Cosmologies of immanence and transcendence: Sahlins and the comparative anthropology of enchanted politics

Abstract

Marshall Sahlins’ The New Science of the Enchanted Universe proposes a new description of cosmologies which Sahlins describes as “immanentist.” Sahlins offers both a novel account of immanent cosmologies and a critique of the existing theoretical presumptions of anthropology, derived from the logic of “transcendentalism.” This article constructs an exegesis of Sahlins’ major arguments and a critical evaluation of the concepts Sahlins creates. This includes questioning the ontological status of “metapersons” in different cosmologies and the universal presence of hierarchy. Ultimately, this article challenges both Sahlins’ account of anthropological comparison and the symbolic reality of immanent cosmologies, which, it argues, remain marked by ongoing transcendentalism. By critiquing the effects of this transcendentalism, this article opens a theoretical experiment: how far can anthropology take its commitment to understand immanence on its own terms? How might this transform what it means to practice anthropological theory and perform anthropological comparison?

Introduction

The world has a final gift from Marshall Sahlins. If gifts contain part of the donor, this can be seen in the performative force of Sahlins’ text; The New Science of the Enchanted Universe enacts its own enchantment. There is a double enchantment in Sahlins’ book: the reader is enchanted by the worlds Sahlins conjures, but they are also enchanted by the allure of anthropology itself, of which Sahlins is perhaps the best representative.

Sahlins (2022: 137) offers a “Copernican revolution” that refuses to treat cosmology as something that revolves around human life. This revolution aims to decenter transcendentalism, the hegemonic ontological distribution of monotheistic and naturalist worlds, and the continuing analytical regime of the human sciences. ¹ Christian transcendentalism created a rift between God and the world, leaving humans autonomous in a more or less stable Nature; in anthropology, the rifts of transcendentalism separate people from cosmology, reduced to (illusory) belief, collective representations or ideology, and Nature, rendered disenchanted and inert. For Sahlins, this is apparent in an ongoing functionalism that, from Durkheim on, sets about explaining phenomena by indexing them as functions of society. Religion, then, as the projection of social ideals (Durkheim) or the mystification of social ills (Marx). These analyses, Sahlins (2022: 3) argues, are produced by the unchecked presumptions of transcendentalism. It is worth noting, about a century after Durkheim, how pervasive versions of this still are; seen, for example, in Ortner's (1978) reading of ritual as a therapy for social problems; Mocko's (2016: 171) reading of rituals of Nepali kingship as enacting discursive “arguments about the social position of the king” in a paradigm of social (re)production; or the existential functionalism that reduces ghosts to grieving mechanisms that “make sense” of death (Poletti, 2016: 277). In all these cases, cosmologies, since they cannot be true, are reduced to something more real: social cohesion, power, the human condition, and so on. Such approaches, a cocktail of “transcendentalist equivocation and colonialist condescension,” fail from the beginning by misunderstanding their object of analysis and, in doing so, “[…] reduce the meaningful relations of a culture of immanence to the status of convenient fantasies of the objective reality—of a world actually without such gods—thus making their culture a fictional representation of ours” (Sahlins, 2022: 11).

Sahlins’ decentering of transcendentalism occurs through his orientation to immanentism, which, he argues, is the ontological reality of “most of humanity.” In ontologies of immanence, ontologies without the rifts imposed by transcendentalism, humans are not the center of the world and live in what Sahlins’ calls a “cosmic polity”:

In a state of immanence, human society is a fractional part of a much larger cosmic society populated by a multiplicity of different kinds of persons, many organised in their own societies, and all more or less ordered, empowered and encompassed by the greatest of them, deities of this cosmic polity—this cosmopoliteia. From all around the world come anthropological calls to go beyond the perspective of society by and for itself, beyond the usual anthropocentricism, and instead to see the human order as integrated in the universal order. (2022: 124–125)

Sahlins (2022: 55) points to the sheer number of ghosts, gods, ancestors, spirits, etc. in the ethnographic register and the diversity of their relations in marriage, play, kinship, and so on. Following what might be called Hocart's axiom, Sahlins asks: “How can we make any progress in the understanding of cultures, ancient or modern, if we persist in dividing what people join, and in joining what they keep apart?” (Hocart, 1970: 23, quoted in Sahlins, 2017a: 23; 2022: 40). It is the cosmo-logic and onto-logic of these relations, taken on their own immanentist terms, that Sahlins develops throughout his text (2022: 11).

In this gesture, Sahlins continues to open new lines of thought and trajectories of analysis. How should one respond to these openings? If a book is a kind of gift, as I suggest, what does this mean for its readers? Can we make a response not given through mechanisms of reciprocity or debt but with a spirit of excess (Bataille, 1989; Derrida, 1992), of seeing Sahlins as inspiration and trying to give more? Sahlins’ work generates concepts that anthropologists should understand and use, but his work also promises further experimentation that can be achieved through a generative critique. This article, thus, offers an assessment of the analytical tools Sahlins provides and asks how anthropological theory might develop new approaches to conceptualize cosmologies of immanence.

Comparative anthropology, generalization, experimentation

Sahlins’ theoretical aim, a Copernican revolution in anthropological theory, is matched by his project's comparative scope, manifest in sources from Inuit North America, Amazonia, Africa, and the Sumerian states. Sahlins’ work can, here, be located in the discipline's ongoing tension between comparison and description, where, in recent years, the particularism of the suture of anthropology to ethnography, dominant since the 80s, has lost its hold. As Matei Candea (2019: 1) writes, “[…] anthropological comparison is back in the limelight and it is the ‘crisis of representation’ itself which is beginning to feel thoroughly passé”.

