Semiotic Review 12: Dialogues between Continental Semiotic and Linguistic Anthropology | article published September 2025 |
DOI: https://doi.org/10.71743/k99f5r76 | Copyright © 2025 Constantine V. Nakassis CC BY-NC-SA 4.0
(Im)personalizing Enunciation
Constantine V. Nakassis
cnakassi@uchicago.edu
Abstract: In this article, I discuss the question of the performative act(ness) of filmic enunciation. Drawing on linguistic anthropological discussions of indexicality, entextualization and contextualization, and metapragmatics, I revisit the discussion of Christian Metz and others of “impersonal enunciation” in light of my own ethnographic studies of the Tamil cinema of South India. I show that doing so draws out the way that an ethnographic attention to enunciation, rather than lead “into” the text, spills outwards into events of cinematic semiosis. Such a shift in orientation demonstrates how the (im)personal nature of cinematic enunciation is not a medium feature of cinema but a situated, semiotic achievement; it is an empirical question and thus deeply political in nature. I conclude with methodological reflections for the semiotic study of cinema.
Keywords: enunciation; indexicality; entextualization; cinema; linguistic anthropology; continental semiotics; Tamil Nadu, India
Preamble (In the Margin of Older Works on Enunciation)
This article is an exploration of the enunciation of images as the result, or outcome, of processes of entextualization and contextualization; in particular, it explores the actness of filmic images in their cinematic contexts. I am especially interested in Christian Metz’s (1995[1987]; 2016[1991]) discussion of impersonal enunciation—as introduced in his critical engagement with Francesco Casetti’s (1995[1983], 1998[1986]) “deictic” account of cinematic enunciation—in relationship to my own ethnographic studies of the Tamil cinema of South India (Nakassis 2023b). This exploration provides occasion to elaborate thoughts on how we might critically think North American linguistic anthropology and continental semiotics of the “Paris School” (Parret 1989) together, in particular, by linking the notion of enunciation (in continental semiotics) with the notion of entextualization (in linguistic anthropology).1
To do so, we might start from a beginning: those two great masters of their generation, Roman Jakobson and Émile Benveniste, who, in elaborating on linguistic structuralism, also pointed out its limitations and thus opened paths beyond them. For Benveniste (1971, 2014[1970]), one key site to synthesize and break through particular aporia in linguistic structuralism was the concept of enunciation, the process by which a virtual “system” (langue in the case of natural languages) is appropriated—instantiated and realized—in an act of discourse (Fontanille 2006[1998]). As Benveniste pointed out, this process leaves particular traces in the énoncé (the enunciated utterance, the precipitated text), which is why we can study this process by inspecting discursive form in vitro. Moreover, such a process is itself presupposed by langue through particular form classes that we find in every natural language, namely, deictics (such as personal pronouns, demonstratives, tense, evidentials, among others). Such form classes are an interface between the virtual and the actual, langue and parole, that is already part of the virtual infrastructure of language (i.e., it is the virtualization of the actuality of discourse, and the way in which that virtuality is, in turn, realized; cf. Bühler 1990[1934]; Hanks 1990, 1992, 2005; Edwards 2018; Edwards and Brentari 2021). Benveniste further suggested that enunciation was the way in which the speaking subject was constituted in and by language and thus how that subject acted through speech: namely, that by enunciating the “I” of langue (the grammatical person)—in relation to a reciprocal “you”—the “I” of discourse (the speaking subject) is entailed (cf. Urban 1989; Paolucci 2020, 2021, 2022).
It is important, I think, to see Benveniste’s comments on enunciation in relationship to Roman Jakobson’s own comments on deixis and the communicative speech event.2 Following Jespersen (1922:123–24), Jakobson (1981[1957]) referred to such token-reflexive denotational indexicals as shifters, because their referent “shifts” across events of usage. (The translation in French, embrayeurs, refers to the gear mechanism—the “stick[-shift]”—in a standard-transmission automobile, which shifts in, and out, of gear.) One of Jakobson’s key insights was that such signs could be productively characterized through the semiotic of the American philosopher and logician, Charles Sanders Peirce.3 Shifters are “indexical symbols.” As symbols (and thus legisigns, or types) their ground turns on a stipulative conventionality, which involves a minimal, transcontextual semantic schema (hence part of langue, as Benveniste noted); yet, as indexes, this semantic schema depends on an “existential relation,” or indexical ground, between a sign-token (or sinsign) and its object (Hanks 1992); in particular, for shifters, as Jakobson showed, this anchoring, or contextualizing, relation between sign and object involves pointing to (“indexing”) elements (or “factors”) of the communicative event in which the sign is used.4 Shifters are not simply indexical, then; they are metapragmatic (Silverstein 1976, 1993; Lee 1997) in that they take as their object the pragmatics (i.e., the indexicality) of sign activity (each such metapragmatics being one of Jakobson’s speech-event “functions”). As such, signs of enunciation—indeed, enunciation itself—are metapragmatic precisely because they reflexively concern—they represent, mediate, regiment—the pragmatics, the doing, the actness of token-instantiations of semiosis.5
It is on this nexus of concepts—enunciation, shifters, deixis, the speech event, and indexicality—that the path of North American linguistic anthropology (following Jakobson) and continental Paris School semiotics (following Benveniste) split, a fork in the road that has led to different itineraries, if also, through these travels, possible common destinations. A reunion, however, requires us to return to that branching point and follow where each path led, and where it might lead in the future. How might we describe this problem space, these respective paths (not) taken?6
Paris School Semiotics: From the Virtual to the Actual
Following the trail of enunciation, early advances in Greimassian semiotics treated enunciation as something to be reconstructed from texts (primarily, literary narratives and later, visual artistic texts; see Logopoulos et al. 2025a for an overview), not as something to be observed in a specific context by the semiotician in vivo, but rather analyzed post facto as a system of relations. In doing so, this approach powerfully and in great detail was able to model the virtual figuration of actness, the way in which texts disclose and model their own actness in their (generalized) textual (narrative) form. As a result, while enunciation involves the conversion of a decontextualized virtual schema into a realized, contextualized instance, this brand of structuralist semiotics—at least, in its “classical” formulation (Dondero 2025a:228)—initially favored focus on one half of the equation: the virtual over the actual, intensional over extensional, text over context; and in particular, on what is “in”—and generatively “under”—the text that makes this realization/extensionalization/ contextualization possible.7
We find this tendency in Benveniste’s writings as well, which, once they hit upon the fact of enunciation (i.e., indexicality) in language, concentrate attention on parts of the linguistic code that presuppose acts of discourse rather than discourse itself. Benveniste never really focuses on any actual examples of discourse as such, that is, actual embodied interaction involving such signs. Recognizing communication ‘out there,’ he remains at the level of the “code” (and written text). This tendency similarly manifests in a certain antipathy, by Benveniste and others (see Landowsi 1989[1983] and Latour 1999 for discussion), to the notions of “communication” and “context.” Both terms, not without reason, have been critically seen to embody Enlightenment ideologies that would reduce language to an instrumental, mechanical "tool" of “communication”—where “information” in messages is “sent” back and forth between a “sender” and a “receiver”8—and that problematically differentiates word (“text”) and world (“context”).9 Given this, between enunciation and communication, text and context, virtual and actual, “classical” Paris School semiotics has tended to give priority to the former, most fully exploring its semiotics so as to explain—and at times explain away—the latter.
Yet by exteriorizing communication and the event of speech (its situation, its context) as something “outside” the code and text, and thus by implication not proper to the semiotic study of enunciation—which is “inside” the text—in moving beyond linguistic structuralism such work has conserved its most foundational gesture, arguably neutralizing the productive advance of the very concept of enunciation: the indexicality in/of language. This results, in my view, in a certain tension and imbalance within the very notion of enunciation, as an account of the actness of semiosis that, at least historically, starts from a position—theoretical and methodological—which places empirical acts of semiosis just beyond its methodological and analytic grasp, as “extra-semiotic” and thus as the purview of other fields (e.g., of anthropology or sociology, as we will see with Metz; also see, e.g., Bertrand 1989[1984]:107, 124, 128, 130).
This has not been without criticism and expansion from within the continental tradition, of course. More recent models of enunciation in Paris School semiotics have taken more seriously the notion of “instances of discourse” (Fontanille 2006[1998]; also see Latour 1999; Paolucci 2020, 2021, 2022), making space for the importance of “reception” and the act of “production” (Fontanille 1998; Fontanille and Dondero 2014), as well as elaborating a semiotics of practice (Dondero 2025a)10 and enunciative praxis (Bertrand 1993, 2000; Fontanille 2017).11 It is here that I see a potential convergence with the semiotics of contemporary linguistic anthropology, which in taking the indexicality of semiosis (and all that this implies) as an analytic anchor, methodologically grounds itself in and departs from what such recent developments in semiotics have tended towards (viz. instances of discourse, practices, events).
