Derrida,Stengers,Latour,and Subalternist Cosmopolitics

Introduction

The concept of cosmopolitics has emerged anew with vibrant senses in contemporary intellectual life. This essay takes cosmopolitics as a useful site for fleshing out and laboring to reconcile two ostensibly-contradictory political and epistemological programs: Jacques Derrida’s ethics of hospitality, on the one hand, and Isabelle Stengers’s and Bruno Latour’s efforts to democratize science, on the other. Most philosophers and science studies scholars would probably assume these authors’ respective senses of cosmopolitics to be incommensurable. Derrida’s (2001) ethics of hospitality openly engages Immanuel Kant’s (1991 [1795]) senses of ‘cosmopolitics’ and ‘cosmopolitanism’, while Stengers (2005: 994, 2010: 79–80) abruptly dismisses the relevance of Kant’s cosmopolitan vision. Likewise, Latour summarily writes off Derrida’s thought as unrepentant reductionism (Latour, 1993: 5–6) and stakes his basic philosophical program in clear opposition to Kant’s metaphysics (Latour, 1999b: 5–7). In the first book to read Latour’s work systematically as philosophy, Graham Harman (2009: 24) even qualifies a passing convergence between Derrida’s and Latour’s programs as ‘a normally unthinkable alliance’.

I approach the labor of this unthinkable alliance not as a philosopher concerned with metaphysical matters, but as an anthropologist working in the field of postcolonial science studies. As such, I’m invested in taking seriously both the radical realist empiricism of Latour’s science studies and the lessons of subaltern historiography that drew from antirealist critiques of representation including Derrida’s metaphysics (see also Hayden, 2005; Watson, 2011). Thus my reading of these two senses of cosmopolitics has a distinct subalternist edge. My insistence on retaining the term ‘subaltern’ marks a political, ethical, and scientific commitment that intellectuals should strive to engage and represent the otherwise-unthinkable or abject.

One doesn’t need the term ‘subaltern’ to be a good political or scientific spokesperson. While most anthropological research doesn’t draw from subalternism, cultural anthropologists engage with and represent otherwise-voiceless collectives and subjects on a daily basis. But such efforts to expose and oppose forms of material, symbolic, and epistemic violence clearly benefit by intensive engagement with conceptual work that considers the limits of representation. So I make the case that while an expanded concept of subalternity comes with intellectual risks, it could help scholars productively retool the science-politics interface and thus expand and improve emergent forms of coexistence, worlding, and hospitality. By critically reading Derrida’s project against Stengers’s and Latour’s, I outline my vision for a subalternist cosmopolitics, an ontological politics that attends to its limits of representation and the forms of violence that its boundary-practices enact. Conceiving a subalternist cosmopolitics requires careful rethinking of both terms.

‘Cosmopolitanism’ and ‘cosmopolitics’ are etymological siblings sometimes used interchangeably in English translations of German and French texts. Many readers may justifiably assume their synonymy. Since cosmopolitanism has become a lightning rod for debates among prominent scholars including Kwame Anthony Appiah, Ulrich Beck, and David Harvey, I should clarify that this essay doesn’t centrally address cosmopolitanism as a neo-internationalist ethic, postnationalist imaginary, or transnational expansion of political sentiments and practices (cf. Appiah, 2005; Archibugi, 2003; Beck, 2004; Beck and Grande, 2010; Cheah and Robbins, 1998; Gilroy, 2010; Harvey, 2000; Morgan and Banham, 2007). While Derrida’s sense of cosmopolitics shares with these discussions a concern for the problems of exile and asylum and a debt to Kant’s political philosophy, it stands out by prominently emphasizing the limits and contradictions of democracy as a practice of hospitality.

Moreover, as Latour (2004b) has pointed out in an exchange with Beck (2004, 2005) concerning this terminological juncture, the humanist sense of cosmopolitanism presumes a highly-circumscribed cosmos. If the cosmos of cosmopolitanism has no place for matters such as trees, marine microbes, phytoliths, mosquitoes, scanning tunneling microscopes, and ghosts, should we call it a ‘cosmos’ at all? Given the ethos of disciplinary hyperspecialization across the humanities and sciences today, we increasingly find ourselves adorned with ontological blinders that render invisible the vast majority of today’s planetary (and extra-planetary) organisms and beings. In the heyday of structuralism and semiotics, before the rise of speculative realism and object-oriented philosophy (see, e.g., Bryant, 2011; Bryant et al., 2011; Harman, 2009, 2011; Meillassoux, 2008), Continental metaphysical and political thought tended to indulge a radical disavowal of beings external to the human-world correlate. The inverse image of this antirealism was the sciences’ structural disengagement with political and scientific worlds and modes of representation. On both sides, these delimitations have stained the legacies of 20th-century thought and politics. This essay, thus, labors to articulate a new alliance, a cosmopolitical program affirming the symmetry of radical realist empiricism and critical attention to the limits of representation.