The anthropology of most of humanity is grounded in the methodology that Sahlins, following Leach, calls generalization. For Sahlins (2022: 14), generalization marks his anthropological concept creation as a kind of “guesswork”. This guesswork is buttressed by the division, drawn from Alan Strathern (2019), between transcendentalism and immanentism. For now, I am focused on this opposition's methodological function. The opposition forms the lever through which Sahlins diagnoses the failures of previous anthropology (which erroneously applied transcendentalist ontological assumptions to immanentist societies), clearing the ground for new guesswork. Here, through a comparative double movement, Sahlins simultaneously uses two methodologies he has used separately throughout his career:

De-Christianizing anthropological concepts through a genealogical account of their origins and through making apparent their failures to analyze ethnographic realties; seen clearly in his article “The Sadness of Sweetness” (Sahlins, 1996) ² ; and

Generalizing comparisons, based on the breadth of human realties in the ethnographic register, to identify broad patterns and generate concepts to account for these phenomena; seen most clearly in his work on kinship (Sahlins, 2013).

These methodologies co-exist in this work via the historically motivated opposition between transcendentalism and immanentism, which aims to make generality possible across cultures of immanence. Sahlins is essentially mobilizing two different modes of comparison at once, one based on opposition and one based on resemblance.

Sahlins’ opposition, or “frontal comparison” (Candea, 2019), between transcendentalism and immanentism is akin to what Robbins (2002) called “the critical use of difference” and functions to relativize anthropology's transcendentalist assumptions. Here, Sahlins’ project shares the ontological turn's use of difference to destabilize anthropology's ontological assumptions and its commitment to analyze cosmologies in their own terms. In this way, Sahlins’ opposition is operationally similar to Strathern's (1988) gift/commodity opposition or Viveiros de Castro's (2012) multi-culturalism/multi-naturalism opposition. Sahlins’ overall project, however, differs starkly from the ontological turn's codification by Holbraad and Pedersen (2017). Sahlins’ generalizing project could not be further from the collapse of scales in what they call “intense abstraction” (Holbraad and Pedersen, 2009). Sahlins, rather than using singular cases to produce ethnographically contingent concepts (Holbraad, 2017), uses ethnographic material to reform anthropological assumptions to produce encompassing generalizations.

From this attempt to displace transcendentalist baggage, Sahlins performs a “lateral comparison” (Candea, 2019) across an array of materials, identifying a uniform set of regularities as cosmic polities. In proposing a generalizing method based on resemblance, Sahlins’ approach departs not only from the ontological turn but also from the methodological thrust of structuralism's own lateral comparisons. While Sahlins and Lévi-Strauss both aim to go beyond cultural specificity, for Lévi-Strauss universality is revealed through difference, achieved through his account of variation; universality is not about resemblance but the integration of variation in transformational sets (see Descola, 2016). Contrastingly, for Sahlins, universality is reached through similarity—Sahlins wants to show not the transformation between cultures but their identity.

The major problem with Sahlins’ account, here, concerns its treatment of difference. Sahlins generates new concepts as generalities, subordinating difference “[…] to the identity of the concept” (Deleuze, 2014: xii). The very aim of generality, in which specific instances partake, inscribes the analysis within a transcendentalist frame. In this context, to rephrase Ingold (2016), continuing “recursion” (Holbraad, 2012) is how one avoids being a transcendentalist abroad in a museum of immanentism. That a given ethnographic situation might be understood as a cosmic polity must be the beginning rather than the end of analysis—any case will transform what we understand as a cosmic polity (see Holbraad, 2012: 264). My aim is, thus, to stretch Sahlins’ account further, creating new recursions, to critique ongoing transcendentalist features of his account.

Axial age, great divide, and transcendence and immanence

The comparative generalizations of Sahlins’ work are structured by the division between immanentism and transcendentalism. On the very first page, Karl Jasper's axial age is invoked. Sahlins writes:

The essential change was the translation of divinity from an immanent presence in human activity to a transcendental “other world” of its own reality, leaving the earth alone to humans, now free to create their own institutions by their own means and lights. (2022: 2, original emphasis)

For Jaspers, the axial age named a transformation away from mythic thought in Greek, Indian, and Chinese breakthroughs that oriented the existential and historical being of man. In Sahlins’ work, the transcendentalism/immanentism split is less a claim about the trajectory of history as an initial opposition, an opening to produce new concepts. Transcendentalism and immanentism can be conceptualized as eternally present modes in constant synchronic tension (as ideal types), just as much as they reflect a diachronic transformation. Reading transcendentalism and immanentism as competing tendencies finds some support in Sahlins’ text; as he describes, even after the so-called break brought about by transcendentalism, “[…] a recurring process fits better with the persistence of immanent elements in all such transcendental regimes” (Sahlins, 2022: 4). Alan Strathern (2019: 82) also notes that features he would describe as transcendentalist appear in ancient Egypt, before the axial age.

Distinguishing his account from Alan Strathern's, Sahlins describes a second axial break. This second axial age begins in Europe around the fifteenth century (Sahlins, 2022: 5). This disenchantment, closer to Weber than Jaspers, moves to more fully separate supernature from the world in which humans live, evacuating the natural world of sentience and agency, and creating a view of humans as autonomous centers of the world. Taking place up until the Enlightenment, this second axial break recalls Descola's (2013: 66) genealogy of the great divide:

In Greek thought and particularly in Aristotle's, humans remain a part of nature. Their destiny is not dissociated from an eternal cosmos, and it is by virtue of the fact that they are able to accede to knowledge of the laws that govern it that they are able to find their place in it. So, for the nature of the Moderns to come into being, a second operation of purification was necessary: humans had to become external to nature and superior to it. Christianity was responsible for this second upheaval, with its twofold idea of man's transcendence and a universe created from nothingness by God's will.

Once God is made to transcend the material world, the material world can be detached and studied in accordance with a set of laws (Latour, 1993: 32–33; Sahlins, 2022: 37); here, the optical illusion of humanity's transcendence gave them the opportunity to study it from outside in what Haraway (1988) calls the “god trick”.

This second break gives Sahlins the opportunity to distinguish, for example, Buddhist cosmopolitics from that of Christianity and the development of naturalism. Indeed, the cosmopolitics of the “transcendentalist,” Buddhist galactic polity (Tambiah, 2013) bears more in common with the pre-axial Incan empire's cosmological pyramid and its overcoding (Clastres, 2010: 146) or the onto-logic of African kingship (Sahlins, 2017c: 347) than it does with Christian empire or European colonialism.