Linguistic Anthropology: From the Actual to the Virtual
To see this, consider the other path taken. Post-war linguistic anthropology in North America was profoundly transformed by Jakobson’s arrival (see Silverstein 2017), in particular, through various reworkings of his speech-event model by those he influenced, such as Dell Hymes (1995[1962]) and, later, Michael Silverstein (1976). In these uptakes, the speech-event model was not seen as grounded by the referentialism of informatic models (even if such referentialism lingered); just the opposite: it was seen as offering a critique of the referential function by provincializing it. From this vantage, “reference” and the “transmission” of a “message” (i.e., a text) between “senders” and “receivers” simply becomes one of the many functions of a text, which is intrinsically multifunctional. Silverstein (1976, 2003, 2004, 2022) pushed this deconstruction one step further by suggesting that Jakobson’s speech-event model—despite its problematic visual semiotics, which “positions”/enunciates its viewer (Fontanille 1989) as nowhere in the model but hovering above it—should instead be taken as inviting us to take the “sign’s eye view” (Silverstein 2004:631n11, 2022:23, 27), that is, that we should see each “factor” (speaker, addressee, code, channel, message, context, etc.) as something that can be indexed (‘pointed to’) by the sign/text. The speech-event model thus becomes a model of indexicality, and each “function” a metapragmatic function that regiments some aspect of the message’s indexicality.
As such, indexicality becomes the lynchpin to study language/discourse, requiring us to start from the way in which signs in particular events invoke—indeed, creatively constitute—the “contexts” (virtual and actual) with which they are indexically entangled. The questions become: What kinds of indexicalities are there? (Many more than the paltry six in Jakobson’s static model, it turns out!) How do they function and what metapragmatically mediates them? How do they interact, indeed, cohere into textures that “dynamically figurate” their own pragmatics (i.e., their own enunciation)? That is, how does the entextualization, the real-time flow of emerging token-signs that are (indexically) juxtaposed with each other, come to (iconically) cohere in some construable/iterable type, forming a patterned “signifying whole” (to use Fontanille’s [2006(1998)] phrase), a text, that is thus bound off from what comes to count as its “context,” as formed through the co-ordinate process of contextualization (Hanks 1989; Briggs and Bauman 1990; Silverstein and Urban 1996)? Put otherwise, what process constitutes an event, a text such that it could index, that it could presuppose or entail, a co(n)text as a condition of its being and meaning? Here, note that on this understanding text and context are not static entities. Rather, they are the precipitates of a process of semiosis; which is to say, they are co-ordinate and co-constitutive of each other, grounded by relations of indexicality, and thus cannot be studied independently of each other. Context, thus, is not an “extra-semiotic” “space” or “container” (wherein “reception” or “production” takes place) but a constitutive feature projected and entangled (as “external”) by semiosis itself.
Yet if linguistic anthropology initially gave much attention to the speech event and (realized) face-to-face interaction (the so-called “micro”), it has often been critiqued as doing so at the expense of, or by unproblematically assuming, encompassing orders of (virtual) mediation (the so-called “macro”); this critique in recent years has led to advances in thinking “semiosis beyond the speech event” (Agha and Wortham 2005; Agha 2007) and to posing the problem of scale as itself semiotic (Carr and Lempert 2016). Here, scholars have increasingly asked, how do processes of entextualization/contextualization give rise to virtual regularities of semiosis (intertextuality, register/enregisterment, genre, scale/scaling, grammatical code, semiotic/language ideology, culture, etc.) that themselves, in dialectical turn, mediate semiotic events in their unfolding? What is the movement from (discursive) instance to (text) type and back again (norm to trope, de-/re-contextualization) as it is reworked and transformed across contexts as the very warp and weft of historical change (viz. as an enunciative praxis)?
In sum—and recognizing the exaggerated, partial, and overly simplified, nature of my overview and comparison—if French semiotics from Benveniste onwards initially moved from the study of deixis to (literary) enunciation and thus from the actual to the virtual (and more recently, towards a semiotics of practices and enunciative praxis, that is, from the virtual back to the actual)—and if it has done so by bracketing the speech event and the problem of indexicality—post-war linguistic anthropology has moved in a counter-ambulation, moving from deixis to the problem of indexicality and the speech event (viz. the problem of entextualization/contextualization), and then more concertedly towards virtual, interdiscursive orders of semiosis. Echoing the theme of this issue, my aim in this article is to close the dialectical loop of realization and virtualization, as differently explored in each tradition, with the hope of creating bridges between them.
Impersonal Cinematic Enunciation
In order to do so, I want to look at how the concept of enunciation was developed by Christian Metz in the study of film, offering a critical rethinking of his account via some concepts from linguistic anthropology, as worked through my own ethnographic research on South Indian cinema.
Bearing out the observations above, foundational work on cinematic enunciation starting in the 1980s (e.g., Casetti 1995[1983], 1998[1986]; Metz 1995[1987], 2016[1991]; Gaudreault 2009[1999]; also see Branigan 1984), used the concept of enunciation as a way to analyze film beyond the impasses of cultural structuralism (cf. Metz 1974[1971], 1991[1971]), yet remained focused narrowly on the text itself, not on film as act per se but on the question of how film reflexively figurates and indexes its actness. While these authors revealed how films enunciate subject “positions” of various sorts (qua narrators, spectators, etc.), such positions remained abstract and virtual (a function of the system of the text in question), exteriorizing the problem of how “instances of incarnation” (Metz 1995[1987]:151)—actual “bodies” (i.e., viewers) in events of “communication”—relate to/inhabit such subjectivities. Indeed, for Metz, the study of actual subjects in actual events of filmic activity (production, reception, etc.) is one that is not proper to semiotics—which is a study of the text, he insists—but other disciplines like sociology.
Of particular interest in the writings of Metz and these film scholars is the way in which they rethink Benveniste’s linkage of subjectivity and enunciation, asking: when film is enunciated, who “speaks” (to whom and) in what capacity? That is, who or what is the subject of film? Here, Benveniste’s (1971:206) distinction of histoire and discours and its relationship to deixis became important (e.g., Metz 1982[1975]). Histoire, for Benveniste, is a narrative énoncé (utterance, enunciated text) that effaces its enunciation (e.g., by suppressing personal pronouns, adverbs, by using particular tenses, etc.; cf. Agha 2007:42–45 on “deictic selectivity”). By contrast, discours precipitates a dialogic (face-to-face, oral) énoncé, a communication which explicitly/transparently bares its enunciation through localizable, segmentable, inherently referential deictic forms (like, “I” and “you”).12 For Metz, the difference between film and speech is precisely analogous to the difference of histoire and discours, which is to say, film is a form of histoire that lacks that which makes speech discours: its co-present, real-time, (inter)personal dialogism. While subjectivity in discursive interaction, Metz supposes with Benveniste, is entailed by the “exchange” of personal pronouns, in film—as histoire—there is no such exchange (the you/spectator cannot respond to the I, presumably) and, further and consequently, there is no I as such (for the maker is only present to the spectator in/as the text, enunciated in their absence). Or rather, the “I” is not a person, but the impersonal machinic film itself. Further, as Metz suggests, the subject enunciated by the énoncé in the cinema—a film text—is the spectator (Metz 1995[1987]:150), a third-person to the film text. As Metz writes (in the context of discussing the work of Bettini and Casetti): “the true têtê-à-têtê ... does not happen between an enunciator and an addressee, but between an enunciator and an utterance, between a spectator and a film” (ibid.; original emphasis). Metz goes on in the next paragraph to argue that the only body that could incarnate the roles of enunciation are the spectator and the film itself, “the body of the text, that is, a thing, which will never be an I.” He writes: “The film is the enunciator, the film as a source, acting as such, oriented as such, the film as activity” (ibid.; original emphasis), an activity of enunciating itself.
This kind of enunciation does not reveal itself, moreover, through deixis (cf. Casetti 1998[1986]).13 Rather, it is metadiscursive: instead of reaching out to its communicative context, traces of enunciation in the film text reflexively fold back in on themselves, thereby diagramming (rather than denoting) their enunciation as “internal” to themselves.14 This makes film “anaphoric”: its deictic gestures are not “true,” they are “simulations” of deixis, involuted into the sealed body of the text,15 in contrast to the putatively authentic home of language, discours (“oral communication”) with its exophoric, “real” deixis.
