Skip to main content
Open Access

Materiality, materialization

University of Chicago


Comment on Hull, Matthew. 2012. Government of paper: The materiality of bureaucracy in urban Pakistan. Berkeley: University of California Press.

What is materiality? And is it of a different order from semiosis?

In this commentary I reflect on the place of materiality—that quality of material phenomena—and materialization—the processes by which such qualities are actualized in social life—in Matthew Hull's brilliant book, Government of paper (GoP). The analytic materiality is central to Hull's ethnography, both theoretically (it informs his engagements with James Scott, Bruno Latour, and contemporary linguistic anthropology, among others) and methodologically (it guides his ethnography and analysis). And yet, while Hull demonstrates the term's pragmatic purchase, he never quite tackles what is “meant” by it head on.1 Bracketing whether this is in fact the point, in what follows I comment on what I think materiality and materialization do in Hull's analysis, motioning to some issues this raises for linguistic anthropology.


Let me start with the beginning of the end: the first endnote (Hull 2012: 259; also see Hull 2003: 291). In this endnote Hull commensurates, and distinguishes, what he calls “graphic artifacts” from “textual objects,” “written objects,” and most important for my discussion here, “text-artifacts.” One reason for this relexical-ization is “to emphasize the non- and para-linguistic semiotic functions of this type of artifact.” While Hull tells us that his graphic artifact is not conceptually different from Silverstein and Urban's (1996) “text-artifact,” there is a pragmatic, or rhetorical, difference. Hull reanimates, but also resists, the terms of his engagement with linguistic anthropology. But what is this difference in emphasis? Why this graphic/textual change? Why “graphic” and not “text”? And, by analogical skip and leap, why “materiality” and not something else? And what might that something else be? To what is materiality opposed such that we necessitate it in our analyses?

I see this rhetorical shift as issuing an important, if subtle, challenge to linguistic anthropology. As Hull alludes (2012: 13–14), linguistic anthropologists have tended toward a semiotic monism in distinction to dualist oppositions of materiality and representation, often through the semiotic category of indexicality (though also metapragmatics and ideology) (Keane 2003). If we take this monism at face value, GoP is not in distinction to, but an exemplar of, linguistic anthropological analysis. And yet, as the first endnote indicates, there is an ambivalence surrounding materiality and semiotics that mirrors a split that Hull implies as running through linguistic anthropology. That is, contrary to its self-representation, in its practice linguistic anthropology tends to privilege a kind of idealism, a focus on “discourse” and linguistic semiosis (and by implication, “meaning”) rather than on materiality. What is disavowed in theory returns in method and analysis. Hence the necessity of Hull's terminological intervention.

While it is easy enough to mount theoretical arguments that reconcile semiosis and materiality or frame their opposition as a nonproblem (indeed, I do it below), Hull's challenge lies elsewhere. How, indeed, do we treat our ethnographic materials? What aspects of semiosis do we focalize? What are the canonical types of cases that we base our larger generalizations upon? Perhaps face-to-face interaction where linguistic semiosis plays a large part is not the basic case upon which to build a more general semiotic anthropology. If so, then GoP calls for us to reemphasize our pragmatic maxim: the value of a theoretical concept is its methodological en-tailments. As I suggest below, Hull's usage of materiality forces us to focus on those actualized possibilities of an object-sign's use beyond the norm of its usage, beyond its putative authority, and beyond its identity as such-and-such an object-sign.2 That is, to (mis)use a term from Elizabeth Povinelli (2011), to ethnograph-ically focus on the “otherwise” of the object-sign.