A major problem with Sahlins’ immanentism is its treatment as the default cosmo-logic of human engagement with the world. Following his reading of Descola's Beyond Nature and Culture, Sahlins (2014) has argued for a default cognitive anthropomorphism as the projection of human-like states on the world. This is unsatisfying because this kind of cognitive claim rests on a “natural” cognitive orientation, violating Sahlins’ methodological aim. (This also, I think, risks treating animism as closer to a default state and thus more natural.) This kind of claim—made through an appeal to the cognitive sciences—is also present in the work of Alan Strathern (2019: 5, 7, 29), where immanentism is the ground out of which the figure of transcendentalism is achieved, through the labor it takes to make distinctions between immanent and transcendent, this-worldly and other-worldly, human and non-human.

The transcendentalism/immanentism distinction risks reading heterogeneous cosmologies backwards as the negation of the world we know. There is nothing itself wrong with this opposition, but it should only be a moment of argumentation; the heuristic opposition should destabilize the taken-for-granted categories of transcendentalism to allow a positive description of a plurality of forms that differ from each other as much as from the current ontological hegemon. Here, treating these as two contemporaneous (if virtual) tendencies is useful. This allows cosmologies not to be read from the point of view of historical development. In this context, a position like Pierre Clastres’ vis-à-vis the State is useful. Clastres argues Amazonian social forms are not a default out of which the State is made (1989: 199 et passim), rather these social forms make consistent efforts to prevent the State from emerging; they are societies against the State (Clastres, 1989). Could it not be the case that there is as much labor in preventing the emergence of transcendence as there is in codifying it in axial age religion? Just as transcendentalism requires recurring processes, might not immanentism too be figured as a recurring process of refusing transcendentalism? Not just ontologies without transcendence, but ontologies against transcendence. This moves “beyond negation,” in Marilyn Strathern's terms (1988), passing from a lack of transcendence to the affirmation of immanence, and beyond the figure/ground relationship that treats immanence as the default out of which transcendence emerges.

Finitude and the outside

The ontological ground of cosmologies of immanence is figured by Sahlins through the finitude of the human situation:

People did not create the gods out of thin air; they only hypostatized the forces that were already conditions of their existence. I repeat: humans did not imagine the gods, they only objectified, or more precisely, subjectivized, the extra-human forces by which they themselves lived and died. The forces were already there. They were not imagined. They were real, empirical, life-giving and death-dealing forces. People only gave them substantive qualities that made them negotiable, or at least intelligible—consciousness, understanding, volition, and intention. They hypostatized these existential forces as persons, hidden souls with whom their own hidden powers or souls might communicate, but who, as the instantiation of these forces were metapersons. (2022: 174–175, original emphasis)

If Durkheim was reversed in order to ground human social life in cosmic relations with a plurality of beings, these cosmic relations are themselves given grounding as the anthropomorphization of existential forces. From this model of cosmology, two problems arise: (1) the status of finitude as an ontological universal and existential ground of cosmological construction; and (2) the status of hypostatization in the dialectic of finite humans and outside forces.

The idea that cosmology is a response to an oncoming death or forces outside one's control is a cliché of interpretation that Sahlins comes close to endorsing. The idea that “belief” in ghosts is (merely) a way of grieving and addressing the contingency of death, for example, is what I call, following Pedersen (2020: 262), “existential functionalism,” where cosmology is a reflex not of social cohesion but an ostensibly universal human condition. Another version of such an analysis, critiqued by Sahlins, is the less existential and more psychological accounts of “belief” in Hume and Malinowski. Sahlins (2022: 16–19) criticizes Hume and Malinowski's accounts of magic as that which one has recourse to in the face of contingency, but ends up with something similar but more general. Magic (or cosmology), for Sahlins, is not only a precursor to actions which are especially risky, but is so tightly interwoven into regular or mundane acts that cosmology cannot only be a psychological reflex in the face of risk or danger. This would suggest, Sahlins argues, “There is some kind of greater human dependency” (2022: 19). In a way, Sahlins merely generalizes the account in a more encompassing description of human finitude and dependence (2022: 27). There is a key ambiguity here that persists throughout Sahlins’ text centered on the question of hypostatization and the reality of metapersons. Are humans vulnerable to outside forces, which they hypostatize as metapersons, or are they dependent on actually existing metapersons, from whom these forces issue and with whose aid they may be ameliorated?

In the spirit of Sahlins’ methodological detranscendentalization, I want to suggest that, rather than a universal ground, finitude may itself be the secularization of a Christian concept or a product of monotheism, more generally. Alain Badiou, in a rather different conceptual project, has argued that the restriction of being to finitude occurs in the context of a monotheistic capture of the infinite to the domain of transcendence and a barring of the realization of an immanent infinity (Badiou, 2022: 75). What Badiou (2022: 76) calls “religious oppression” consists in condemning the pride that would allow for a subjective encounter with infinity in its immanence (in which beings would aspire “to be equal to God”) and the celebration of the virtuous “internalizing of finitude as destiny”. Jadran Mimica, reviewing James Weiner's use of Heidegger's poetic ontology in the ethnography of the Foi (Papua New Guinea), challenges the ground of phenomenological construction in human finitude. Like Badiou, Mimica notes the emphasis on finitude is a product of the break or rift imposed by cosmologies of transcendence. Mimica's comments on Weiner's Heideggerianism can be transposed onto the themes of dependence and finitude that are seen as motivating forces for the construction of cosmology in Sahlins’ text:

The mood of man's perplexity and the overwhelming sense of the world—“dependent men” like the “tragic” or “guilty man,” the man “overcome by nature,” […]—all these are the vintage chords of the Western existential blues, with their roots in the Judaeo-Christian tradition of the human and cosmic fall. (Mimica 1993: 84)