Finally, despite his statements that attention to actual events of cinematic semiosis (“reception” by viewers or “production” by filmmakers) are beyond the purview of the semiotician, who is only concerned with the text and “the general, or rather, the generic truth” of “THE spectator,” Metz repeatedly argues for the impersonality of cinematic enunciation by appealing to viewers.16 (This is the result, I would suggest, of the internal tension to the notion of enunciation discussed above, manifested by the haunting figure of the “empirical spectator,” a point I return to in the conclusions.) Directly after Metz’s quote about the “the true têtê-à-têtê” above, he adds, “It is also the common feeling ... that the ‘subject’ is the spectator.” Similarly, in the next paragraph, after writing that “The film is the enunciator” Metz writes, with no supplementary qualification (no “also”), “This is how people think: what the spectator faces, what he has to deal with, is the film.”17
Each of these arguments about the cinema (that it is “monodirectional”; that there is no exchange across the screen; that cinema is founded on absence and not presence; that spectators feel/think cinema to be so) can, and must be, problematized on empirical grounds, as I do in the sections that follow (see also Nakassis 2016, 2020, 2023b). But they also are problematic for the theoretical and methodological effects they entail. Notice how the way Metz poses the problem produces an alibi, a justification, for limiting the study of cinematic enunciation to the film text in “itself” (as disclosed to the analyst as spectator). Yet by cleaving enunciation from communication, semiotics from empirical social science, text from context, “universal” textual processes (“the general, or rather, the generic truth” of “THE spectator”) from sociologically or culturally particular ones, Metz makes film semiotics unable to account for the indexicality of enunciation, for the actness of images, their capacity to performatively “shift-in” in empirical contexts of happening to performative effect.18 Moreover, it makes the semiotician unable to account for the fact—in fact, even to properly appreciate—that cinema is not the same everywhere, that is, is not all like Euro-American narrative film as watched by mid-century, bourgeois viewers (Hansen 1991). Do, indeed, all spectators think as Metz says? And are all (narrative, fictional) films “monodirectional,” unresponsive to their audiences (and vice versa)? And how would we know? That is, can it be known just by the analyst looking at film form per se? And if, instead, we must follow enunciation not “into” the text but instead look to the processes in which film texts unfold in some evenemential, social “context,” might we end up with a different view on the whats and whens of (im)personal enunciation?
In asking these questions, I aim to draw the study of enunciation closer to the semiotics of linguistic anthropology, bringing enunciation back to its indexical roots and thus to the study of actual events of cinematic semiosis, not as “incarnations” of pregiven virtualities, but as the cradle from which the virtual continually emerges and manifests. Overall, I see this as consonant, though in slight tension, with developments within continental semiotics that have taken up and generalized Metz’s insights (e.g., Bertrand 1993; Latour 1999; Paolucci 2020, 2021, 2022). Such developments have variously argued that impersonal enunciation is logically (Bertrand 1993), ontogenetically (Paolucci 2022), or evolutionarily or cosmologically (Latour 1999) prior.19
Here, Claudio Paolucci’s (e.g., 2020, 2021, 2022) recent writing on the “illeity” of enunciation is closest to the view I develop in what follows. Paolucci notes that the so-called “third-person” (or non-person as Benveniste put it)—as a grammatical category—is not opposed to the personal pronouns but rather encompasses them: it is thus not an opposition between [I-you] and [he/she/they/it] but rather a differentiation of [[I] and I+not-I]. On this view, subjectivity (and reference) is indeterminate independent of some process, some event within which subjectivities (an I, a you, a they) are assigned and distributed.20 This process is enunciation, and it is “impersonal” in the sense that the il (literally, 'he' in French; i.e., the non-person, the encompassing category of I+not-I, or ‘he/she/it/they’) has “primacy” over the “I.” Paolucci’s concerns are primarily, as I take it, about the question of subjectivity and its genesis. What is important for our discussion is the conclusion that we remain agnostic to whether enunciation is “personal” (sensu Benveniste) or not, but rather see this very distinction as the emergent effect or outcome of “events” (Paolucci 2022:52), of an “illeity process” (ibid.) of semiosis.
What is the structure of such a processual event? And when we empirically turn to look at such events of semiosis what do we find? We find, I suggest, indexical processes of entextualization and contextualization; we find the personalization and impersonalization of enunciation. But what’s more, as we see in what follows, we find that this very question of the im/personality of enunciation is itself at issue in such “illeity” events; that is, there is a politics of enunciation internal to enunciation that concerns the very question of its im/personality. There is an internal fold, or reflexivity, thus, that wrinkles any simple answer to the question of the im/personality of enunciation. Put otherwise, if events of enunciation involve the very question, Is enunciation (im)personal? (via the coordinate question, is this énoncé im/personal?), and if semiosis is only ever encountered with this question in play—because this question itself propels semiosis, driving it, serving as its condition of possibility (thus standing behind, and mediating, the philosophical conclusion that enunciation is impersonal)—then this politics is lodged within, and thus “before” in a sense, that very conclusion.
Here, thus, I would resist a tendency in the above discussed works (from Benveniste’s to Metz’s to Paolucci’s) to pose the problem of enunciation as either/or: that enunciation is one or the other (personal or impersonal); that one is first (impersonal) and the other second (personal); that one is primary and the other secondary or derivative (in terms of importance or function). Rather, I would suggest that we see impersonal/personal enunciation as co-ordinate processes, as co-constitutive dimensions of semiosis that, in specific contexts, are configured and realized in variable ways to variable effects.21 It is this variability that we must first attend to, and from which we may then generalize. For me, thus, the issue isn’t whether enunciation in film is or is not impersonal, but how we pose and pursue that very question as something to empirically study; indeed, as something to study because when we look at what the “illeity” of enunciation consists of we find a reflexive tussle over the becoming of enunciation. To theorize this requires that we move beyond abstract models or decontextualized texts (sentences like “It is raining” or song lyrics or film texts) into empirical processes of contextualized semiosis.
To do this, in what follows I reconsider cinematic enunciation from the question of the way in which images may be constituted as enunciative acts in their contexts of happening through the dialectic movement of virtualization/realization, entextualization/contextualization. I do this in two interlinked ways, one methodological and one theoretical.
Methodologically, I rethink enunciation through a consideration of an ethnographic example from my fieldwork on the Tamil cinema of South India, instead of only “reading” textual examples from a familiar canon. In particular, I want to consider a cinematic context, and a particular film, within which audiences threaten to hold the director (and his body, not just the film) responsible, personally, for the film they took him as enunciating as a particular kind of performative act of discourse (in the instance, of caste insult to them), despite his attempts to impersonalize his narrative as a mere histoire.
Theoretically, I situate enunciation in relation to linguistic anthropology’s discussion of indexicality, metapragmatics, and entextualization, as well as Goffman’s (1974, 1981) deconstruction of the speaker and hearer roles in discursive interaction.22 I argue that rather than leading us into the (storied) filmic text—as seemingly sui generis and autonomous, impersonal and machine-like (as Metz and others would have it)—the concept of enunciation, revisited and rethought ethnographically via the legacy of Jakobson in linguistic anthropology, can lead us “outwards” into events of cinematic semiosis and their social contexts and effects; or rather, it can help us deconstruct this very binary of inner and outer, text and context by making the dialectic process of their very formation the object of our analysis.
Ultimately, thus, I suggest that the impersonality or personality of enunciation is the outcome of semiotic processes in and by which enunciation is itself enunciated—that is, entextualized and contextualized—, achieved as an enunciation in some such way (viz. as personal, impersonal, some blend of the two, etc.), which is to say, these are the precipitates of dynamic processes of the personalization or impersonalization of enunciation. As the outcomes of empirical processes of entextualization/contextualization, such achievements are tenuous, perspectival, negotiated, contested; which is to say that the (im)personalization of enunciation is a political issue that is constitutive of semiosis precisely because it is one of its engines—it is what parties to semiosis are often trying to figure out, to fix, control, or avoid, and thus cannot be taken as pre-given or theoretically determinable outside of other enunciations, or put otherwise, outside of some social co(n)text. It is not just the “(personal) subject” that has to be enunciated by an “impersonal” enunciation; the very impersonality of enunciation itself has to be achieved; sometimes, through its personalization, as we’ll see.
What kind of semiotics—theory and method—do we need to account for this?
A Personal Enunciation
To think through these questions, I turn to the 2004 Tamil film, Kaadhal (‘Love’; dir. Balaji Sakthivel), a small-budget movie that was a surprise hit in South India at the time; a surprise because, unlike the blockbuster, star-centered films that dominate the Tamil film industry (see Nakassis 2023b), Kaadhal was a story-centric, realist film that lacked any known actors and was framed, in the press and by the film itself, as based on a true story (Figure 1).23
Figure 1. The opening frame of Kaadhal (2004; dir. Balaji Sakthivel): ‘This … a screenplay that was created based on a true incident’ (“Itu… uṇmaic cambavattai ātāramākakkoṇṭu uruvākkappāṭṭa tiraikkatai).