Critical Cosmopolitics

Derrida develops his concept and practice of cosmopolitics most clearly in ‘On Cosmopolitanism’, an essay conceived for and presented to the International Parliament of Writers (IPW) in 1996. He took the occasion to address the efforts of the IPW to establish a network of cities of refuge hosting persecuted intellectuals. Dusting off the term from Kant, Derrida describes the establishment of such cities of refuge as ‘a new cosmopolitics’ (Derrida, 2001: 4, emphasis in original). He refers here primarily to Kant’s essay ‘Perpetual Peace’, which proposes a system to progressively unify world-politics. This cosmopolitan world-politics served Kant as a guideline for extending routes of commerce, a system of rights and laws that would facilitate interactions among representatives of different nations.

In Kant’s (comparatively ill-regarded) political writing, Derrida locates a potent conceptual and ethical prospect: universal hospitality, or, ‘the right of a stranger not to be treated with hostility when he arrives on someone else’s territory’ (Kant, 1991 [1795]: 105). ¹ While Kant’s notion of hospitality pertained to the expansion of early modern European economic control into the emerging colonial world, Derrida extricates universal hospitality from this historical milieu. He effectively transforms Kant’s ethic of hospitality from a broad commitment to welcoming ‘strangers’ (and thereby expanding markets) into a claim about persecuted subjects’ rights to asylum.

In justifying the establishment of cities of refuge, Derrida invokes pervasive contemporary forms of violence, including censure, terrorism, and persecution by both state and non-state actors. Although Derrida recognizes that victims of such violence are ‘nearly always anonymous’, he and the IPW focus, with some justification, on ‘intellectuals, scholars, journalists, and writers – men and women capable of speaking out (porter une parole) – in a public domain that the new powers of telecommunication render increasingly formidable’ (Derrida, 2001: 5, 6). Where Kant sought to establish the right of all strangers to enter all territories, Derrida focuses principally on the role of cities in hosting the persecuted, and – even more specifically – persecuted intellectuals.

Of these many victims, whose access to technologies of speech has been targeted, Derrida cites none. In fact, he states, ‘to cite the best known would risk sending the anonymous others back into the darkness (mal) from which they find it hard to escape, a darkness which is truly the worst and the condition for all others’ (Derrida, 2001: 6). In the context of the lecture, this statement marks Derrida’s divergence from the IPW’s elevation of specific political figures, particularly Salman Rushdie, to the status of celebrity exiles. As Gideon Baker puts it, ‘what Derrida intended here was to identify cities of refuge, and asylum in general, with the foreigner in all her foreignness, with the nameless stranger, something which the IPW, like all hosts, fell short of in its offer of a limited, in this case literary, hospitality’ (2010: 91; see also Kelly, 2004). On the one hand, Derrida extends the promise of radical hospitality to all those constituted as externalities by state violence. On the other, this radical promise seems compromised by Derrida’s readiness to dispose with the non-innocent act of naming the other. This evasion of specification, singularization, or the empirical seems somewhat disingenuous and may even presume a form of authorial sovereignty, as if Derrida’s voice and inscription were the ur-text (or, at least, a particularly powerful form) of political representation.

The mutuality of representation and elision invoked here reflects Derrida’s established ideas, particularly his notion of différance, the claim that any representation is destabilized by its own production of externality. But, following Gayatri Chakravorty Spivak’s (1974, 1988) defense of Derrida’s earlier work, the engaged intellectual’s awareness of the limits of representation should not paralyze the messy work of knowledge production. Intellectuals have an ethical obligation to innovate new ways to speak and care for non-subjects whose experiences initially appear to be unrepresentable within dominant political and academic institutions.

Despite this evasion of the empirical, to which I return momentarily, Derrida’s essay treats difficult and timely problems of asylum, refuge, exile, and hospitality. As Simon Critchley and Richard Kearney (2001: ix) point out in the preface to On Cosmopolitanism and Forgiveness, Derrida’s concern with refuge reflects his interest in the fraught politics of immigration in mid-1990s France, in the wake of state crackdowns on the sans papier and resulting mass demonstrations. It is this context that apparently compels Derrida to take the term ‘cosmopolitanism’ from Kant and rationalist state discourse, affirming the city – rather than the state – as the site of refuge. In this respect, Derrida and the IPW seem to scale down Kantian cosmopolitanism, invoking the possibility of a democracy to come that no longer requires the state as a guarantor of individual rights.

Derrida diverges from Kant’s vision of the political scale of cosmopolitanism, but he ultimately reiterates Kant’s effort to grapple with a central contradiction. Namely, while the cosmopolitical ideal upholds the promise of unconditional hospitality, the right to refuge must be limited in practice. For Kant such limitations took the form of inter-state treaties regulating issues of refuge and immigration. At the close of his essay, Derrida (2001: 22–3, emphasis in original) poses this tension as the key contemporary problem:

It is a question of knowing how to transform and improve the law, and of knowing if this improvement is possible within an historical space which takes place between the Law of an unconditional hospitality, offered a priori to every other, to all newcomers whomever they may be, and the conditional laws of a right to hospitality, without which The unconditional Law of hospitality would be in danger of remaining a pious and irresponsible desire, without form and without potency …

