ‘peace offering’, specifically, as a marking ritual, but his approach has wide application to all sacrifice. Ruane (2013) analyzes gender in the sacrificial laws of the Hebrew Bible, concluding that the texts are highly concerned with both human and animal gender and control of reproduction. Beyond the Hebrew Bible itself, the writings of Philo (ca. 25
The strand of ancient Judean religion that eventually came to dominance, via connection to political power, around the middle of the second millennium
For those who did not espouse the same ideology of centralization in Jerusalem, alternate temples did exist at different time periods in Elephantine, Leontopolis and Mt. Gerizim, but even these sites could not serve all of the Diaspora (see Gruber 2013: 82-83). Judeans, therefore, had to deal in various ways with the absence or remoteness of a sacrificial site. Examining the evidence for these alternate practices has become a fruitful avenue of study. On Judean household sacrifice, see Rosenblum (2013). On Judean sacrifice outside Jerusalem post 70, see Sandwell (2007), Graf (2008), and Petropoulou (2008: 145-204).
After 70
In the absence of the Temple, creative ritual improvisation would surely have taken place among Judean groups. Some of this has left evidence in our written sources. The Passover, for example, was recast in some Rabbinic texts so that the requirement to sacrifice was considered to be fulfilled by other practices including prayer and Torah reading (see Bokser 1984; Brumberg-Kraus 1999, 2004; Rosenblum 2010; Marks and Taussig 2014). The Rabbinic texts show a general tendency to transform ritual practice into textual argument, as McClymond (2011a) argues (discussed above); see also Rosenblum and Ullucci (2014).
Responses to 70
Sacrifice in Early Christianity
The academic study of religion and indeed the term ‘religion’ itself, in its modern English usage, is largely the product of the Protestant Enlightenment and the European Colonial experience (Arnal and McCutcheon 2013; Nongbri 2013). As a result, Christian, particularly Protestant, interpretive biases have a long history in the field. This is particularly true in the realm of sacrifice. A strong anti-ritual and anti-sacrificial bias, entwined with Protestant anti-Catholic polemics, shaped much early work. Only now is this being overcome in analysis of Christian responses to sacrifice. A number of recent articles and monographs have addressed this topic, all dealing with the same basic issues and concerns: (1) the process by which Christians rejected animal sacrifice; and (2) the Christian interpretation of Jesus’ death as a sacrifice and/or a replacement for sacrifice. The fact that Christianity ultimately became a tradition without animal sacrifice is quite remarkable historically. The offering of plants and animals to the gods had been one of the key practices of the ancient Mediterranean for nearly 10,000 years by the time of the emergence of Christianity, yet within 400 years of that emergence sacrifice had declined significantly and would eventually be banned by Christian emperors. Such a major shift in ritual practice requires explanation. It is also clear that an alternate discourse on sacrifice developed early on within some Christian groups. Jesus’ death, the Eucharist and a host of other things were put forward in Christian texts not only as ‘true sacrifices’, but as superior to all other physical offerings.
It is commonly assumed, even among some New Testament scholars, that the fundamental break with sacrifice came right at the beginning of Christianity, with the teachings of Jesus. Evidence for this, however, is completely lacking. Nowhere in the New Testament is the historical Jesus portrayed as explicitly rejecting sacrifice. Quite the opposite: the New Testament Gospels depict Jesus participating in the Jerusalem Temple cult (Mark 14.12-31 and parallels). Paul is also depicted as participating in sacrifice (Acts 21.23-24; 24.17-18). There is, nevertheless, a strong tendency to push the Christian interpretation of Jesus’ death as a sacrifice and the Christian rejection of sacrifice as early as possible. Heyman, for example, argues that a sacrificial interpretation of Jesus’ death is found in the New Testament (2007, especially p. 116). Other examples of an attempt to push sacrificial interpretations of Jesus to the earliest period (the period of Paul and the Gospels ca. 40–110) include Daly (2009) and Siker (2011). Much of this scholarship revolves around the Letter to the Hebrews, since this is the only New Testament text that unambiguously refers to Jesus as the sacrifice, which replaces the animal sacrifices of the Jerusalem Temple. (Hebrews is also unique in its vocabulary, its assertion that Jesus is a priest and the circuitous way it ended up in the canon.) For recent analysis, see Eisenbaum (2013), who addresses the reasons for Hebrews’ critique of the Jerusalem cult; Ullucci (2012: 90-96); and Finlan (2011), who argues for spiritualization of sacrifice in Hebrews (also discussed above). See also Naiden (2013b: 287), who incorrectly attributes the text to Paul.