The motifs, especially dependence, that recur in Sahlins’ text bear traces of an ongoing transcendentalist reading of cosmology. While Sahlins does not dwell on death, it is central to the notion of finitude. Indeed, in immanent cosmologies, death does not mark an absolute ending but a transformation: “The Foi, I believe, are not mortals in an unqualified universal sense, since it does not follow that ‘death experienced as death’ invariably makes mortals mortal in all life-worlds” (Mimica, 1993: 91). If the problem of the afterlife, as in Christianity, or of nothingness, following the death of God, does not occur as such in cosmologies of immanence, mortality has a different ontological status. ³ Cultures of immanence frequently display what Mimica (1991: 40; 2020; 2023: 138) calls an ouroboric cosmology, an image used to note the continuity of the cosmos. In the movement of beings through the immanent surface of an ouroboros (or what Handelman and Shulman, 2004: 25, in their work on Śaivism, describe as a Möbius cosmos), death is one moment of ongoing transformation (Sahlins, 2022: 48; Mimica, 2023: 138), and cannot be the site through which humans are rendered finite juxtaposed to a separate infinity. Immanence refuses to distinguish the finite and the infinite; finitude only becomes finitude when infinity is captured in regimes of transcendence.

The key movement in Sahlins’ account of cosmology that embeds finite humans within a cosmic polity is hypostatization, a concept used by Sahlins to describe deities as the cultural reification or personification of material forces. Hypostatization is the movement through which cosmic forces are given cultural identity through the figure of divinity. In the moments of the text which discuss hypostatization (distinct from moments of emic exegesis elsewhere in the text), Sahlins sets his approach as a symmetrical reversal of Durkheim in which he follows Giambattista Vico. For Durkheim, cosmology is “an external idealization of an internal force” (Sahlins, 2022: 22), whereas, for Vico, it is “an internal hypostatization of an external force” (Sahlins, 2022: 22–23). Here, as above, cosmology concerns the unknowable but necessary forces obscured by human finitude, which people render as gods. The problem with the reversal of Durkheim is that deities are still at root representations; the difference is a self-portrait designed to affirm collective identity (Durkheim) versus an anthropomorphic depiction of external forces (Vico). Whether cosmology is a projection of society outside or a domestication of outside forces in society, the derealizing effect remains the same.

Despite arguing, in a reversal of Marx, for “determination by the religious basis” (Sahlins, 2022: 3), when treating deities as hypostatizations of external forces, Sahlins even appears as a “cultural materialist,” insofar as he locates “[…] the source of the gods and immanence in material forces that shape the human condition” (Ewing, 2022: 936). This account of hypostatization—a kind of psychological or existential functionalism—is at tension with other aspects of Sahlins’ account of cosmology. An alternative way of describing the relationship of deities to forces comes in the exegetical moments of the text, when Sahlins’ is describing the internal cosmo-logics of emanation. In these moments, in the logic of the cosmologies at hand, deities appear as the cause of both social life and forces outside it. They are the ground of the enchanted cosmos (not hypostatizations of it) (Sahlins, 2022: 28). As the ground of being, gods manifest immanently as distributed persons (a concept taken from Gell, 1998). Sahlins writes: “Partible beings, the gods are present for all that they are distant; they are potent agents of humans’ fate for all that they are unseen” (2022: 69). Here, the account of gods as hypostatizations of forces clashes with Sahlins’ notion of the distribution of divinity. Through hypostatization, the primary reality of cosmic forces is (by the work of culture) rendered as deities; whereas, through distribution, forces and culture derive from the ontological power of the deity.

Sahlins’ book reflects an antinomy between a descriptive concept of immanence and a diagnostic interpretation which ties this immanence to a broader topology of human finitude. These competing accounts display an ongoing tension between transcendentalism and immanentism. The immanentist reading, which does not treat deities as culturally specific hypostatizations of natural forces, provides a way of describing the pervasiveness of cosmology in everyday life without an existential or psychological functionalism. It also does not rest on the logic of human finitude. On this reading, cosmology names an immanent causality where “[…] the spirits enact the actor's actions” (2022: 25); that is, cosmic agency does not cover over contingency but is expressed in both contingency and successful human action. Here, pace Malinowski, magic does not supplement human action to (psychologically) make success (feel) more likely, rather successful action embodies magic as its immanent cause.

Is the enchanted cosmos symbolic?

The tension between Sahlins' two accounts of divinity is a manifestation of a lingering problem in his work regarding the status of the symbol (and thus, implicitly, the relationship between symbolic reality and material reality). Sahlins’ discussion of language's power is telling; language, for Sahlins, is powerful for two reasons:

First, because the arbitrary relation between the signifier and its worldly referent establishes the spoken word as an autonomous, creative force; and second, because speech has potent pragmatic effects on interlocutors and their situation independent of any physical contact. (2022: 46–47)

These are two rather distinct claims, and they do not necessarily pair nicely. In fact, to give an appropriate cosmopolitical reading of the pragmatics of speech in enchanted worlds, the symbolic logic of arbitrary signifiers and the autonomy of language must be disrupted. The logic of the signifier as arbitrary distinctions, superimposed on nature, in structural and symbolic anthropology rests on a nature/culture dualism; the signifier is arbitrary because nature is not meaningful in-itself and becomes so only through the human production of symbolic constructs. Here, the autonomy of the signifier and the separation of the symbolic order are transcendentalist in the most extreme way possible.

Accounts which make signification arbitrary are incredibly recent in the historical scope of Sahlins’ account of axial ages. As Talal Asad (1993: 60) notes, the logic of symbolic meaning emerges in the history of Christianity and is secularized in anthropological practice. In the same way the nature/supernature distinction emerges at a distinct point of rupture, as Sahlins (2022: 36–37) points out, the language/world distinction also emerges at a point of rupture. Just as it is only through the newfound opposition between nature and supernature that supernature is figured as transcendent, only in its opposition to the world is language found to be arbitrary and autonomous. Foucault, for example, notes the power of language before the Classical Age (seventeenth century); here, the signifier is precisely not arbitrary because it connects to what it names as the “prose of the world.” Foucault writes: “In its raw, historical sixteenth-century being, language is not an arbitrary system […]” (2002: 38). Language was figured, in the sixteenth century, as the medium through which the world and thought could be understood through a regime of resemblance, recomposing God's original language (Foucault, 2002: 40). It was the seventeenth century's invention of a binary model of signification that dissolved “the profound kinship of language with the world […]” (Foucault, 2002: 47).