Through the language of hospitality, Derrida takes the act of inviting an outsider to enter one’s home as the model for democratic practice (see also Derrida and Dufourmantelle, 2000). Derrida’s effort to grapple with the problem of hospitality is well-trod intellectual terrain (e.g. Barnett, 2005; Candea and da Col, 2012; Westmoreland, 2008). But exploring Derrida’s cosmopolitics requires us to dwell momentarily on the ontological status of this subject afforded the promise of hospitality. Derrida receives the subject of hospitality categorically as l’etranger, the stranger or the foreigner. L’etranger is not an absolute alterity, but rather a subject with the potential to become recognizable and visible to law and society:

[A foreigner] is not only the man or woman who keeps abroad, on the outside of society, the family, the city. It is not the other, the completely other who is relegated to an absolute outside, savage, barbaric, precultural, and prejuridical, outside and prior to the family, the community, the city, the nation or the State. The relationship to the foreigner is regulated by law, by the becoming-law of justice. (Derrida and Dufourmantelle, 2000: 73)

The ontological and empirical status of the foreigner potentially subject to acts of hospitality remains problematically indeterminate in Derrida’s work. The crucial difference between Derrida and the subalternists who engaged his philosophy of the limit resides precisely in their differential willingness to circumscribe the unknowable within an empirical or historical context. For example, Spivak (1988) explores the experience of 19th-century widow self-immolation in the same frame that she interrogates the parameters of academic historical discourse today that render such an act normatively incomprehensible, pointing to late-modern scholars’ epistemic continuity with colonial administrators. Likewise, readers encounter Dipesh Chakrabarty’s (2000) defense of the epistemic legitimacy of Bengali historical narratives that present spirits as agents inciting rebellion in the same frame through which he explores how historians would problematically deny the sociohistorical truth of this claim.

Such subalternist accounts enable readers to identify and interrogate the parameters or limits of the knowable. The subaltern is a known unknown, a knot in the fabric, a being that scholars can circumscribe and engage only if they seriously interrogate the boundary conditions of their own thought, and perhaps not even then. Derrida seems satisfied to interrogate boundary conditions without risking positive knowledge. He offers no tools for unraveling the knot beyond contemplating production of l’etranger. He explicitly refuses to risk empirical engagement with forms of life or being that resist immediate translation into our regimes of knowledge. Derrida can’t stomach the ‘pious and irresponsible desire’ of an unconditional law of hospitality because he considers such a radicalization of democracy to be impotent, maintaining that its sense collapses in internal contradiction. But if this is the case, if radical democracy is utopic piety, so is Derrida’s refusal to cite a single victim of violence, an attempt to avoid complicity in sending others into le mal, darkness, evil, abjection. Perhaps Derrida (2001: 6) doesn’t send ‘the anonymous others back into the darkness’, but only because he doesn’t risk inviting them into his text (or his home?).

In critiquing and reforming Indian colonial historiography, the subaltern studies group provided numerous examples of how to interrogate the epistemological parameters of historical knowledge production by writing the collective subjectivities and knowledge practices of the abjectly marginalized into Indian history (as well as recognizing the limits of this project) (e.g. Chakrabarty, 2000, 2002; Guha, 2002; Spivak, 1988). They gave voice to those otherwise exiled from the historical cosmos. This is a key point of distinction from Derridean ‘hauntology’ that clarifies the political value of interrogating the limits of representation, history, and knowledge.

Subaltern studies nevertheless entailed problematic risks. Subalternists’ detours into the haunts of the historical commons risked conflating the temporarily unknown with the permanently unknowable. Some critics reasonably consider claims to the mutuality of inclusion and exclusion in ontological politics a tautological and formal problem, verging on a red herring. For example, Warren Montag (1998) critiques Spivak’s (1988) defense of Derrida in ‘Can the Subaltern Speak?’ on such grounds. He claims that the assertion of a complete disjuncture between forms of political/academic representation (such as Michel Foucault and Gilles Deleuze’s gestures to a French Maoism [Foucault, 1977]) and specific Indian historical practices (such as Hindu widow self-immolation) took subaltern studies down a ‘transcendental turn’. Montag proposes that Spivak’s delimitation of the term ‘subaltern’ relegates marginalized subjects to positions of figural externality. In effect, he suggests that such a position extends the formal production of externality and the attendant destabilization of sense and logic termed différance into a space where it manifestly doesn’t belong.

This risk remains when we invoke the subaltern after the heyday of subaltern studies. But I think that a subalternism true to Derrida’s cosmopolitics might be much more than a mapping of semiotic thought concerning representational gaps onto political and academic terrains. As Martin Hägglund (2008) emphasizes, Derrida’s concept of hospitality fit within a broader vision of deconstruction as the very process of temporal change. Deconstruction is ce qui arrive, whatever arrives. Subalternism sought to grapple sincerely with marginal narratives as they arrived. It treated their divergence from the hegemonic structures of colonial historical narrative as a call to consider the violent exclusions of Western political and epistemic representation.