The film takes place in Tamil Nadu, in the southern city of Madurai, wherein romantic love blossoms between the film’s protagonist Murugan—a poor, Dalit (formerly “untouchable”) mechanic—and Aiswarya, his middle-class Thevar lover, a schoolgirl from a so-called “dominant” mid-level caste. The lovers elope to the metropolitan city of Chennai to marry but before they can consummate their marriage, they are discovered by the girl’s paternal uncle. Aiswarya and Murugan return only to be violently separated, each savagely beaten and cursed with calls from Aiswarya’s frenzied family members to kill them both. While questions of caste in the rest of the film run just under its surface, in this wrenching climax sequence they erupt into explicitness. As her father repeatedly smashes Murugan’s head onto a large rock, Aiswarya cries out that she’ll do whatever he asks (Figure 2). He demands that she rip off her tāli (a necklace tied by the groom during the wedding to effectuate marriage) and repudiate the union. About to smash in Murugan’s head with a large stone, she acquiesces.
Figure 2. Aiswarya yielding to her father, as he savagely beats Murugan
Years later, in the neighboring town of Dindigul, we see Aiswarya riding on the back of a motorcycle with her new husband (of the same caste, as inferred by viewers with whom I spoke) and a small child. At an intersection, she looks over and sees a beggar (Figure 3).
Figure 3. Denouement of Kaadhal: Aiswarya recognizes a crazy beggar, her ex-husband Murugan.
She recognizes the tattoo of her name on his chest. Murugan. She faints. Awakens in a hospital, crying. Wildly yelling Murugan’s name in the street, she looks for him. Finding him, she falls to the ground, wailing and hitting her head in lament: ‘I made a mistake, I thought you’d forget about me; I lost you to protect you; you’ve become crazy because of my mistake! I’ll never leave you again!’ (Figure 4—left) Murugan doesn’t recognize her. Behind her, the husband has come outside with their baby (Figure 4—right).
Figure 4. Left: Aiswarya falling at Murugan’s feet; right: her husband finding her in the street
Aiswarya hears the baby and cries out, recognizing the irreversibility of her fate. The stoic husband betrays no anger. He walks toward her and pats the baby on the back, pointing to it, that it needs her. He hands the baby to her, then softly touches her head. He turns to Murugan and brings him around, and with both of them on his arms begins to walk down the street. Here, the narrative proper ends as a rolling text appears, echoing and elaborating on the opening frame’s claim that the film was based on a true story (Figure 5).
Figure 5. The rolling text in the final frames of Kaadhal (dir. Balaji Sakthivel)
Events of violence surrounding inter-caste union are common in South India—and it is not unusual that they end in the murder of both man and woman, so-called honor killings, by the girl’s family (Srinivasan forthcoming). This is particularly the case when the boy is Dalit and the girl is from a land-owning higher caste such as the Thevars, as in the Madurai region.
The realism of Kaadhal, thus, for many viewers with whom I spoke, was how the film captured the acute anxieties and real-life experiences of violence surrounding inter-caste romance. Indeed, as the director Balaji Sakthivel told me in a 2008 interview, Kaadhal was inspired by one such event that happened in his hometown of Dindigul when he was a teenager. Walking by a home in his neighborhood, he overheard a conversation between a young Thevar woman and her mother; the daughter hesitantly asked her mother what would happen if she loved a man outside their caste; the mother, without hesitation and to the young to-be-director’s horror, frankly said she would chop her to pieces. Weeks later, the boy with whom the girl was in love (and which her hypothetical question was about) disappeared. This is how fanatical (jāti veri) Thevars are about caste, he said.
Yet this was not the event that Kaadhal cites in its conclusion as the “true story” that it claimed to be based on. In the rolling text, the film cites the experience of someone else, framing itself as a cinematic version of a railway story told to the director by the husband, itself putatively a historical event that happened. Yet, that story, as Balaji Sakthivel revealed to me in a 2008 interview, was made up. The director never met such a man. It was simply another “script,” a lie (poy) as he put it, one he was publicly revealing for the first time in our interview. But why lie given the horribly common, and instantly recognizable, nature of such events of inter-caste love and its violent outcomes? Why wasn’t one “script” enough?
As he framed it to me, Balaji Sakthivel intended Kaadhal as a critique of caste, to show the hypocrisy of its patriarchal misogyny, to show the cruelty of caste as it really is, something hero-oriented commercial films, he implied, fail to do; but more than critique the painful realities of caste, he also wanted to show the possibility of humanity, to show that violence was not the only response to challenges to caste purity and that, indeed, like the second husband in the film (and at the railway station), one could accept that women love before marriage, and even across caste lines. But if so, why the rolling text at the end of the film, with its reflexive linguistic enunciation of the “I” of the director in direct address to the “you” of the audience? Why the need to lie and frame the story as historical fact, as another man’s story when the fates of inter-caste romance are known to all? Indeed, the rolling text wasn’t part of the original script, only added after the film was shot, just a week before its release.
As Balaji Sakthivel narrated it, he was afraid (bayam). He wasn’t courageous (tairiyam) enough. Would the audience accept the film’s ending as he originally wrote it, without any such framing text laid over as the husband, his wife, and her ex-husband walk down the street, arms entwined? Or would they have felt that he botched the climax (climax-le sotappiṭṭān ṭā), dismissing the film, and its caste critique, as ending with a fantasy, with a totally unbelievable character? Would the “male chauvinistic mind” of the audience, he said, a mindset consumed by caste (jāti) and race (inam) and the purity of its women, hate the film as a whole, disidentifying with it and asking in disgusted disbelief, ‘What man would do that?’ Such a viewer, he ventriloquated, would not: ‘I would never eat the half-eaten leftovers (literally, ‘spit food’) of someone else’ (innoruttan sāppiṭṭa ecci sāppāṭu sāppiṭamāṭṭēn). As the director noted, such a viewer would feel revulsion and would reject this possibility (in the film) despite, or precisely because of, the fact that such things are common in everyday life.
Balaji Sakthivel’s worry was not simply one that he projected outward to an imagined spectator, however, an empty “position” waiting to be incarnated by some “body” on the other end of the future screen, absent to him. It was first and foremost a worry that was based in the reception of and communication with an embodied, empirical viewer: himself. Habituated to caste discrimination, its seed planted in his mind by his own family, the director recounted that when he watched the first cut of his own film he felt uneasy (neruṭal) with its ending. His years of immersion in a casteist society raised a resistant voice in him that he feared would be there for other viewers. To answer that voice, he put the rolling text, a ‘fictional story’ (karpanai katai) whose function was thus to force the film’s viewers to maintain identification with the film, to force the viewer not to dismiss the unreality of the husband character but instead to accept his existence, and perhaps even embrace his own enlightened attitude as their own.
At the same time, the rolling text was also a “disclaimer,” akin to, as Balaji Sakthivel put it, the fine print at the end of a newspaper put to avoid defamation suits (māna naṣṭa valakku; literally, ‘honor-loss/destruction case’). But if it was a disclaimer, what was it disclaiming? And if there was defamation, whose honor was defamed? The enunciative disclaimer was, of course, his own, this “I” of the narration. What was disclaimed was that the opinion of the film, its caste critique, was not his. It was someone else’s story. And it really happened. And the defamation being disclaimed was that of the film’s audience members. The rolling text, in effect, said “it is not I showing you a man from your caste (or a caste like yours) accepting ‘half-eaten, saliva-drenched food (i.e., a spoiled woman) from a Dalit.’ It is not I preaching to you, it is not I criticizing you, it’s not I defaming you.”
Note, then, that while the husband accepting Aiswarya is framed by the film and its director as an admirable act of humanity and even of ‘sacrifice,’ this very framing “within” the text already presumes upon a competing enunciation outside of it (that is, indexically projected by it): namely, that accepting a soiled wife is a humiliating dishonor, not only to the fictional character but also to his (represented) caste (the Thevars), and further, to the viewer who would identify with either. What is being disclaimed through the enunciated veil of reality, then, is a representation that would cinematically cuckold the Thevar caste and castes like it; what is being disclaimed is a personal enunciative insult by the director to the upper-caste male viewer who feels himself negatively interpellated by the film as a whole (i.e., the issue isn’t reducible to the linguistic deictics that appear at the edge of the film text). Such an interpellation would undermine the film’s political message to abandon caste fanaticism and its patriarchal hypocrisies by eliciting both. The film’s personal enunciative framing as historical truth, then, invites the spectator to take the film up as an authorless, impersonal representation for them rather than a personal, performative act of offense directed at them by a personal enunciation issued by the director; that is, it invites us to entextualize the film as histoire rather than discours.