The otherwise of materiality

Phrases like “The semiotic and material forms of files and their circulation …” (Hull 2012: 160) conjoin the semiotic and the material, but also oppose and distinguish them. Is materiality of a different order from semiosis? As Hull points out in GoP's introduction, if we follow the Saussurean tradition of semiology, then yes. But if we follow the semiotic of Peirce, then no. As Peirce (1992) argued, there is no semiosis without materiality, or better put, without what he called Secondness— roughly, the resistance or force exerted by that which is beyond us when we come into relation with it (see Keane 2003). For Peirce, this metaphysical category is distinct from, but presupposes, Firstness: pure possibility, quality, that which is unto itself. Firstness, however, does not “exist” unless materialized in a Second. Hence, for example, the quality of blackness must be embodied so as to be experienced, say, in a particular token-droplet of black ink. Further, both Firstness and Second-ness are in relation to and presupposed by what Peirce called Thirdness: regularity, law, habit, that which mediates Seconds with respect to each other. An example of a Third is the law of gravity or a textual genre, say, a “file”—that social type that groups together and organizes a set of singular, token-documents (and their possible uses, circuits of circulation, etc.) into some identifiable whole. As Peirce suggests, there is no general class (or type)—a Thirdness—without specific instances (or tokens)—Seconds—and no instances that can be experienced except as through their qualities—their Firstness. And vice versa. Qualities are only experiencable as actualized in particular instances, and those instances are only intelligible (i.e., can be recognized as “the same” across space and time) insofar as they belong to a type. In short, no semiosis without materiality, and no (experiencable or intelligible) materiality without semiosis.

While Hull seems to follow this view of semiotics, he also seems to hold on to the distinction, even opposition, between the material and the semiotic. In such disavowing moments, his use of materiality is similar to other post-Saussurean /poststructuralist analytics like differance, event(ness), the Real, difference/repetition, whereby materiality functions as a wedge to open up a space beyond semiosis (or, more often, “discourse”). It captions an exteriority, that which cannot be signified, that which escapes “meaning” even as it makes it possible. Implicitly, then, discourse—and by metonymic slippage, semiosis—is understood as abstract, idealist, and nonmaterial.

Hull's account, of course, is much more complex than such simplifying dualism. In his ethnographic analyses, materiality appears as intertwined, complementary, or even tangential, to semiosis. Think, for example, of the BQB villagers who, in managing to get their hands on the lists of who qualifies for compensation for built-up property, fraudulently add or scratch a name off it (chapter four). As Hull shows, the weight of paper (the sheer amount of interdiscursively tangled documentation), the distances between and accessibility of physical sites (the plots of land and the office), and the perdurance of (un)authorized inscriptions on paper all conspire to undermine, and repurpose, official discourses and government schemes. Here the relationships between discourse and materiality are complex, and neither can be thought independently of the other, even if they are assigned distinct causal roles. Similarly think of the proliferation of (sect-specific) mosques discussed in chapter five. Individuals get ahold of planning documents so as to squat in the very sites where the bureaucracy plans, in the future, to build mosques, prolep-tically anticipating the very action of the bureaucracy, materializing such plans in ways that go directly against the reason why the government planned the mosques in the first place: to obviate sectarianism. The result: the very form of the documents has changed. Now it is planners that anticipate, hiding mosques in maps by coding them as schools or other buildings, attempting to discursively intervene in the entailments of the unauthorized, material circulation of their planning documents by altering the document's form.

These incredibly rich examples show that for any object-sign there is always some other context, some other use, some other interpretation. These examples show that any object-sign—here, a paper-file—can get lost along the way, turned against it supposed function. Which is to say, there is always a gap between some formulation of an object and the range of its possible (mis)uses (and formulations therein). Materiality, here, captions the exteriority of any one semiotic event relative to some metasemiotic formulation of it. That is, it could always be otherwise. But note that this notion of materiality—that which allows this otherwise3—is only exterior to one (authoritative/normative) sketch of the interaction (viz., a planning document is used by bureaucrats to determine where a mosque will be built by the government) but not to others (viz., a planning document is used by villagers to squat), or to semiosis more generally. Materiality is, in this sense, that quality that opens up the object-sign to other competing and recalcitrant metapragmatics, even as it closes it off from others (cf. Derrida 2002: 151 on “materiality without matter”). We might see materiality, then, from two interlinked perspectives: as that which limits, or hems in, possibility; or as that which surpasses the limit or norm, that which is possibility itself. I emphasize the latter for two reasons. First, it is this sense that I find operative in GoPs ethnographic examples. Second, this reframes materiality not as a negative condition (on semiosis or discourse) or, even worse, a kind of necessity or essentialism but as a positive, operative principle sensitive to the fundamental open-endedness of semiosis.