What is cosmopolitics in this light? Drawing from subalternist ethics, ‘cosmopolitics’ might designate scholars’ and activists’ efforts to imagine, produce, and govern more livable worlds (past, present, and future). Our acts of worlding unfold through the constant experience of encountering the unknown, facing and responding to ce qui arrive. Cosmopolitics might also become a set of techniques for responding with sincerity and generosity to existing forms of scientific and political governance. It’s a response and not a reaction to these sites of governance in that it approaches them empathetically as important institutions operating with distinct ontological politics that it cannot fully supplant or succeed. Yet this cosmopolitics would also foreground the failures of present political and scientific systems to adequately represent and alleviate widespread suffering among diverse beings – the subaltern – of which humans are only a fraction (cf. Kirksey and Helmreich, 2010; Mitchell, 2002).

Derrida’s and the subalternists’ contributions to my sense of cosmopolitics can thus be summarized in two points. First, we must embrace openness to whatever arrives, a willingness to host unexpected guests. Second, we must attend to the violence intrinsic to this act of hosting the stranger. Hosting is, at once, an ontological, empirical, and political problem. Reconceiving processes of politico-scientific representation as the act of hosting will entail a Latourian ontology of ‘irreduction’. As Latour puts it, ‘nothing is, by itself, either reducible or irreducible to anything else’ (1988: 214; see also Harman, 2009: 12–15). A guest should not be reduced to the terms of the host’s ontology or representational system. Whether the guest is a newly-discovered microbe in a scientist’s laboratory, an undocumented worker, a political uprising contesting state violence, or a feral cat calling for food at the door, it must be taken as a singular event or emergent actor introducing world-altering claims and conditions. Unexpected arrivals surprise. But hosts must work to accommodate guests in the form of their presentation, as they are given in experience.

While this is a prescription for how we might better live together, it’s significant to note its congruence with existing scientific practices. For example, drawing on Ludwik Fleck’s (1981 [1935]) understanding of experimental practice and his own research into the history of experimentation in 20th-century molecular biology laboratories, Hans-Jörg Rheinberger (1997) illustrates that, contra Karl Popper (1959 [1934]), some experimental systems are not designed primarily to test hypotheses. Instead, they are open systems that condition the possibility of encountering truly novel epistemic things:

A researcher … does not, as a rule, deal with isolated experiments in relation to a theory, but rather with a whole experimental arrangement designed to produce knowledge that is not yet at his disposal. What is even more important, the experimental scientist deals with systems of experiments that usually are not well defined and do not provide clear answers. (Rheinberger, 1997: 27)

But openness to the arrival of unexpected strangers is not enough. Enacting hospitality constitutes limits that estrange beings, exiling them into an exteriority haunting the commons. We can’t invite everyone into our texts, homes, and parliaments. Hospitality entails exclusion. We welcome some strangers and turn others away. An unconditional hospitality is impossible (Derrida, 2001: 103). Derrida’s stance with respect to this coproduction of the positive and the negative, or the empirical and the critical, didn’t amount to an ethical affirmation of hospitality or justice. The impossibility of unconditional hospitality sends Derrida, again, scurrying back into the philosophical archive, pointing insistently to the aporias of Western epistemology. This work is important, but it begs ethical criticism and empirical symmetry. The essential impossibility of unconditional hospitality is no excuse for the refusal to attempt partial, situated acts of representation, invitations into the text, the home, and the parliament.

Such acts will never amount to hospitality-as-such, to the essence of hospitality, to representation without negation. They will always be non-innocent. But we shouldn’t invoke the limits and violence of positive representation as an excuse for refusing the empirical, refusing to name the oppressed because others go unnamed. Critical cosmopolites have a crucial archaeological task: to excavate the middens, or refuse heaps, of our existence, accounting critically for the haunts of our insistent empiricism. This is my vision of subalternism (and, for that matter, science). It takes Derridean critiques of presence and representation seriously, but diverges in insisting that we represent despite its imperfection, attending to the violence, negation, and abjection inherent to our acts of positive representation.

Empirical Cosmopolitics

Derrida subsumes the unknown to the unknowable. The effect of this subsumption with respect to political matters such as the repression of intellectuals’ voices is stunning. Derrida’s effort to combat violence entails a refusal to name oppressed activists and intellectuals. His formulation of hospitality as a cosmopolitical project thus suffers from an ethical and empirical paralysis. It encounters subjects arriving and demanding hospitality with the refusal to represent or host them on the grounds that true democracy is impossible in essence. One hosts others not out of a desire to condition their sustained presence or out of a call to respond to their demands, but out of the necessity of being through time, of enduring the existential process of reality. Being itself is the experience of encountering others (seemingly of all ontic kinds).

The vocabulary of hosting mobilizes the home, oikos, as a pleasant metaphor for the ecology of relations termed ‘time’. We host. We become-other. It’s a metaphor present in Stengers’s (2010, 2011) and Latour’s (2004a, 2004b) versions of cosmopolitics, though they emphasize the institutional forms of representation that they know best: science and politics. While Stengers’s and Latour’s historical and empirical accounts focus centrally on institutions of knowledge production, the sciences, they rather insistently narrate these institutions’ knowledge practices in political terms.