To do this, the director had, importantly, to contort his critique of caste to a casteist vision of the world. Indeed, this fear of an uptake of the film as an image-act of caste insult didn’t just necessitate the (re)framing of the film in its first and last images; it infused the very thread of the film’s narrative. Hence, at every juncture where the lovers are afforded a moment of sexual intimacy, their physical union is frustrated. When Aiswarya and Murugan reach Chennai, with nowhere to go, they stop by a “mansion” (or hostel) for bachelors where Stephen, a friend of Murugan, stays in a room with six other men. While they cannot stay with Stephen—women are not allowed in such mansions—they manage to use his room and the public bathroom to shower when his roommates are out for the day. With the lovers finally alone in a private, conjugal space, the scene breaks into a conventional, symbolic song-and-dance sequence, “Kiru Kiru,” that depicts the lovers’ sexual desire as they dance across a series of extra-diegetic locations (Gopal 2011). The song ends with a mansion-resident discovering Aiswarya’s bindi stuck to the wall of the bathroom stall where she had showered. Realizing there’s a girl in the mansion, he gathers other residents in the mansion to investigate. Returning to the diegesis, the camera cuts to a shot of Aiswarya and Murugan in the room, just about to kiss on the mouth yet interrupted by a knock at the door. Later, the lovers attempt to get a registered marriage but are unable to because they lack a place of residence, and so try to find an apartment to rent. By the end of the day, they have again failed to secure a private space of conjugality. To pass the night, Stephen recommends that they go to a local theater to watch the last two films of the night, and then take a bus to Tindivanam (a town about 130 km south of Chennai) and back.24 As it turns out, Aiswarya is having her menstrual period. After a night riding the bus, the next day the couple manages to rent an apartment of their own and conduct their marriage. Yet on the morning of the first day in their new conjugal space, Aiswarya’s uncle has discovered their whereabouts and arrives to take them back to Madurai.
While in 2018, Balaji Sakthivel explained the purpose of his “disclaimer” as securing spectatorial identification, in 2008, when he admitted to me his “script,” he framed it as a problem of filmic reference, saying, in a stretch of discourse wracked with hesitations, hedges, codeswitches, and ellipses (original in left column; translation in right column):
The first time I’m admitting, in this interview, but uh, it’s a kind of script I think, because nobody, no male- it’s a male dominated society, nobody will admit it. ((No one)) touch paṇṇinā- avaḷe vantu touch paṇṇ((inā)) they won’t ((release)) it. Nān vantu anta paṭattule avaṅkaḷukku sex varrā mātiri, appaṭiyē atilē
some limit was there. So, that’s also not such a reality Kaadhal, as far as my humble knowledge is concerned. The first time I’m admitting, in this interview, but uh, it’s a kind of script I think, because nobody, no male- it’s a male dominated society, nobody will admit it. If ((no one)) touched- if he touched her, they won’t ((release)) it [the movie]. In the film, it was like there was about to be sex, in that way in that
some limit was there. So, that’s also not such a reality (in) Kaadhal, as far as my humble knowledge is concerned.
Here, thus, the issue is not just one of spectatorial identification, nor even about the director’s capacity to represent the act of sex between a Dalit boy and a virginal Thevar girl (be it directly or through euphemism). It is the capacity to narratively emplot such a sexual union as part of the story, that is, to serve as a referent in the fabula. In real life, he admitted, they would have had sex. But in reel life, they couldn’t.
Notice how the performativity emanating from the very thing that is the basis of the realist conceit of the film—that it will reveal the truth of caste and sexuality—infects the representationality of the filmic image and its narrative structure, breathing life into both as the avoidance of that very performative force. “Nān tavirttuṭṭēn” (‘I excluded/avoided it’), Balaji Sakthivel said in 2008. The semantic range of the verb he used, tavir, includes the English senses ‘to expel, exclude, interrupt, prevent, frustrate, control, restrain.’ The interruption of the kiss, the bureaucratic hurdles to their registration, the inability to rent a home, Aiswarya’s menstrual period, her family—the narrative structure of the second half of the film blocks, frustrates, expels, and restrains Aiswarya’s sexuality (Anand 2005; Rajangam 2016). The representation of female sexuality harbors a performative force, and its intersection with caste renders this unmanageable and, in the end, effaced from the diegesis itself.
The repercussions of such a representation are not imagined and were certainly part of the director’s hesitations. Cheran’s 1997 Bharati Kannamma, for example, a story about a Dalit (Pallar) man in love with his Thevar landlord employer’s daughter, triggered protests and attacks on movie theaters and buses from Thevar groups for the way in which it was taken to dishonor their caste (Nambath 2003; Rajangam 2016). Kaadhal avoided such violent protests by, in the end, protecting the chastity of the dominant castes that it attempted to critique for their patriarchal hypocrisy and casteist violence.
Enunciated in this text (like all other texts) is thus already a history of enunciations—indeed, an enunciative praxis—one which staves off what it ironically already prefigures and enables. Consider, for example, the not exceptional recontextualization of Kaadhal in the south of Tamil Nadu, as reported by the journalist and author S. Anand (2005). When Murugan was beaten in the climax of the film, S. Anand relays:
An aspiring filmmaker friend who watched Kaadhal in a Madurai cinema talked of how Thevars—the dominant ‘backward caste’ of the southern districts—in the hall shouted aloud: ‘Fuckers, this will be your fate if you think you can get our girl.’ Dalits watching the movie in the southern districts were intimidated both by the depiction of the hero and by the participative enthusiasm of the Thevars among the audience.
And later, in the final scene, when Aiswarya “spots Murugan as a mad man at a traffic signal” (Figure 3), he reports: “This is where some young Thevars in the audience shout in Madurai’s cinemas: ‘Keep off our women.’” In such moments, the film is entextualized not as realist representation critical of caste violence but as an act of caste violence itself. Contextualized by these metapragmatic utterances (‘Fuckers, this will be your fate if you think you can get our girl,’ ‘Keep off our women’), the film image is indexically projected from the screen into the timespace of screening and onto the bodies of co-presence Dalit viewers. In this violent, shifted-in act of uptake, an impersonal realist representation is ontically transformed into—indeed, because it has already given itself over to becoming (as a potentialization of)—a personal, performative act of intimidation and threat, precisely against but also through the simultaneous entextualization of caste critique and realist aesthetics that the film attempted to articulate. Notice, further, how the personalization of the film text involves bodies being found for the virtual enunciation of the film, bodies being presenced, as when the offending director or actor is sought out—in his person, as offended viewers in the Tamil context sometimes do (see Nakassis 2020, 2023b); or through his proxies, as in public protests against public property (as in Bharathi Kannamma)—or when co-present spectators in the cinema hall become proxies for characters on the screen itself, where some viewers (re-)enunciate the film by recontextualizing it, projecting it onto other viewers as the screen of their ‘caste fanaticism.’
There are many issues at play in this example. But for our consideration, three sets of questions emerge: First, what kind of images are these, that reach beyond the screen in and by and despite representing narrative worlds? And what makes their actness, this indexicality possible? Put otherwise, under what conditions and through what evenemential (illeity) process does an image stand or become an impersonal representation or a personal performative act, histoire or discours, or both? Or rather, in what ways do these two metapragmatic modalities of semiosis interact with each other to variable effect? Further, how are we to think of the film text itself as an actual, communicative act—rather than simply a space within which acts are virtually enunciated or simulated by the text?
Second, how does this implicate the question of subjectivity and its distribution? Here, I am concerned not simply with the question who enunciates (‘speaks’) per se, but more specifically with the question, who is taken to be responsible for the act of enunciation and why? Is it the putative author, the director, Balaji Sakthivel? Or perhaps, the addressee, the audience, whose voice, after all, is already inscribed in these texts and their constitutive uptakes?25
And finally, how do we account for the indexicality, the context, the actness of the image in answering the above questions? Put otherwise, to what extent are answers to the above already “in” the text itself, recoverable from their traces on the film text itself precisely because they are all already simulated in and by it? That is, can we fall back on Metz’s and others’ position that the semiotician is not a sociologist and that all we need to know about signification can be had through textual, rather than “contextual,” analysis? Or, perhaps, is this a false dichotomy (between semiotics / social science, text /context), one we precisely need to overcome? In the next two sections I take the first two sets of questions, and in the conclusions the third.
The Act/ness of Enunciation
Let’s begin with the first set of questions and ask, why did some viewers of the film take it up as an act of insult in the event of its screening rather than merely a story, merely a representation, and why did they hold the director as the subject who enunciated that insult to them? To use the terminology of Paris School semiotics, why was this representation shifted all the way in, eliciting the affective passions of Thevar viewers toward the director and co-present Dalit viewers in the space of exhibition? Put otherwise: if the image has an indexical component, how are we to know to what it points, that is, “where” its context is (in the world of the diegesis or beyond it or elsewhere)?