Marisol de la Cadena, in the Andes, citing Foucault, describes willakuy as “[…] the act of telling or narrating an event that happened […]” in a way that makes this happening present through a performative entwinement with what it names (2015: 28). Here: “[…] no separation exists between the narrative (the word) and the event (the thing). Or better said: in willakuy, there is no word and thing mediated through meaning. A willakuy performs the event” (2015: 29). So too, Sanskrit mantras possess direct efficacy because Sanskrit is the expressive language of primordial being and directly instantiates what it names. Vivienne Kondos writes:

Sanskrit is no ordinary language, but is said to be the original one from which all others evolved. Nor is it just a language but those vibrations which burst forth at the creation, the manifestation of the Absolute in that form (sabda). (2019: 282)

Sanskrit does not symbolize an absent deity, nor do deities personify the power of Sanskrit. Rather, Sanskrit mantras, as direct expressions and manifestations of divinity, are powerful not because they are autonomous but because they resonate with the cosmos which they express: “[…] if the language itself contains the primordial sounds, then they are also cosmic forces” (Kondos, 2019: 283). Mantras possess efficacy whether or not the subject of their enunciation understands them (Michaels, 2016: 256); they are definitively not symbolic signs to be interpreted. ⁴

Indeed, the ethnographic record is replete with refusals to perform exegesis, insistences on the belonging together of language and world, and accounts of pragmatics in which signs are not symbols and which produce effects outside the arbitrary logic of signification. Gell's fieldwork in Papua New Guinea with the Umeda exemplifies this point:

Among my Umeda informants I found none willing to discuss the meaning of their symbols—to discuss their symbols as symbols “standing for” some other thing or idea, rather than as concrete things-in-themselves. In fact I found it impossible to even pose the question of meaning in Umeda, since I could not discover any corresponding Umeda word for English “mean,” “stand for,” etc. Questions about symbols were taken by Umedas as questions about the identity rather than the meaning of a symbol: “what is it?” not “what does it mean?” (Gell, 1975: 211, quoted in Asad, 1993: 61, n7)

Here, the methodological warrant of symbolic anthropology falls short. If, following Hocart's axiom, “we” cannot separate what “they” join, there can be no autonomy of the signifier.

For Sahlins, symbolic language is cosmogonic: “[…] speech is the symbolic, life-giving power of the creation of cultural order. In which case, humans are not only spirits, but the original spirit, the genesis of spirit” (Sahlins, 2022: 47). If speech, for Sahlins, is cosmogonic, it is through the creation of culture. Combined with the model of hypostatization, symbolic speech is simultaneously treated as the mechanism for attributing agency to the forces of the cosmos and the misrecognized origin of divine power:

Humans are here considered as limited beings, their own existence dependent upon and subordinate to a cosmic host of metahuman powers. But in this enchanted mode of unhappy consciousness, humans’ symbolic power is the model of divine power itself. Divinity or spirit is the hypostatization of speech, of symbolic power. “In the beginning was the Word, and the Word was with God, and the Word was God” (John 1:1). Human speech is the essence of divine power. (Sahlins 2022: 48)

This line of thought, Sahlins claims, is not anthropocentric because it is not a projection à la Durkheim; however, it is just as derealizing. Leaving aside the reference to Hegel's unhappy consciousness, grounding cosmology in human action appears as the transcendentalism Sahlins’ text promised to avoid. To follow the example of mantras, used above, Sahlins’ argument would be better in reverse. Divine power is the essence of human speech since mantras produce pragmatic effects through their participation in cosmic power. Why in a text that proposes a revolution of anthropological thought, that proposes to stop grounding cosmology in the social, does Sahlins regress on this point? Why do ethnocentric (transcendentalist) concepts like “belief” or “myth” get removed (Sahlins, 2022: 13–14), but the symbolic remains untouched and newly celebrated? I want to suggest the symbolic is kept in place because it grounds the endeavor of generalization. Generalization is grounded in commonality of resemblances because all cultures are symbolic (and this is a universal human faculty). It is this methodological commitment to the symbolic that prevents an account of pragmatics. Here, Sahlins’ Copernican revolution does not go far enough. If one follows the immanence of the cosmos to its end, pragmatics cannot rely on a separation of domains. If one embeds speech in a pragmatics of cosmic participation, then the immanent casualty at work in a willakuy, or uttering a mantra, can be given efficacy internal to their own cosmo-logics.

Hierarchy: encompassment and distribution

A major argument Sahlins advances, from the perspective of the cosmic polity, is that, taken on their own terms, all cosmologies are hierarchical. Hierarchy is parsed in two ways: (1) as an onto-logic internal to cosmologies, where encompassment describes the relationship of a singular deity to their plural manifestations; and (2) as a universal cognitive operation, embedding entities in encompassing categories, rendering hierarchy implicit in the act of perceiving the world through the symbolic logic of token-type.

Hierarchy is given, for Sahlins, through the one and the many. Sahlins develops the cosmological motif “Divinity is one” (Lienhardt, 1961: 156, quoted in Sahlins, 2022: 105) through several cases culminating with explication through the Dinka:

Like East African Dinkas’ Nhialic (“Divinity”), gods are one in essence and multiple in forms. Not simply the one and the many, but the One over and in the Many. Dinka Nhialic is a partible or distributed person: in being—and/or in power, to put a fine point on it—present in all the “spirits,” great and small. (Sahlins, 2022: 109)

For Sahlins, the cosmos manifests as the partible person of the deity who, in their distribution, maintains an encompassing control over their distributed processes. This unity, described by Sahlins, resonates with many cosmologies in the ethnographic record; think of the unity of kwoth, reported by Evans-Pritchard (1977); Hindu accounts of cosmic unity (brahman) underlying diversity; or the Iqwaye cosmos, made out of the anthropomorphic deity, Omalcye (Mimica, 1988: 74–87, 1991: 40–43; Scott, 2007: 11). ⁵ The morphology of a deity becoming a distributed person recurs in myths of cosmogenesis and produces a distributed oneness; however, in many such cases, the immanence of this distribution means the one cannot encompass the many.