Latour’s (1993, 2004a) central concept-work undermines the ontological distinction between the human and the nonhuman. In so doing, he conceives a unified institutional apparatus, an alternative to existing forms of political and scientific representation. He promotes an irreductionist ‘flat’ ontology, a refusal to accept that some beings are intrinsically more real or consequential than others (see, e.g., DeLanda, 2002: 58). All actors can associate with all others. Mickey Mouse, the Mayanist epigrapher David Stuart, the Guatemalan state, mitochondrial DNA, the ontological turn, Gabriel Tarde, the former planet Pluto, and my hardcopy of Politics of Nature are all equally real entities (though they are not all equally consequential).

Latour’s cosmopolitics follows from this ontological position and thus remains largely consistent with his earlier ‘actor-network theory’, a set of sociological techniques for understanding the production of scientific facts (Latour, 1987, 1999a; Latour and Woolgar, 1986 [1979]). The empirical basis of Latour’s (2005) flat ontology and unified politico-scientific common world issues from his ethnographic work on science in action. Against arguments that solid scientific accounts of the world emerge either through the formal processes of science (as claimed by some logical positivists, Mertonian sociologists, and internalist historians) or through the social contexts of science (as claimed by some sociologists of scientific knowledge, externalist historians, and poststructuralists), Latour’s empirical studies reveal that scientific facts emerge through practices that do not distinguish a priori the social from the natural, society from science (Latour, 1999a; Latour and Woolgar, 1986 [1979]). There’s not merely extensive traffic across the boundary between science and society; science, society, and the boundary are embroiled emergent properties of the traffic itself. Society and nature are internally-heterogeneous actors that do not exceed (and emerge out of) the sum of their parts, social scientific and scientific knowledge practices, bodies, and objects. In the terms of Gilles Deleuze and Félix Guattari (1987 [1980]: 12–13), they are maps that should be put back on the territory. ²

If no actor can claim ontological primacy in accounts of scientific knowledge production, we need an explicitly empirical cosmopolitics as the complement to our critical cosmopolitics. Social explanations of science and natural explanations of society are red herrings distracting ethnographers (and other wanderers?) from the task at hand, a determined empirical engagement in, with, and of the world. Latour is most explicit and adamant about this persistent empiricism in Reassembling the Social (2005). There he elaborates the metaphor of the fieldworker as an ant poking around, refusing to shift scale and indulge explanations, generalizations, or critical frameworks. The researcher can do no more than diligently trace the network with the aspiration to produce a good description, a good account, a good map to reterritorialize on the topos of the real.

This program is deeply scientific and deeply political. In When Species Meet Donna Haraway (2008) lauds Latour’s and Stengers’s vocabulary of ‘politics’, stretching the etymology a bit to point out the term’s relation to polite in addition to polis. For Haraway (2008: 92), hospitality, good manners, or politesse, is a cosmopolitical and biopolitical practice of ‘articulating bodies to other bodies with care so that significant others might flourish’. For those like me skeptically wandering alongside anthropology’s multispecies turn (see, e.g., Kirksey and Helmreich, 2010), these bodies take form as a multitude of co-inhabitants dwelling together in the anthropocene’s precarious common worlds. ³ Haraway (2008: 19–27), a protagonist of this movement, thinks that Derrida (2002) didn’t venture far enough into the ontological open in exploring the implications of his naked encounter and very French exchange of le regard with his cat one morning in the bathroom. He opened his home to that feline. And he opened his ontology and politics to the sans papier (well, sans nom). But he didn’t open his ontology and politics to sincere engagement with the cat and its human companions or biologist spokespersons. Derrida didn’t seriously consider the cat’s ‘umwelt’ (von Uexküll, 2010 [1934]), its phenomenal world and experience independent of a human correlate.

From a Latourian perspective, such an anthropocentric position is evident in Derrida’s (2001) equation of ‘intellectuals, scholars, journalists, and writers’ with those ‘capable of speaking out (porter une parole) – in a public domain’. Latour (2004a) teaches that spokespersons require ‘speech impedimenta’, chains of mediators that both enable and hinder speech. For the botanists and pedologists that Latour (1999a) describes in his ethnographic essay ‘Circulating Reference’, such impedimenta include maps, rusty nails and tags affixed to tree branches, plants, notebooks, shelves for organizing plant samples, compasses, clisimeters, topofils, pedocomparators, dirt, and Munsell charts. Scientists’ representational competencies are predicated upon their engagements with such worldly things. They are no more or less necessary to addressing matters of scientific concern than telecommunication technologies – or, for that matter, parliaments – are necessary to addressing matters of political concern.

By reading Derrida’s ethic of hospitality against Latour’s efforts to innovate more realist and realistic scientific practices, it becomes possible to designate two key acts of deferral or exclusion at work in Derrida’s cosmopolitanism. It’s useful to distinguish here between implicit and explicit acts of deferral. Haraway has already attended effectively to the implicit act of deferral in Derrida’s philosophy: the circumscription of legitimate epistemological objects to the human subject and the world. The cat that he allows into his philosophical reflections quickly loses its furry and fleshy worldliness, becoming an empty placeholder for the problem of whether animals suffer, writ large in principally metaphysical terms. So his implicit deferral is exemplified by failing to take the cat seriously. Perhaps more problematic is the explicit deferral; he can’t seriously advocate specifying the voiceless, in all of its non-innocent partiality.