Consider this last question: “where” does the index point? In a film, is it to the shifted-out world of the diegesis, or perhaps the shooting spot, or, in Kaadhal, perhaps the train station where the director said he was told the story? Or is it shifted-in in the here-and-now of theatrical exhibition? Is it histoire or discourse, to use Benveniste’s and Metz’s terms? Or, to use a different but commensurable tripartite distinction, is the indexical sign “calibrated” to its object (what it “points to”) reflexively (where the sign points to its own act/event of enunciation, as in “normal” discours, where I denotes who utters “I”), reportively (where it points to another event of enunciation; as in reported speech), or nomically (where the narrating event points to no particular event as such; as in mythic speech and fictional narratives)? (See figure 6 for a brief summary and Silverstein 1993, 2021, and Nakassis 2020 for more in-depth discussion of these three types of calibration.)
As is clear, and recalling Paolucci’s points about the “illeity” of enunciation (and PERSON deixis), there is no pre-given answer to this question. Indeed, what is peculiar about indexical sign relations in general is that, as types, they are underdetermined. As noted above, indexicality involves a semiotic ground whose meaningfulness turns on an existential relation between a token-instance of the sign and some (metapragmatic) frame within which its “origo” (Bühler 1990[1934]) is situated (i.e., enunciated).26 And yet, as sign relations whose meaningfulness turns on their status as token signs, every indexical sign is itself only enunciated in relationship to other co-occurring signs; which is to say, is entextualized as part of an unfolding texture with respect to which its meaning is “fixed.” While “I” in the genre of everyday conversation where no other signs contravene its reflexive calibration may denote speaker of the token-instance “I” (and indeed, where many signs—gesture, gaze, bodily hexis, previous discourse, collateral knowledge—effectively converge on just such an interpretation), its being co-textualized by a matrix clause that grammatically subordinates it (as in reported speech) or its being said with a sarcastic tone of voice are enough to “shift” its indexical target elsewhere.
As this implies, the calibration of indexical signs turns on a process of entextualization which is also, thereby, a process of contextualization; it turns on the way in which textures (of enunciations) emerge in ways to project and build a context, in which enunciation at an encompassing level (of a “text”) becomes possible. Here, “context” is projected by an unfolding entextualization of signs as one part of an indexical relationship between what is constituted/carved out as an “inside” (a text) and an “outside” (a context); rather than an inert container of the text, “context is another language (langage)” (to quote Maria Giulia Dondero from a workshop discussion of a draft of this paper; also see Landowski 1989[1983]:98, 101), one that is invoked/created by sign activity. If entextualization is a real-time movement of de-contextualization, of bounding off a virtual achronic text from its real-time event of realization, this is only also through the implied coordinate process of contextualization, a dialectic by which an indexical relationship of text-in-co(n)text is precipitated. While, as the precipitate of an indexical process of entextualization/contextualization, such a context is equally virtual, this process and its product, a text-in-co(n)text, always involves an actualized (or realized, in the terms of Paris School semiotics) instance in a situation filled with other signs, signs that interact with and potentially resignify such a text in novel, unique ways. It is this diagrammatic whole, this text-in-co(n)text, thus, within which calibrations are figurated/enacted, be they reflexive, reportive, or nomic.
Figure 6. Metapragmatic Calibration Types
This means that the necessary scope of analysis for any particular act of enunciation is always beyond itself. Benveniste’s histoire, an énoncé that effaces its traces of enunciation is such a text-in-context achievement (just like discours is as well); and yet, as Benveniste reminds us, every histoire is also still an act of communication (and thus discours). After all, it takes place in some event (of reading, watching, hearing), and thus stands in (indexical) relationships to other signs that indexically anchor it in this and other events (to which, in its explicit, localizable form it feigns indifference), signs that contextualize this instance in a history of events and in some here-and-now. There is, in other words, always a residual reportive calibration to every nomic calibration; and always a residual reflexive calibration to both. Which is to say, these distinctions of calibration (reflexively, reportive, nomic)—like the distinction of histoire and discours (and the impersonal and personal enunciations they figurate)—are not kinds but dimensions of semiosis; they implicate and intertwine with each other. Indeed, they play off of each other in complex ways. Further, how they so relate are not predetermined but emergent from semiosis itself. They are the effect of entextualization/contextualization.
We see these points well evidenced in our discussion of Kaadhal above. In Kaadhal, we find the emergence of a film text—both across its production history and its events of exhibition—that is calibrated to multiple time-spaces: to a fictive (nomically calibrated) diegetic world (where the indexes are contained to the screen), to a reportively calibrated world (where the indexes are relayed to another event of discourse, to the putatively real—but in turns out actually fictively emplotted—story by the husband told to the director, and from there to the “real world”), and to the reflexively calibrated event of film screening. Moreover, these multiple calibrations—this histoire and discours—are intertwined with each other, activating each other in various ways; the histoire of the diegesis is always residually unfolding (reflexively shifted-in) within an event of screening whose shifting-out (reportive and nomic calibration) is also shifted back in; in Kaadhal, the nomically calibrated story is projected not just onto the screen but is reflexively calibrated in the space of the theater itself. Moreover, each such calibration presupposes the others; the story of Kaadhal was written anticipating that it would be so reflexively calibrated, just as the viewers took up the image as an act here-and-now-directed-at-us precisely because of its status as a (nomically and reportively calibrated) histoire.
Here, thus, the film text—including its narrative, its impersonal histoire—is already itself enunciated as a personalized act of discours, not simply in the sense that every narrative presupposes some act of narrating, but in the sense that what is depicted (the narrative/denotational text, the image text) is itself an act that is interactionally realized in the event of its narrating/exhibition.
The Production Format of Image-Acts
Here, we turn to our second set of question regarding subjectivity: if a sign, a filmic text, is enunciated, who or what thereby enunciates, and in what capacity? Recall that for Metz and others, it is not a person but an abstract, impersonal textual function that enunciates: narration itself, “a story that seems to tell itself” (Aumont et al. 1992[1983]:96).
Putting aside its arguments about film for the moment, we might ask whether the founding supposition of this view is itself justified, namely, the primal scene of the performative inhabitation of subjectivity—the act of appropriating the “I” and exchanging “I”s and “you”s—that is posited as the default, “normal” mode of discours: “têtê-à-têtê” “oral communication.” Does discourse actually, normally, or even essentially works like how Metz (and Benveniste) naively imagine it to be (see Paolucci 2022, which also questions this view)? Like all primal scenes, this one is more ideological fantasy than empirical fact.27
Deconstructing this primal scene and its metaphysics of presence is one major import of Erving Goffman’s (1974, 1981) partitioning of “speaker” and “hearer” into what he calls the production format and participation framework of an utterance, respectively. As Goffman notes, and as the example of Kaadhal attests, the speaker (or enunciator more generally) is not a coherent, self-same role in discourse. Rather, enunciating comprises an array of semiotic partials or actants (Fontanille 1989) that may or may not be aligned in or embodied by a single actor: what Goffman called the author (the role fraction who composed some utterance), the animator (the fraction who enacts it, the “talking machine,” as he [1981:144] puts it), the principal (the fraction responsible for it), or some combination therein.28 One may be a subject, an “I” with a stake in an utterance in each of these rather different ways—in fact, as Judith Irvine (1996) has argued, in principle in an innumerable number of ways—just as each such partial may be distributed across or constituted by a number of agents (not all of which are human). While we ideologically assume a self-present speaker who inhabits all such roles at once (Derrida 1988), an empirical glance at actual events of talk reveals this not to be the case at all.
If we follow Goffman, then, the distinction in theories of filmic enunciation between histoire and discours, begins from a misunderstanding of the basic facts about discours itself: discours too has an impersonality (see Paolucci 2022:47). Or to put it another way, discours itself slips into histoire, insofar as it precipitates any number of what Bakhtin (1982) called “voices” in one and the same “simple” utterance of an “I” (or “you,” or any other indexical term, as it turns out).29 And while this is a major point that Paris School semioticians have emphasized in arguing that enunciation is first impersonal (shifted-out), we should also stress its reverse implication: namely, the personalization of the impersonal, the always already shifting-in of histoire.30 Here, the evenementiality of enunciation reveals processes of personalization and impersonalization as multiply entangled, non-predetermined dimensions of semiosis, indeed, as the outcome of the politics of semiosis as it plays out in and beyond the text, on and off the screen. Without seeing this, we wouldn’t be able to account for cases like Kaadhal.