Several recent books have offered models for how and why Christianity rejected animal sacrifice. Stroumsa (2005) puts this change into the context of a series of developments taking place in Late Antiquity around the concept of the self, but he does so within an anti-sacrificial, supersessionist framework (2005, especially pp. 59-63). Petropoulou (2008) argues that a direct and powerful encounter with Jesus led early Christians to rethink their religious beliefs (2008: 290). She posits that the idea that God needs nothing was new and original to Christianity and that this led ultimately to the cessation of sacrifice when metaphorical interpretations of Jesus as a sacrifice undermined actual physical sacrifice (2008: 250-81). Her claim that Christians were the first to articulate a god that ‘needed nothing’ is untenable, since this idea was common stock for philosophers centuries before the dawn of Christianity. The idea that sacrificial metaphors ultimately undermine physical sacrifice is also articulated by Finlan (2011), but is critiqued by Klawans (2006) and Ullucci (2012).
My own work (Ullucci 2012) has attempted to show that the process whereby Christians stopped sacrificing is much more complex. What becomes the ‘winning’ Christian position on sacrifice takes a long time to develop and is as much the product of inter-Christian conflict as it is conflict with Roman and Judean sacrificial practices. Rather than an original moment of revelation from Jesus, or even a neat development in sacrificial interpretation, I tried to show that the early Christian texts are evidence of various cultural producers in competition, each deploying discourse on sacrifice for different strategic ends. A rejection of sacrifice cannot be found in the earliest Christian sources, with the exception of Hebrews. When rejections of sacrifice do begin to appear in Christian texts, the positions of different authors are often quite incompatible, even contradictory. For example, the idea that the Eucharist is equivalent to a sacrifice, first voiced by Cyprian in the third century, stands in stark incongruity with Hebrews’ notion that Jesus sacrificed once for all, negating any need for future sacrifices (for further discussion of Cyprian, see McGowan 2014b). Both of these positions conflict with the opinion of Justin Martyr that the Jerusalem cult was only given by God as a stop-gap measure to keep the Judeans occupied in the period of time before the coming of Jesus (see Ullucci 2012: 90-117). This incongruity on sacrifice is illustrative of the varied and contested landscape of early Christianity itself.
The form of Christianity that won had the added problem of attempting to position itself as the fulfillment of, and true heir to, a superseded Judean tradition (it is important to note that not all early Christian groups made this claim). To do this they had to explain: (1) the existence of the sacrificial laws in the Hebrew Bible; (2) the historical existence of the Jerusalem Temple cult; and (3) its later destruction by the Romans. They also had to explain their own non-participation in sacrifice given that their own sacred texts, the Hebrew Bible, clearly mandated them. We should not underestimate how logically and ideologically precarious the beliefs of these early Christians were, or how vulnerable these beliefs were to attacks from other Christians who did not see themselves as heirs to Judean tradition, or even worshipers of the Judean god. The answers that the group that won produced vary greatly and are, again, often incommensurate (Ullucci 2012: 119-36). In the end, it is the success of Constantine and his heirs, with their support of one form of Christianity, that cemented what became the Christian master narrative on sacrifice, namely that it is a rejected practice that God never really wanted, which was superseded and replaced by the death of Jesus and the Eucharist.
King (2013) illustrates the complexity of inter-Christian competition by showing how non-canonical Christian writers and groups participate in this competition to define Christian positions on sacrifice. Unlike the form of Christianity that wins, many of these Christians do not develop a positive sacrificial theology for Christian practices; they do, however, still deploy sacrificial interpretations and metaphors, but in ways that are very different from canonical Christian texts. King’s essay shows that what ultimately became the Christian position on sacrifice was by no means universal or inevitable. It is the product of a long and highly contentious process of cultural production and competition.