If we hope to achieve more vibrant, collective, ecological, and hospitable common worlds, this will not do. A critical cosmopolitics that obsesses with the violence of representation to the point that it founders there has little utility for making more open, livable, and ontologically-heterogeneous worlds. This is very much the project of Latour’s cosmopolitics, as an effort to rethink our systems for representing so-called ‘humans’ and ‘nonhumans’ with the objective of producing peaceful worlds that do not lend ontological primacy to the human/nonhuman distinction or, presumably, any other ‘great divide’ (Latour, 1993). In this sense, thus, Derrida’s and Latour’s cosmopolitical programs emerge, symmetrically, as mutual supplements.

But significant tensions between the programs persist. Latour’s use of the term ‘cosmopolitics’ doesn’t derive from Kant’s sense. He takes the term from Stengers, who has explicitly distanced her concept from the Kantian etymon, pointing out that the project for peace cannot begin with the extension of Western values as a universal framework for governance:

Although the idea of peace among various peoples must have some significance, we need to start not like Kant from promises the West might flatter itself for propagating, but from the price others have paid for this self-definition. It is not so much peace that we have brought to other peoples and ourselves, but a new scope, new methods, and new modalities of warfare. (Stengers, 2010: 79)

In treating the cosmos as an ecology of practices, Stengers (2010: 10–11) conceives cosmopolitics as the process of affirming what Deleuze and Guattari (1994) term ‘new immanent modes of existence’. Stengers seeks to make the sciences and the history of the sciences more attentive to the empirical constraints of scientific practice. These constraints, these experimental arrangements, constitute new modes of existence, new ways of inhabiting and conceiving the world for scientists and their epistemic objects alike. Stengers insists that scientific propositions emerge out of specific material systems that endow these propositions with truth-values. The meaningfulness, felicity, and consequentiality of scientific facts emerge through scientists’ specific material practices, their tinkering or empirical wandering with and within experimental systems. As a result, it’s a mistake to assume that the propositions emergent out of specific systems have intrinsic epistemological or empirical bearing on exogenous worlds or modes of existence.

In her careful historical accounts, Stengers concentrates primarily on faulty assumptions that new scientific programs disqualify their predecessors. Her archaeology attends, for example, to problems with how the rise of dynamics disqualified thermodynamics, and how the rise of theoretical quantum physics subordinated phenomenological physics to the status of mere experimentation with ‘heterogeneous empirical phenomena’ (Stengers, 2010: 261). But this method challenges a broader set of normative scientific assumptions as well. As Stengers (2010: 11) puts it, ‘the sciences do not owe their existence to the disqualification, with which they are identified, of so-called “prescientific”, or nonrational, knowledge’. Stengers’s cosmopolitics, thus, opens up the possibility that ostensibly contradictory scientific programs or metaphysical systems can coexist peacefully, affirming their independent logics of composition and knowledge production.

The assertion that the sciences could coexist peacefully with worlds that entail different metaphysics should not be confused with cultural or epistemological relativism. As Latour argues, the very possibility of relativism presumes a kind of ontological externality, Nature, that has a singular or absolute integrity and logic. Relativism indulges the conceit that humans construct their own webs of symbols or fields of power against the backdrop of a timeless, unchanging Nature. Relativism cannot coexist with a more rigorous ecological thought. An ecological approach to ontology and relationality evades the dominant modern duality pairing multiculturalism with what Latour, following Eduardo Viveiros de Castro (1998), terms ‘mononatralism’:

… by losing mononaturalism, the collective frees itself at the same time from multiculturalism. Up to now, pluralism had never been anything but a rather facile tolerance, since it never poured out its generosities except by drawing on an unchallenged common fund. By losing nature, we also lose the fragmented, dispersed, irremediable form that it gave, by contrast to all the multiplicities. … Relativism would disappear with absolutism. There would remain relationism, the common world to be built. (Latour, 2004a: 220)

Stengers and Latour thus advocate an ontological pluralism, conditioning and affirming the growing trend in anthropology to substitute the problem of ontology for the problem of culture (e.g. Candea, 2008; Carrithurs et al., 2010; de la Cadena, 2010; Viveiros de Castro, 2004). It’s crucial here to acknowledge that the ontological pluralism of Latourian cosmopolitics is itself a question of cohabitation, coexistence, and hospitality. How might we build a more open world, a world where scientific and political spokespersons could work in common towards the goal of making our collective-becoming and becoming-collective more peaceable? To some readers this problem might look very much like a humanist cosmopolitanism. But the ‘we’ composing new worlds, representing in new parliaments, and hosting in new homes is not exclusively ‘human’.

Latour’s program reforms our representational systems in order to accommodate the ontological pluralism of contemporary coexistence. He adopts an empirical openness to all actors that arrive at the door. He hosts spirits alongside photons. But I suggest that he does not effectively attend, as Derrida and the subalternists rightfully insist that we must, to the violence of composition, representation, and hosting (cf. Puig de la Bellacasa, 2011; Strathern, 1996; Watson, 2011).