In the case of Kaadhal, if we ask who enunciates—and in what way (in what production format)—we find a political field of contestation, a multiplicity that is a dialogic space of negotiation, not a “generic truth” of what “the spectator” thinks. To what subject, then, we can ask, was Kaadhal’s narrative anchored and in what capacity?
Recall that Kaadhal was framed by its maker and its audiences as a “realist” film, and, as we saw, as a reporting of another narrative—one given in a face-to-face discourse on a train, a historical event, a “true” story. Such a film asserts a kind of enunciation that would seem to nicely confirm to the film theorist’s concept of histoire, detached from the here-and-now exchange with its audience, where the director recedes and the “story” tells—indeed, reality reveals—“itself.” I say “seems” since, as we saw, its representationality crucially depended on the retrodictive, reflexive enunciative gesture of the rolling text at the end of the film, where the director’s “I” appears in address to the “you” of the audience. In this reflexive enunciation, the director presences himself to speak, to narrate. Yet does he speak? And if so, in what role? Here, again, we confront the question of production format, and of the responsibility for the image, for is this “I” an author, a principal, or a mere animator? And of what section of the film? It is precisely this question that metapragmatically regiments—indeed, discomfits—the constitutively open-ended enunciation of Kaadhal.
With the closing rolling text, the narrative and the event of screening are attempted to be rendered asunder, attempted to be disallowed from superimposition. That is, they are reportively calibrated to each other, relayed through the enunciated subjectivity of the director who points to himself so as to frame everything else in the film as a different kind of act—not an insult of you, the viewer, but the telling of a story for you. Yet this is a strange enunciation, for the director, in assuming the “I” of narration stakes a certain authoritative claim to the histoire precisely in order to disavow his responsibility for it.31 He is the mere relay (its animator), the one who “tearfully submits” the film to us in “cinematic form,” not the narrative’s author, nor—most importantly—its principal (the one whose position is staked by the text, who is responsible for it). If the director assumes the “I” of enunciation, it is only for a moment, and only to disavow his enunciation as author and principal for the rest of the film. Who, indeed, enunciates? And for what stretch of semiosis?
Notice, then, how the director’s entextualization (or framing) of his film as histoire is itself a discursive act within (and at the outer edge of) the film text; and moreover, an act in dialogic response to another potential entextualization of his film as a performative act of insult. This virtual entextualization of the film-as-insult metapragmatically regiments Kaadhal by threatening its status as histoire—and, in doing so, thereby actualizes itself as a “voice” within the film’s textual structure (indeed, it is what necessitates the “lie” of the rolling text as a rejoinder). Here, the production of histoire is a response to the imminent and unruly actness of the image—its reflexive calibration, its always already residual status as discours—and as such is itself an illocutionary act, a “disclaimer” as the director put it: “In and by this act of discours, I hereby disclaim this histoire as my discours.”
This internal(izing) enunciation works to cleave off the narrative from the event of its narration, and it does so, note, by exchanging an “I” and a “you” with an audience, first virtually and then actually; or rather, first actually, with the director’s own response to his own film, then virtually as the director’s imagination of other such possible viewers, and then actually again, with this virtual “you” realized by a rolling text which requests that the audience take the film’s images as narrative and not as an image-act of filmic discourse. In doing so, the rolling text reflexively stages for us the distinction of histoire and discours, performative image-act and representational narrative-text. Yet in so doing, across this chain of entextualizations, virtual and actual, Kaadhal also gave itself over to future unanticipated recontextualizations, (re-)enunciations of the film by its viewers in the theater as casteist intimidation and threat. Enunciation is a political and ideological gambit whose success is far from determined, indeed, is only determined after the fact, never solely within the confines of the image that appears but of a film text in its cinematic contexts, that is, as the effect of multiple chains of enunciations that we access in/through the empirical relationship of text-in-context.
Conclusions; or, On the Site of Analysis
Let me end this discussion with some concluding thoughts that offer possible landmarks for our intellectual sea lanes across the pond.
(1) The filmic image is not necessarily “impersonal” (Casetti 1995[1983]:120; Buckland 2000:52–76) nor is oral communication strictly “personal.” Impersonal or personal, histoire or discours, shifting-out or shifting-in, representationality or performativity—not only are these not either/or oppositions (for one and the same text may be both or not-quite either), they are contingent achievements of semiotic processes that are themselves perspectival and relative; which is to say, the status of an image (or any sign process) as one or the other (or both) are not ontological or analytic questions that can be settled as such. Rather, these are in-principle open questions, and they are not ours (alone) to answer. Which is to say, they are intrinsically political questions. The cases that we have looked at reveal a contestation over the coordinates of enunciation within enunciation itself, not simply before or after the fact but as a constitutive ambivalence in itself. This is because the entextualization of the enunciated text is always a phase of a dialogic process and a part of an encompassing co(n)text; this ambivalence at the heart of the enunciated text is a result of the fact that every enunciation already responds to and anticipates another enunciation, mediating it proleptically even as it is determined by it in futuro (i.e., retrodictively). That an image may function as histoire, or as a performative image-act of discours, or some other combination, is a situated achievement that is the effect of semiotic labor of various kinds (not just the presence or absence of deictics or other indexicals), always caught in fraught relations with other interpretive frames and political projects. This means that enunciation is not an objective fact to be logically reconstructed in the abstract (even if we can, and must, in a sense, do this as analysts) but must be documented through attention to empirical situations of semiosis. Such enunciations, thus, do not only virtually figurate possible stakes, but also actually map out a terrain where what is at issue is what and how is a text enunciated. Put in other words, “enunciation” is not simply an “external” analytic frame we bring to bear on the analysis of discourse, nor an isolable object of analysis per se, for the question of enunciation is itself already part of the phenomenon itself—we are not the only analysts of enunciation, for it is already being theorized within itself by those party to cinematic semiosis. Directors, actors, viewers, film critics, film theorists, among many others—their sophisticated theories/ideologies of enunciation are very much part of how and what is enunciated, when, for whom, and with what effects. Moreover, such dialogic relations leave their marks in the text, indeed, they entextualize images, realizing them in all their complex polyvocality. But how to hear or see or otherwise perceive them?
(2) This means that we cannot focus on medium specificity but rather must focus on semiotic functionality, functions that far outstrip any medium or modality. Any medium or modality of semiosis wherein indexicalities can be modulated (i.e., calibrated in different ways) will also be modulated in the im/personality of what is enunciated. As it turns out, this is all of them, since all sign processes are materially borne, and thus unfold in textualized arrays of signs in time and space with which they indexically transact. To speak about the personalization or impersonalization of enunciation—as an empirical process and an achievement—is to speak about enunciation as always proceeding via a dialectic of entextualization and contextualization, not only/primarily through localizable, segmentable, explicit/referential traces of enunciation “in” a text (like personal pronouns or anaphors) but by distributed, global, metapragmatic functions that stretch across and breech the texts-in-context through which im/personality is figurated, contested, taken up, reformulated, institutionalized, and so on.
(3) Cases like that of Kaadhal (see Nakassis 2020, 2023a for others) empirically refute Metz’s claim that “I” and “you” are not exchanged in cinema, that films do not hear the responses of their viewers, do not touch them, encumbered by the matter of communication.32 Tamil films, like all films, anticipate future uptakes/entextualizations by (en)registering histories of entextualization. And not just before the fact, such films also respond to fan uptake after the fact. Tamil films at the time of my fieldwork were not uncommonly recut or even reshot when fans were upset. And in these films, film stars talk directly to their fans in their films, they mark their deictic here-and-now in the event of exhibition, just as fans talk back to them in the theater (Nakassis 2017). All such signs are deeply mediated, to be sure, yet this is not simply a simulation of presence except from the ethnocentric perch of the analyst who refuses to take such cinematic practices and media cultures on their own terms. “This is what spectators [or at least some spectators!] think,” in other words. Certainly, the temporality and mediators of dialogism in cinema and oral conversation differ, but that is no grounds to believe cinema as such to be a “mono-directional” or “non-interactive” medium (Metz 1995[1987]:144). It simply is not the case.