Beyond the specificity of cosmo-logics, Sahlins moves from a descriptive account of the one and the many to generality grounded in “class logic.” For Sahlins, this logic manifests as token-type relationships—originally developed in Islands of History (Sahlins, 1985)—which he describes with the double movement of encompassment (one) and distribution (many) (Sahlins, 2022: 85). Here, the description of cosmo-logics through encompassment manifests as the cognitive operation of class logic:

In sum, as a categorical notion, “class” is a hierarchical order of things, a symbolic topology of token and type, which when hypostatized in a world of immanent person-beings generates structures of authority and power far superior to humans, and to which they must submit—and endeavor to appropriate. (Sahlins, 2022: 132–133)

Token-type is, then, read through Louis Dumont's (1970) hierarchical encompassment as a general condition through the syllogism: all humans perceive the world through categories, all categories are hierarchical, therefore humans are creatures of hierarchy. As Sahlins puts it: “Homo Hierarchicus: hierarchy is built into the human way of thinking the world; it is ontological, insofar as the elements of the world are classed or categorized” (2022: 131).

As with Sahlins’ account of hypostatization and symbolic power, an account of cosmo-logics in their own terms is distorted by an attempt to ground this in an ostensibly universal cognitive and symbolic structure, (re)introducing a transcendentalist bent to the analysis. The over-generalizing and emptying of the concept of hierarchy reduces hierarchy to the structuring processes of cognition and logic. When conceived of as a class in which things participate, encompassment is given the ontological characteristics of transcendentalism in which the encompassing category is singular and the encompassed manifestations are plural. I want to suggest the very logic of categorization and classes misses cosmo-logics of causality. When reading cosmologies as causal logics, immanentist cosmologies frequently make metaphysical operations that confound the morphological relationship of the one over the many; here, the logic of a multiplicity contained by an encompassing force (be it deity or symbol) fails to take the logic of immanence far enough.

Starting on the “home ground” of hierarchical encompassment, South Asia, one sees both cosmologies which celebrate the unity of being and cosmo-logics which challenge Sahlins’ account of hierarchy. The diversity of Hindu cosmologies makes generalization difficult. In more axial age-inflected versions of Hinduism, centered on the Upanishads and Vedanta's reading of them, the one is opposed to the many; the unity of being (brahman) is opposed to the multiplicity of its illusory fragmentation (maya) (Dyczkowski, 1987: 37–38). Note, here, that precisely the versions of Hinduism where the one and the many are hierarchically valued are the versions most fitting the label of transcendentalism. ⁶

In her work on Nepali Hinduism, Kondos points out the conceptual failure of Dumont's encompassment; rather than an encompassing category, unity is a subtractive point of inexistence, something reached by discarding the illusory attachments of plural being (Kondos, 2019: 173–174). Divinity is one but not as an encompassing metaperson or symbol. Here, the One, rather than a circle encompassing multiplicity, is best imagined as the center point of a mandala (bindu), beyond the concentric expansion of space and time (Gray, 2006: 92). Unity, in this case, also cannot work as the type for the world's tokens, since it is noumenal and without quality (nirguna), unable to provide a model for the phenomena to copy.

Alan Strathern (2019: 94–95) describes Tantra as an immanentist subversion of transcendentalist logic; here, an altogether different account of the one and the many appears. Against metaphysics that subordinate plurality to unity, some versions of Hinduism embed plurality and unity within each other, such that the one is existent in the many and the many expresses the cosmos’ unity (Dyczkowski, 1987: 37, 41–42). For example, the development of Kashmir Śaivism is described by Dyczkowski (1987) as an integral monism. Here, as Deleuze and Guattari write, “MONISM = PLURALISM” (2013: 21); or, in Marriott's terms, there is “diversity without dualism” (1976). Tantric Hinduism and Buddhism develop cosmo-logics that announce the immanence of oneness to multiplicity and vice versa, in cosmic topologies that resemble a Möbius strip rather than an outside encompassing an inside.

Developing his own account of pre-axial age cosmology, Don Handelman (2014: 99) argues that the concept of encompassment derives from post-axial age cosmologies, where deities become absolute creators. Handelman argues this transcendentalist account of encompassment relies on a logic he calls inter-gration, where cosmic unity or coherence is established from an outside (the realm of a transcendent god); immanentist cosmologies, contrastingly, display what he calls intra-gration through which the cosmos composes itself from the inside without ontological rifts. In Handelman's discussion of Mahayana Buddhism in China's Hua-yen school, one sees the analytical failure of Sahlins' one-many hierarchy and the dynamics of cosmic intra-gration, holding together from the inside through the cosmos’ autopoiesis. Handelman writes:

The Hua-yen cosmos has no centre, or, if there is a centre, “it is everywhere. Man certainly is not the centre, nor is some god” (Cook, 1972: 4). Note that the Hua-yen cosmos has no external boundaries, unlike the absolute, virtually impassable boundary between God and human being to which the surviving monotheisms have accustomed us to as natural and commonsensical. The Hua-yen cosmos is not enclosed from outside itself, in contrast to our understanding of the kind of holism suggested by Dumont's idea of “encompassment.” (2014: 97)

The distribution of cosmic unity through all its points is described through the figure of Indra's net, a network of jewels each of which refracts the totality of jewels composing the net (Handelman, 2014: 97; Lidke, 2017: 30). Anthropologists inspired by holography—used to describe refractions, “each region [of which] contains a total structure ‘enfolded’ within it” (Bohm, 2002: 188)—have drawn out descriptions of the one-many in ways more suitable to cosmo-logics of intra-gration. This can be seen, for example, in the work of Wagner (2001) in Papua New Guinea, Martin Holbraad (2012: 168–169) in Ifá divination, Jeffery Lidke (2017) in Newari Tantra, or Handelman and Shulman (1997: 194–198) in Śaiva mythology.