Here Latour and his followers do not necessarily need to appeal to Derridean philosophy of the limit, an act that would put at risk their devout commitment to a realist ontology. The limits of representation may appear much closer to home. For example, Maria Puig de la Bellacasa (2011: 101, n. 8) productively identifies Latour’s reluctance to politicize the limits of representation as a discordance between his and Stengers’s respective concepts of cosmopolitics:

Stengers’s cosmopolitical proposal gives a prominent space to the ‘victims’ – those who retain no power to represent themselves; to groups who disrupt, or fall out of the cycle of representative politics; and to the ‘idiots’, who don’t want to be ‘included’ and cannot ‘contribute’ because they feel that ‘there is something more important’ than the proposed issue.

The confounded lawyer who employs Bartleby is mobilized by the scribe’s disengagement from their daily labor; he’s called to coax Bartleby into the common labor of the office. But he eventually moves away from the office and abandons the scribe. Stengers reads the lawyer’s abandonment as a non-cosmopolitical act, a failure to cohabitate peacefully, negating the relevance of Bartleby’s mode of existence. We could say that the lawyer failed to be a good host, or failed to render his common world vulnerable to an ostensibly contradictory mode of existence. Through this example Stengers presents cosmopolitics as a commitment to opening worlds to practices or beings that seem otherwise inexplicable. If, when confronted with something he could not understand, the lawyer had insisted on proceeding with business ‘in the presence of’ Bartleby, he would have become a cosmopolitical actor, a cosmopolite: ‘Giving this insistence a name, cosmos, inventing the way in which “politics”, our signature, could proceed, construct its legitimate reasons “in the presence of” that which remains deaf to this legitimacy: that is the cosmopolitical proposal’ (Stengers, 2005: 996). Stengers’s cosmopolitics, thus, entails the insistence that we build worlds alongside – if not necessarily with – those working from wholly different blueprints.

In Politics of Nature, Latour further develops the political and metaphysical implications of Stengers’s cosmopolitical proposal. Cosmopolitics, for Latour, addresses two problems, which he tends to designate as ‘modernism’ and ‘political ecology’. The first problem entails the separation of powers between representational systems designed to account for nature and society, which Latour addresses by further elaborating the ‘parliament of things’ (Latour, 1993: 142–5). The second problem is a newer theme for Latour and bears more directly on the question of his unlikely alliance with Derridean and subalternist thought. Namely, Latour advocates a mode of composition or worlding that resists the seductions of a monistic and totalizing ‘political ecology’ that purports to take everything into account:

It seems, indeed, that in passing from modernism to political ecology, we pass from the imprescriptible right to ignore the majority of beings to the necessity of excluding none of them. Complexity, ‘total connectivity’, the global ecosystem, the catholicity that wants to embrace everything, all this is what always seems to accompany the erecting of an ecological way of thinking, a way of thinking rightly persuaded that in the final analysis everything is interconnected. (Latour, 2004a: 198–9)

Latour establishes a set of formal procedures for determining which actors should be taken into account and how they should be arranged for peaceful and productive cohabitation within the commons. Each emergent actor presents itself to a commons whose denizens are charged with determining whether it would aid in producing a more peaceful world, a world characterized by a progressive degree of flourishing. Latour (2004a: 102–16) indicates that the process of denying entry to the commons should always be explicit. Those denied entry become ‘small’ or ‘little’ ‘transcendences’ (Latour, 2004a: 196, 199) afforded the right to appeal their exclusion. It remains important to recognize that Latour’s parliament differs from that of the ‘modern constitution’ because it lacks the power to withdraw the existence of excluded entities. It even ‘trembles at the possibility of committing an injustice’ (2004a: 179) by producing enemies that will haunt the commons and potentially become co-actors in a future iteration. In other words, exclusion is provisional; excluded actors are neither negated nor relegated to the space of an absolute externality. They reserve the right to appeal.

But is the right to appeal a sufficient guarantee to immunize against the violence of exclusion intrinsic to actual modes of representation and cohabitation? In designing a representational system as a thought experiment supposed to supplant the romantic seductions of both the modernist settlement and political ecology, Latour seems to affirm and extend an equally non-democratic divide between actors included in and excluded from the commons. Each actor excluded from the common world, each being that denizens of the commons do not host, becomes a ‘small transcendence’, a provisional and produced externality. Ultimately, ‘small transcendences’ seems to be a misleading designation for those excluded from the common world, even if they are afforded the right to appeal.

The violence of turning emergent actors away at the door is irreducibly situational, and – as I’ve put it elsewhere – ‘we still need social activists, scientists, and politicians to relay between the cosmopolitical citadel and the worlds of normatively marginalized actors, learning to listen to the oppressed’ (Watson, 2011: 67). Trembling at the potential injustice of exclusion is not enough. We should be specific and empirical in addressing the process of exclusion as the production of subaltern actors and innovating novel practices of listening as risky techniques of cosmopolitical care. What violence does the lawyer enact by failing to cohabitate with Bartleby? What violence did colonial administrators enact by criminalizing sati? What violence do parliaments enact by ignoring climate change? Climate change, as a set of propositions, may not seem on the surface to be a subaltern. But the silencing of climate change’s scientist spokespersons is a serious matter of concern. Silencing produces subaltern strangers that haunt the common world. When those who are called to action by climate change cannot be heard, denizens of the commons jeopardize the very possibility of their collective becoming; climate change calls attention to the precarity of the commons. The repressed returns. Perhaps climate change is a ‘hyperobject’ (Morton, 2011) that threatens to swallow the commons whole (cf. Clark, 2011). Subalternist cosmopolitics foregrounds the question of how to relay into the haunts of the common world and innovate means to listen to provisionally excluded actors at constant risk of failure.