But if materiality is that which lies beyond any particular semiotic event, what lies beyond but other semiotic events? Indeed, in Hull's examples, the exteriority invoked by materiality always implies other events of sign use. Materiality has everything to do with the necessary possibility of interdiscursivity (or perhaps we should say intermateriality), the (meta)indexical relation of one semiotic event to another (Silverstein 2005). Taken together, the otherwise of materiality is also the object-sign's Secondness, its resistance to being pinned down totally and forever, how it resists being in any one agent's control, exceeding the interactions in which it is deployed by spilling into others.

This recalcitrance is, in certain respects, retroactive, imputed to the token-object when it ends up being used otherwise and elsewhere. We only know it after the fact, even as it seems, curiously, to have always already been there, a potentiality waiting to happen. In this sense, materiality denotes the immanent/imminent possibility, or quality, of (self-)difference that is seemingly “in” the phenomenon in question. Not strictly exterior to semiosis, materiality is the virtual exteriority presupposed by semiotic events, the outside that is inside the sign, which calls for and anticipates its future reanimation and misappropriation, that is, which calls for and anticipates a next sign. This implies, however, that materiality is not “in” anything at all, even if we often experience it as such; rather, it is a relationship across events of semiosis, a property of a whole social arrangement. The fetish of the material is seemingly, then, a function of the retroactive/proleptic temporality of the otherwise.

As Hull shows us, this quality must be actualized. But how? The materialization of this potentiality into actual events is an achievement, the result of actors enmeshed in their social projects as they come to be entangled with each other. It is a social process that we can and, as Hull forcefully shows, must study. I would further suggest that materiality in the senses that Hull discusses can only be known and experienced—and thus materialized or made to exist—relative to the events wherein such materiality is pragmatically consequential. It is a function of those events. Every event of object-sign use elicits/creates certain materialized qualities from the object-signs it deploys (Nakassis 2013b): burning documents turns on paper's combustibility, using paper as an toy airplane turns on its foldability, storing it turns on its perdurability, et cetera. The point is that materiality, however else we may construe it, is mediated by actual events of encounter and (mis)use and, this being crucial, by how those events are reflexively framed for and by participants. The ontology of materiality, then, is epistemological. While the (non)perdurance of paper and the erasability of pencil (but not of pen) (Hull 2012: 144–45) matter in certain interactions, the electron spin of the atoms making up paper do not. While stamps matter to the ability of written discourse to be “anchored” in the world, the qualities of handwriting style do not (Hull 2012: 131). Why? Within the metapragmatic framing of paper in the ethnographic encounters that Hull discusses, they are not taken to make and thus do not make a pragmatic difference. Performatively, they do not matter.

Moreover, what paper is—its materialized qualities—is dialectically related to this (meta)pragmatics. Hence the numerous examples Hull gives of how paper-files materialize as a function of the social entanglements that they mediate; that is, as a function of the reflexivity of those participants whose social relations and activity are mutually oriented to the very file that their common orientation performatively materializes. Think, again, of those plans to build mosques discussed in chapter five.

Performativity and materialization

In his classic discussion of performativity, J. L. Austin (1962) was concerned with what he dubbed “explicit performatives”—sentences that “masquerade” as statements but do not quite abide by truth functionality. By virtue of fulfilling some set of felicity conditions—that is, being appropriate to some metapragmatic provisos vis-à-vis successful utterance in context—the mere use of some token-form (“I hereby declare you man and wife”) is enough to instantiate a social type (marriage), and hence its attendant entailments. For Austin, performativity was a wedge, an exteriority to the propositionality fetishized by the analytic philosophy of his time. Performativity opened up for study a linguistic space of material force that worked through but was distinct from denotation. As linguistic anthropologists have shown, the explicit performative functions by its denotational content reflexively sketching the very event of its utterance, its reflexive calibration to the token-singularity of its use being “nomically” calibrated to the social types, or institutions, which it invokes (see Nakassis 2013a for discussion). On this analysis, the felicitous performative involves a particular configuration: one where the (material) token is the (conventional) type.4