Finally, it is important to note that the anti-sacrificial positions of many Christian cultural producers do not necessarily reflect the beliefs and practices of all or even most early Christians. Archaeological evidence makes it quite clear that many Christians continued performing sacrifice to saints, martyrs and honored dead. See the section on archeological evidence above.
Conclusions
Much like the Christian master narrative, the present state of research on animal sacrifice is also the product of a long and highly contentious process of scholarly production and competition. The past two decades have been, in many ways, enormously fruitful. I believe some of the most significant trends shaping current and likely future approaches to sacrifice include:
The recontextualizing of sacrifice within the broader scope of ancient ritual and the de-privileging of sacrifice as the most significant aspect of ancient Mediterranean religious practice.
The de-privileging of ‘blood’ sacrifice and the abandonment of the notion that it is fundamentally different from non-animal offerings.
A move away from viewing sacrifice, and all ritualization, as enacted myth or as actions that encode universal meanings or essences.
A move away from grand theories which attempt to explain all sacrifice as the expression of some universal human anxiety, desire, guilt or violent tendency.
A decoupling of sacrifice and excessive violence.
A better integration of archeological evidence with textual evidence, combined with increased circumspection about how much texts about sacrifice truly reflect what non-experts thought about sacrifice.
The recognition that sacrifice was part of ancient competitive discourses on a vast range of things, including (inter alia): family and group identity; ethnic and civic identity; elite identity; inter and intra-religious polemics; religious and philosophical apologetics; religious authority; the nature of the gods and the universe; plant, animal and human fertility; the legitimacy or illegitimacy of acts of murder; and the propriety of meat-eating. Various people in the ancient Mediterranean were actively engaged in competitive debates over what counted as sacrifice. Thus, the definition of ‘sacrifice’ and its symbolic and range were open for debate among ancient practitioners themselves.
The final point above requires more discussion and promises to be the most significant current scholarly trend in the study of ancient Mediterranean sacrifice. Major and fundamental disagreements remain concerning how sacrifice should be defined, theorized and analyzed. In some cases, different scholars’ conceptualizations of ‘sacrifice’ are so different that they hardly seem to be analyzing the same thing. This appears to be one of the biggest sources of confusion in current work. The category ‘sacrifice’ is amorphous—it just means too many things. See for example the discussion of ‘human sacrifice’ or ‘spiritual sacrifice’ above. There are competing classificatory systems used by scholars in determining the object of study (on competing classificatory systems, see Arnal and McCutcheon 2013). What one scholar considers an example of sacrifice would not be considered sacrifice by another scholar—to say nothing of what non-scholars, confessional theologians, violent religious extremists and barbecue aficionados (see below) consider sacrifice. The broadness of the term is the result of its use in so many different types of ancient and modern discourses (martyrdom, reciprocity, commensality, Catholic transubstantiation, Protestant atonement, etc.) for such a long period of time.
This situation of confusion over what we mean by ‘sacrifice’ is not unique; scholars constantly deal with a range of amorphous and contested categories (e.g., religion, Christian, terrorist, etc.). This situation does, however, place a demand on scholars who would enter the world of sacrifice; they must be clear about what they mean by the term. Scholars must be reflexive and transparent about what, for the purposes of analysis, they are considering sacrifice and why. This simple act of definition could clarify an enormous amount of the confusion seen in the present literature. We cannot escape the problem that our categories (sacrifice, religion, etc.) are created, not given. But this only becomes a significant problem when these created categories (along with their created boundaries and distinctions) are taken as—or presented as—natural and universal. There is no right definition of sacrifice, and the definitions scholars use—simply because we must use definitions—must not be presented as universal or definitive.
Here again, I think, current theoretical work on the created category religion is helpful to scholars interested in sacrifice. Nongbri (2013) argues that it should not be the goal of scholars of religion to find the most ‘correct’ or ‘best’ definition of religion, but rather to analyze the activity of defining religion, eo ipso. He advocates ‘giving up on the essentialist project of finding “the” definition’, in favor of asking ‘Who is doing the defining and why?’ (2013: 155, emphasis original). This strikes me as a very valuable approach to studying sacrifice. It does not mean that we surrender our right as analysts to define our object of study; it simply means that we must be transparent about the definitions we create—first, that they are in fact created, and second the reasons behind their creation. More importantly, we must recognize that sacrifice was an amorphous and contested category millennia before our academic institutions were formed. Our goal is not to find the most correct definition but to find the most useful one to whatever project of redescription we are undertaking. These scholarly definitions must always remain provisional and redescriptive, not normative or dogmatic.