These are irreducible cosmopolitical problems, and Latour provides a skeletal framework that we must now flesh out. Much is elided when we address the problems of peacefully cohabitating with self-immolating Hindu widows and with climate change using the same totalizing conceptual apparatus, even as a machine for generating openness to whatever arrives. But we cannot return to a dualist separation of powers or to a monist political ecology. A unified politico-scientific system, cosmopolitics, remains useful as a fragile, federal guideline for a multitude of disparate, situated common worlds.

Puig de la Bellacasa (2011) poses similar questions in tracing the powers and limits of Latour’s concept ‘matters of concern’. ‘Matters of concern’ functions to expose the deeply anti-empirical implications of ‘matters of fact’. Where matters of fact are presumed to be final, closed, and independent of their material construction, matters of concern retain and represent relations with the actors that made them possible. Puig de la Bellacasa enacts a feminist intervention into such varied matters by drawing out the affective senses of the term ‘concern’ that Latour ignores. Where Latour’s ‘concern’ effectively denotes collective interest, Puig de la Bellacasa (2011: 89) emphasizes how the term also evokes ‘trouble, worry, and care’. As she puts it:

Understood as affective states, concern and care are … related. Care, however, has stronger affective and ethical connotations. We can think on the difference between affirming: ‘I am concerned’ and ‘I care’. The first denotes worry and thoughtfulness about an issue as well as the fact of belonging to those ‘affected’ by it; the second adds a strong sense of attachment and commitment to something. Moreover, the quality of care is more easily turned into a verb: to care. (Puig de la Bellacasa, 2011: 89)

To address such issues politically and scientifically, concern doesn’t suffice. As Puig de la Bellacasa emphasizes, we also need to enact forms of care. The term ‘care’ foregrounds problems of vulnerability, marginality, precarity, and structural violence. There’s no reason to expect that marginal or abject actors – the subaltern – whose experiences, narratives, and forms of being are unrepresented within and even unrepresentable by contemporary institutions of politics and science would have clear spokespersons in Latour’s proposed parliament (Watson, 2011). They aren’t just matters of concern. They’re matters of care. We need care. Care means taking representation slowly. Care means attending to the gaps, the ignorances produced by our knowledges, the subaltern holes and knots in the fabric (Chakrabarty, 2000: 106). Care means hospitality. Cosmopolitical care might also mean anarchism, as an activist reconstitution of the commons in a finite world with no recourse to the holy trinity of God, Nature, and Neoliberalism. I’m not certain.

Coda

We live in difficult and challenging times with respect to how we think about the problem of the commons. The commons, what we hold together and – presumably – what holds us together, requires reconstruction in rigorously nonhumanist and anti-correlationist terms. ⁴ Subalternist cosmopolitics entails producing cohabitable worlds and corollary forms of local governance with actors that don’t share the cognitive, bodily, and metaphysical forms of human being. Our political spokespersons shouldn’t continue to assume that human beings are the only legitimate actors meriting rights to select a body of representative proxies. The prospect that a political system constructed to represent human interests and constituencies might be able to know and care for the ontologically disparate worlds and experiences of other-than-human beings may amount to the present era’s predominant ideological delusion.

Our political systems can’t act ecologically because they weren’t built to do so. They’re artifacts of an 18th- and 19th-century romanticism that presumed the absolute externality, immutability, and permanence of Nature. We live in a more precarious age, and one that requires more heterogeneous, experimental, and scientific forms of governance. The human shouldn’t be conceived as the one and only host in today’s oikos. The parliament shouldn’t assume that humans representing other humans can exercise sovereignty over the entirety of the Earth as we know (and don’t know) it. Subalternist cosmopolitics is a prescription for conceiving politico-scientific governance that’s immune to human exceptionalism and sensitive to the impossibility of abstract justice. The twilight of the anthropo-political age has arrived. The coming night will entail a great deal of critical and empirical wandering, tinkering with and within the small corners of a co-inhabited cosmos as we seek to produce more just and livable worlds, worlds with brighter dawns.

So what’s the ethnographer to do? To be an empirical wanderer, a host calling attention to how guests impose their own ways of being-in-the-world? Guests are obdurate. Perhaps it’s worth recalling here that the French words for guest and host are the same: hôte. In his book The Parasite, Michel Serres (2007 [1980]: 16) puts it this way: ‘There are some black spots in language. The field of the host is one such dark puddle. In the logic of exchange, or really instead of it, it manages to hide who the receiver is and who the sender is, which one wants war and which one wants peace and offers asylum’. We must both risk and remember: risk extending asylum, empirically, to emergent matters of care; and remember that who’s the host and who’s the guest remains an open question.