If, as I would suggest, the (mis)uses of documents that Hull discusses are performative (and in more than one way), what are the felicity conditions on such performativities, on how documents and other pieces of paper are made to get work done? There are many. GoP is an ethnographic compendium of them, detailing how paper is made to have various pragmatic effects in various contexts. One recurrent feature that enables, and to some degree unites, the various performative effects that Hull describes is the (token) singularity of the file (as a type) (Hull 2012: 144–45, 151). In the Islamabad bureaucracy, there is one, and only one, “copy” of the file. (In fact, under this metapragmatics the file cannot be copied. While its pages can be copied, the file as such cannot—which is to say, copies as copies do not carry the same performative effects.) This is, of course, what enables the mischief. If you can get your hands on the file (or the relevant document in question), if you can alter it, add your name to it, delete that reference, or change that number (chapter four), you have effectively changed the bureaucratic reality that the file purports to denote. (Compare this, however, with the performative economy of plans in chapter five.) Material inscription under such a meta-pragmatics is performative entailment. Hull traces this performative regime back to a particular British metaphysics of truth functionality ideologically vested in the fixity and transparency of written documents (2012: 246). It is this metapragmatic, or metamaterial, function that enables and constrains the file's performativities.

Jacques Derrida's (1988, 2002) engagement with speech act theory is relevant here. Derrida's (1988) critique of Austin precisely parallels Hull's intervention into linguistic anthropology. Ultimately, Derrida suggests, Austin analyzes performativ-ity in precisely the way that its coinage was meant to obviate: he treats the explicit performative by its ideality, enclosing it in ways that uncannily mirror the philosophical analysis that the performative undermines. Derrida charges Austin, in effect, with disregarding the material substratum of the performative (in this case, its status as a “locution”) (also see Derrida 2002). As a monkey wrench in speech act theory, Derrida's (1988) citationality turns on the gap between (locutionary-)token and (illocutionary-)type. Every identifiable object-sign must abide by a “code” that governs its transcontextual iterations. At the same time, that principle implies that every token-iteration has the possibility of differing from its type (also see Briggs and Bauman 1992; Silverstein and Urban 1996). Every sign can be cited, repeated in a new context to new effect. It is this property—namely, that through its iterability the object-sign's origin and identity can be “lost along the way” (1988:12)—that defines, among others, what Derrida idiosyncratically calls “writing.”

Is this not what GoP is about? How any document can be used differently in another context, literally lost in the morass of paperwork, stolen or bought out of the office? How the bureaucratic performativity of a document can be turned against itself, and in doing so entail other performative effects (perhaps even the opposite of its “code”)? There is a difference, however, from Derrida. The regime of performativity that Hull discusses is based on the metapragmatic proviso that, as I noted above, a file abides no iterations. Recontextualization in this regime of per-formativity is rematerialization: new inscriptions on, and new physical contexts of, the token-type. The issue, in the Islamabad bureaucracy at least, is not simply that every sign-type can be cited and repeated, but that every file as a token-object can be dis-placed, re-placed, re-written. And moreover, that action on the token is the equivalent of altering its type. Hence the peculiar performativity of the file is linked to the fact that the token is always already the type. If Derrida draws our attention to the token-type relation as a way to counterweigh our idealist tendencies, Hull draws our attention to the quality-token relation as a way to do the same.


Hull writes: “A file is a chronicle of its own production, a sedimentation of its own history” (2012: 117). If paper mediates social relations, it is mediated by them. Or rather, paper's materiality is materialized by serving as a medium for social action, as a contact zone, a site of interaction and entanglement. To function as a medium is to be materialized by and in the very events that it makes possible. Materialization, then, is one aspect of mediation. Or, as Hull shows, there is no mediation that is not a process of materialization. Further, to study mediation-materialization requires attention and sensitivity to those qualities of semiosis that perhaps our academic practice is wont not to materialize.



1. Of course, materiality means rather different things in different parts of the book. In these comments I draw out but one thread from this diversity.

2. Rather than “artifact” I use the composite object–sign to highlight the semiotic simultaneity, and potential otherwise, of token and type.

3. This is not to say that there are not other kinds of qualities besides materiality that enable the otherwise.

4. Hull's (2012: 132) comparison between ritual and files and Silverstein's (1993) comparison between ritual and performatives is telling here.

Constantine V. NAKASSIS

Department of Anthropology The University of Chicago
1126 E. 59th Street
Chicago, IE 60637 USA