Where holographic logic distributes oneness in the cosmos without encompassing it, Pierre Clastres reports a different challenge to the model of the one hierarchically encompassing the many. Clastres identifies the motif of the One with State societies (1989: 217), ⁷ and describes Amazonian metaphysics as cosmologies against the One (against the State) (2010: 276). This is a cosmopolitics of multiplicity without an inter-grating One: “Savages want the multiplication of the multiple” (Clastres, 2010: 274). (Note that, for Clastres, “savages” are described tautologically as those against the State.) For Clastres, this can be seen in the development of cosmologies without centralization or unification: “Almost always, as a matter of fact, the cultural practices of these people develop without implicit reference to a single or central figure of the divine […]” (2010: 122).

Clastres’ multiplicity, against Sahlins’ universalization of cosmic hierarchy, can be seen in the response of Americanists to Sahlins. Trying to generalize the topology of hierarchy, Sahlins (2022: 82, 85) argues that hierarchy is central to the putatively egalitarian societies of Amazonia because multiplicities of animal-beings are encompassed by their singular species-masters. Here, the claim that divinity is one is already watered down from the African cases from which it originally derived. Against Sahlins, Carlos Fausto (2022: 949) describes relations of mastery as a series of asymmetrical relations that are themselves multiplied and dispersed. For his part, Descola (2014: 297) describes relations of species-masters as specific relations of dominance, distinct from ontological hierarchy, which, in Amazonia, is a subordinate relational mode in cosmologies dominated by shifting and reversible relations of predation and gift. Fausto renders Amazonian cosmologies along Clastrean lines, describing the acts of metapersons as originally dispersed in a way that prevents a hierarchical distribution ordered beneath or around them: “The pyramidal model is absent, as is the notion of a ‘prime mover’—even the demiurges that made the world as it is are not One, nor are they Firsts” (Fausto, 2022: 949). Here, Fausto describes a topology of intra-gration that defies Sahlins’ encompassment:

Amazonian cosmic topology, in brief, is not hierarchical (“the rule of the high priest”), but heterarchical—which, it should be stressed, is not necessarily the same as being symmetrical or egalitarian. A heterarchy has also the potential to organize itself in various and dispersed asymmetries which are never totalized as an all-encompassing structure. (Fausto, 2022: 949)

Where, in Amazonia, the cosmos’ self-organization produces a heterarchy, Handelman (2014: 98) notes that other cosmologies of intra-gration are hierarchical, as seen in Hua-yen Buddhism and throughout Indic cosmologies. Following Handelman, the task of theory, in response to Sahlins, is to conceptualize hierarchy in a way that does not reproduce transcendentalism.

Are gods (like) kings?

For Sahlins, if hierarchy is the universal condition of all societies, so too is “something like” the state: “Something like the state, the cosmic state, is the general condition of humankind—even in the state of nature” (2022: 137). Acephalous societies, Sahlins (2022: 127–130) argues, are nonetheless embedded in a cosmic hierarchy. For Sahlins, the cosmic state as a universal form becomes apparent in the homologies between Sumerian and Inuit cosmologies. According to Sahlins (2022: 170), both are hierarchal; both rely on the agency of divinity to achieve the conditions of human life; and both have unity of divinity, which is both centralized and distributed through cosmic encompassment. Sahlins’ development of “the cosmic polity” derives from a formula he discovers in Hocart: “the straightforward equivalence, king = god” (Hocart 1970: 74, quoted in Sahlins, 2017a: 23). Does the “=” suggest this equation is reversible? That is, does god = king, or softening the formula, as Sahlins does (2017b: 162, n3), do gods possess kingly powers?

Following the critique of encompassment as the general condition of hierarchy in the previous section, I want to use Australian material to present an account of cosmopolitics where humans, metapersons, and various non-human persons negotiate their ontological conditions of existence where the rules are not enforced by metaperson sovereigns. Handelman's model of intra-gration might give a more clear reading of the status of ancestor metapersons; here, one sees a cosmic polity governed through its intra-gration. Dreamings is the term used for both ancestor metapersons and their trajectories across an already-there landscape, which they transform. Dreamings’ movements, encounters, and relations generate the Law, a concept used to translate various indigenous terms across Australia. The basis of the Law, rather than prescriptions generated from the outside, can be seen as intra-grated, generated internally through relationships. Thus while:

Law is often expressed as rules about behaviour […] what Law seems most fundamentally to be about is relationships. Dreamings determined sets of moral relationships—country to country, country to plant and animal species, people to country, people to species, people to people. (Bird-Rose, 2000: 56)

As holding together and held together as a bundle of moral relationships, the Law is akin to what Signe Howell (2017: 144) calls “cosmo-rules.” Cosmo-rules are not enforced by ancestor metapersons from the outside—Dreamings do not inter-grate the cosmos from the outside like sovereigns. As Howell (2017) puts the situation among the Chewong of Malaysia, contra Sahlins, there are “rules without rulers.” In Australian myth, Dreamings are themselves punished for breaching taboos (see, e.g., Glowczewski, 2015: 140–141), which would suggest cosmo-rules are not the jurisdiction of ancestor metapersons acting with “kingly” power, but something different since the rules that emerge in Dreaming are the lingering, virtual product of encounters that have enduring force as the cosmos hangs together. The rules of the Dreaming are relational; there can be rules without rulers when the rules are generated out of the relational constitution of an intra-grated world, held together through the mutual embedding in a positional and dynamic schema generated through the forces of Dreamings. The intra-gration of the cosmos means changes to parts affect the whole, producing feedback relations:

Each part of the cosmos has within it the potential to expand; each is potentially a runaway part, pushing and testing until it is stopped by others. But each part, as it becomes active triggers other parts which are interconnected to it and which stop the runaway process. (Bird-Rose, 2000: 97)

The Dreamings do not dictate Law from outside, their intra-actions generate Law embedded in the spatial dramatization of relations. Unlike the de jure universality of sovereign power, Law is, here, situated and emplaced in context-sensitive perspectives. The complexity and spatial distribution of Dreamings’ movements mean any principle of unification is thwarted from the beginning; because powers are emplaced, they cannot be held together all at once. Recalling the holography of Indra's net, “[…] the view from the inside is one of a multi-centred world in which each centre is structurally equivalent, and linked, to every other centre” (Bird-Rose, 2000: 220). The distribution of multiple centers without unification renders cosmic relations the negotiation of differential access to powers or forces. As Barbara Glowczewski (2015: 137) describes in Warlpiri cosmopolitics of the landscape, territory and ritual performance, Dreamings manifest as fundamentally multiple, “as energies that circulate among participants”. Persons, here, are not neatly encompassed by a singular totem that acts as their species-master, rather: “Every person carries within themselves a multitude of becomings, including totemic becomings: he or she is always a constellation of totems, not just a single totem” (Glowczewski, 2015: 136).

The Law concerns, then, feedback relationships between the present and the Dreaming, as Glowczewski (2021: 164) describes. The Law does not concern the sovereign exception to punish, but the dis-intra-grative effects of relational mismanagement, which violate the webs of intra-dependence established by the Dreamings. There is no outside from which Dreamings could rule (or of which they could be hypostatizations); as embedded in relations, which are emplaced on a landscape, they cannot encompass human sociality.

Elizabeth Povinelli provides the example of a creek-metaperson in northern Australia. Tjipel is a teenage girl who became an estuarine creek after losing a fight with an old man. The creek-place of Tjipel's being enfolds her relations and trajectories (Povinelli, 2016: 93). In her emplaced being, Tjipel is able to establish what Povinelli describes as normativity (2016: 94–95). Tjipel's human kin need to care for her through the norms established through her entanglements; here, normativity cannot be established from outside but is embedded in localizing intra-actions (Povinelli, 2016: 102–103). Failure to care for Tjipel, Povinelli reports, would lead to a withdrawal of her life-giving and sustaining relationality (2016: 94). Here, one sees the potential force of Sahlins’ science of the cosmopolitics of immanent cosmologies. Tjipel is an immanent metaperson with whom relations must be maintained to continue human flourishing in the world. But the idea that Tjipel manifests sovereignty would skew the cosmopolitical sensitivity of the arrangements that are established through the emplaced dynamics of intra-relationality. This kind of situated normativity, embedded in placed relations, precisely develops rules without rulers in which metapersons are not a model from which any kingly polity could establish itself.

Conclusion: immanence and holographic comparison

What makes a text a unique kind of gift is its ability to keep giving, to be read and reread, and to enter new compositions as its components are newly challenged and affirmed. Just as the emergence of transcendentalism required a “recurring process” (Sahlins, 2022: 4), the same is true of detranscendentalizing anthropology. The best way to read Sahlins’ work is as a gift we should take as a spur, continuing an ongoing process of detranscendentalization. Sahlins’ gift should be distributed without being encompassed. What I offer, here, is one of many ways to receive Sahlins’ gift. In my reading of The New Science of the Enchanted Universe, I have taken aim at specific instances of transcendentalism (finitude, deities as hypostatizations, the autonomy of the signifier, and so on); however, more general problems occur in the structure of Sahlins’ argument as a whole. Cosmologies of immanence, I have argued, cannot be conceptualized through hierarchical encompassment; following Handelman, this is a transcendentalist notion. Problems intensify if one notes that hierarchical encompassment is not only the schema Sahlins claims to be at work in all cosmologies, but is also the structuring principle of his project of generalized comparison. The cosmic polity is a singular and general type that hierarchically encompasses multiple and specific cosmologies, which are its tokens. If one were to follow the trajectories of Sahlins’ project at its immanentist best, it must be taken up again. What would an immanentist comparative anthropology look like? One starting point for an immanent comparativism can be potentially found in the holographic model I used to displace the image of encompassment. In this model, singular cosmologies are no longer encompassed by generality which they instantiate but are diffractions of univocity without a center. Such a holographic model is hinted at in Carlo Severi's suggestion of a comparative approach based on the composition of variations in the work of Anton Webern:

Once the theme has been presented together with its first variation, the second will be a variation not of the theme but of the first variation; and the third variation, for its part, will be a variation of the second, and so on. In this way, one obtains a structure that generates not only variations organized around a center but also other centers that can produce other variations. This model makes it possible to illustrate a way of organizing differences that does not recognize any preestablished center of epistemological priority. (Severi, 2015: 12, original emphasis)

Conceived holographically, each variation, itself a center, contains a metonymic refraction and internal redescription of the total composition of which it is a part. Despite the conceptual failure of his hierarchical encompassment, such a model can be seen in Dumont's development of Leibniz. The inventor of calculus’ holographic cosmology is used by Dumont to illustrate a model, where, like the jewels in Indra's net: “Each culture (or society) expresses the universal in its own way, as does each of Leibniz's monads; and,” Dumont argues,

it is not impossible to conceive of a procedure—complicated and laborious it is true—which would enable one to pass from one monad or culture to another through the intermediary of the universal taken as the sum integral of all known cultures, the monad-of-all-monads, present on everyone's horizon. (1986: 210, original emphasis)

Dumont's holographic image, I suggest, makes a nice virtual backdrop for comparison in the way Rane Willerslev (2011) has described. Comparative anthropology without the god trick, unique to the transcendentalist world, means setting Dumont's monadology in “perpetual motion” (Holbraad, 2012: 265). Immanence, here, can only be perceived in motion, just as a song can only be heard in the progression of its sonorous becoming. From this perspective, Sahlins’ revolution, then, does not reach a final equilibrium but, rather, is better conceived of as a continuous revolution, a recurring process.