Coda: Handbooks on Sacrifice and Disseminating Scholarship
I would like to end this exploration in a slightly unusual place: the various handbooks, companions and guides that have proliferated in academia generally (on this phenomenon, see Arnal and McCutcheon 2013: 31-56). These books are important, because they often form, or at least seek to form, the first impression many undergraduate and graduate students get of a topic, either in an introductory class or as a prolegomena to their own research. As such, these guides would hopefully represent a distillation of the state of the field. Unfortunately, a perusal of some of these volumes shows that the insights of the last several decades of scholarship are not reaching into these potentially influential scholarly niches. Indeed, several of these ‘go-to’ handbooks contain entries on sacrifice that are significantly dated and distorting.
The ‘sacrifice’ entry by Olson (2011) in Routledge’s Religious Studies: The Key Concepts, for example, repeats old assumptions about the centrality of violence and death in sacrifice (the long shadow of Burkert and Girard), as well as privileging notions of self-sacrifice and substitution, which play a key role in Christian interpretations. For example Olson argues that ‘Sacrifice is a personal act in the sense that a sacrifice gives his
Robbins’s (1998) entry on sacrifice in Taylor’s influential Critical Terms for Religious Studies repeats this same overemphasis on death and reflects a lack of theorization of ritual in general. The main example of sacrifice around which the entire essay is shaped is the story of the binding of Isaac in Genesis 22, which Robbins asserts is ‘the Hebrew Bible’s exemplary story of sacrifice’ (1998: 291). The notion that the akedah, or binding of Isaac story, is exemplary of normative sacrifice shows how deeply Robbins’s views of sacrifice are affected by a particular strand of interpretation, largely rooted in Christian theology, which focuses on sacrifice as substitution. The idea that any one narrative could encapsulate the essence of any ritual is itself a fallacy, but beyond this, the akedah is a terrible candidate to serve as an example. This story is not at all representative of the sacrificial rules and procedures in the Hebrew Bible, much less of the ancient Mediterranean generally. There is an enormous amount of scholarly literature on the akedah and interpretations vary widely (see Delaney 1998 and Halbertal 2012), but however it is interpreted, the story in Genesis 22 is certainly not an example of typical sacrifice.
A positive example of a handbook doing a good job is the balanced and informative entry by McClymond (2011b) in The Blackwell Companion to Religion and Violence, which—correctly and ironically—refuses to equate sacrifice with violence.
Finally, a sophisticated if slightly dated discussion of sacrifice has entered the mainstream in the work of Michal Pollan. The famed food writer, best known for The Omnivore’s Dilemma (2006), includes a responsible and fairly scholarly discussion of sacrifice in his latest book Cooked (2013). In the chapter on barbecue, Pollan explores the evolutionary history of cooking meat and killing animals to obtain meat. He points out the omnipresence of ritualization surrounding meat eating in ancient and modern times. Although his discussion overstresses the importance of anxiety over the death of the animal, it mainly focuses on the ways sacrifice indexed communities and hierarchies, based on the work of Detienne and Vernant (1979). While perhaps not all that could be hoped, Cooked at least brings some balance to a broad range of non-specialist readers who might see sacrifice only as a backward and barbaric practice, rather than a complex, contested and endlessly illuminating nexus of discourse and ritualization.
Footnotes
Acknowledgements
I would like to thank the editors of Currents, and particularly Jordan D. Rosenblum for inviting me to write this essay. I hope some will find it useful in approaching the sprawling subject of sacrifice. With such a large subject I am sure many important works have been left out, for which I apologize. Hopefully what is here is a useful prelude. I am also thankful to Nathaniel DesRosiers, Kevin McGinnis, Jessica Pesce and Stanley Stowers, who read and commented on various parts of the project and manuscript, and to Philippa Townsend and Stephanie Kraft who bravely read the whole thing.