Yet under what ideological and historical conditions could anyone (i.e., Metz) think it to be so? That is, why does film seem to be analyzable and coherent under conditions of extreme decontextualization? An inspection of the history of European and American cinema reveals that the notion of a non-interactive, inert text hermetically sealed off and ever absent from its viewers is not only an ethnocentric view (other cinemas are not like this), nor is it only a disciplinarily partisan view (not all of Western cinema is like this either, only the narrative full-feature film that film studies has historically focused on; Casetti 1995[1983]:124); it is also ahistorical. This view effaces what is exactly the point: the political achievement and institutionalization of what Metz takes as the impersonality of cinematic enunciation. As Miriam Hansen (1994) has shown, the institutionalization of “classical” spectatorship that became the “essence” of cinema for film studies, was a reorganization of and differentiation from the so-called “primitive” institutionality of theatrical experience that came before it: the “cinema of attractions” (Gunning 1986) with its deeply shifted-in, performative, and interactive spectacle.33 Further, this institutionalization was itself an elite economic and moral project; an economic project because the gentrification of cinema, as a working class form of entertainment, was linked to political economic interests to bring a new type of audience into the theater, one that would pay more for tickets to see more narratively coherent, realistic, novelistic films in less bawdy, rambunctious, “civilized” spaces; and it was a moral project in the US insofar as it was seen as a way to reform and uplift the vulgar practices of the ethnic working classes, the immigrants, and all those who needed to be properly socialized to white middle-class norms of American social being.
Important, here, is how this was a politics of entextualization/enunciation; indeed, the above transformation of Euro-American cinema was part and parcel of the de-indexicalization of film (abstracting it away from the theater itself into the nomically calibrated world of the self-contained fictional text, into histoire); just as it was a shift in the authoritative basis of enunciation/entextualization (where it became the text itself, and its absent auteur, that narrates rather than a co-present lecturer or a projectionist); just as it was a cleaning-up of the social (non-referential, non-narrative) indexicalities of all of the above as linked to racialized and class-linked images of personhood. Enunciative praxis indeed!
The kind of cinema that film theory has typically picked, thus, are low-hanging cherries, a “classical” cinema that is the result of a historical, political project that produced not only a type of cinema, but a culturally and historically naturalized ideological experience of it, as composed of seemingly auto-entextualized texts, so seemingly autonomous from their contexts that their indexical particularities fall away—pruned away with the right semiotic tools—to reveal perfect beautiful structures, self-contained and filled with a spirit which we can divinate without recourse to any “incarnation.” A fetish.
(4) We should refuse any hard separation of (external) “communication” and (inner) “enunciation,” and all related spatializing, analytic distinctions that work to produce hard, discrete divides.34 Semiotics, avers Metz, should be a science of the text divorced from its social contexts; which is to say, should be the study of that entextualized whole which can be analyzed under conditions of radical decontextualization, i.e., which can stand under “plausible” analysis solely through the entextualizing work of the analyst and their intuitions. Obviously, this approach doesn’t work when our intuitions fail us, as anyone who works outside of their own social milieux knows (though the point holds even when we work in our “own” “culture”). Anything beyond the film text (narrowly construed) for Metz is a “risk” (1995[1987]:149–50), and thus beyond the scope of the semiotician, as we noted at the outset. What anyone actually does when engaging a film is thus, for Metz, an “exterior exchange,” an “external support” for “text orientations of the text” (ibid.:151; emphasis in original) because the text is already “finished” when we see it. (To see the text as unfinished, indeed unfinalizable, undoes this, needless to say.) As Metz writes, “Research about enunciation does not make use of such [empirical] methods; it is most of the time based on film analysis and remains ‘internal’” (1995[1987]:151). On the next page: “Textual analysis, even when enunciative, remains textual analysis. If you want information about audiences and filmmakers, you have to go and get them on the spot” (ibid.:152). Luckily, the analyst is always there, where he always already is (“on the spot”), available for “information” in his own act/event of analysis. Yet, for Metz, this enunciation need not lead to solipsism or contradiction, for, as noted, what the analyst is trying to understand is not “the spectators’ actual reaction” (ibid.:152) but rather, to recall, “a general, or rather, a generic truth; it concerns THE spectator. Anyone can find it within himself” (ibid.:157; emphasis in original);35 indeed, abjuring any empirical research, one must find it within himself or risk the total collapse of this enterprise.
Ironically, it is precisely through this (Cartesian) intuitionism, this confidence in already knowing and already having the right categories of analysis that a certain empiricism—of the most naïve variety—is welcomed through the backdoor.36 Indeed, underwriting Metz’s argument, as we saw, are a whole set of (problematic) empirical assumptions that “spectators think like this” or “feel” like that. Yet given his methodology, how could Metz, the semiotician ever know otherwise?
(5) At issue, then, is not simply a theoretical question of enunciation. At issue is a methodological question: namely, how does a particular methodology compel (or even naturalize) a theoretical position, and vice versa, how does our theoretical approach dictate a methodology? Metz theorizes like a sovereign who concedes methodological territory to others so as to shore up his own internal territory, concluding that textual (internal) analysis is the domain of film studies and semiotics, while everyone else (sociology, psychology, anthropology) can toil in the wilds.37
Yet the above theoretical propositions compel us territorialize differently: namely, that semiotic analysis must always proceed empirically, never from first principles or intuitions, never from the armchair but in the “field”; for semiosis is not “in” a text, a medium, a code. It is a dialectical movement that precipitates insides and outsides, texts and contexts, where what is “inside” is constituted by an exteriority (within and without), and vice versa, what counts as “context” is already mediated by a process of projection from “within.” To my mind, this troubles the notion of “relevance” that often functions as a “immanent” principle in (post)Greimassian semiotics to discern what is the “proper” object of semiotics and what is “extra-semiotic,” for what is at issue here is precisely the semiotic process through which an inside and outside of the “text” is precipitated (indeed, enunciated), that is, through which “relevance” is produced as and belied by the lure of the text.
Which is also to say, methodological individualism will never do for the kind of semiotics we require. We have to do exactly what Metz assures us we need not do: “get out of the text” (1995[1987]:153; in Metz’s sense of text, of course). This is not because we abjure textual analysis, but because it is precisely the division of text from context that we need to deconstruct and pragmatically retheorize as an indexical process of entextualization/contextualization. This requires us to wed, blur, and do away with the division between formal semiotic analysis and empirical social science. Indeed, empirically (and ethnographically) approaching semiosis is, in my view, the best way to do textual analysis. There is no way to analyze the structures of enunciation in the examples we have given above—to even know what “the” “text” is—without taking the “risk” of looking at “actual” “incarnations of instances.” Moreover, having done so we realize that these are not “abstract” “positions” in a purely virtual structure but a movement of entextualization and contextualization, an illeity process and enunciative praxis that, again, can only be studied through a movement beyond the limits Metz (and others) would set for us. For there are no encounters with virtual structures that are not always already mediated by the movement of actual sign processes (Metz 1995[1987]:158), no texts independent of processes of entextualization/contextualization, no enunciations independent of communication (i.e., other enunciations), and thus no alibi or apology for not starting from the mediation of token-signs of semiosis in their events of happening, that is, starting from indexicality and all it implies.
Perhaps all this has become a truism; but it only becomes “true,” that is, does the work of truth in guiding the labor of inquiry, when one carries it out in scholarly practice. It is not enough to acknowledge “context” or “practice” in theory and then ignore it in our practice; rather, we must deconstruct the divide between these supposed antinomies by analyzing semiosis in/as in vivo, in situ, cultural and evenemential contextualizations.38
Acknowledgments. This paper emerged, initially, from a lecture, “The Actness of the Enunciated Image: Notes for a Linguistic Anthropology of Cinema,” presented to the Atelier de Recherche en Anthropologie du Langage, Centre des Liens Sociaux, (CERLIS) (Paris, June 26, 2019) and a related workshop paper circulated at the Wenner-Gren conference, Sense and Semiosis: Bridging Conversations between Linguistic and Visual Anthropology (September 26–28, 2019, University of Chicago). The ethnographic analysis of, and some of the concerns explored in, one of the examples in the workshop paper (Arunachalam) went on to be published in Nakassis 2020 and the other (Kaadhal) in the fourth chapter of Nakassis 2023a; a revised version of the original paper was then circulated to and discussed with the year-long working group, Entextualization/Enunciation: Bridging Linguistic Anthropology and Paris School Semiotics on December 8, 2023 at the University of Chicago Center in Paris. Thanks to the discussants of the workshop papers, Martin Lefebvre in Chicago and Tatsuma Padoan in Paris, and to participants in both workshops: in Chicago, Chris Ball, Meghanne Barker, Jenny Chio, Steven Feld, Anna Grimshaw, Barbra Meek, Norma Mendoza-Denton, Teri Silvio, Rupert Stasch; in Paris, Enzo D’Armenio, Felix Danos, Maria Giulia Dondero, Aurora Donzelli, Mariem Guellouz, Bertrand Masquelier, Alvise Mattozzi, Urmila Nair, Franciscu Sedda, and Shu Takeda. Critical feedback on the article by two anonymous reviewers of Semiotic Review and Tatsuma Padoan have significantly improved the paper.
Endnotes
1. Here, I limit myself to a narrow sampling of the continental semiotic tradition, focusing on the work of Metz on enunciation and, more broadly, works that identify themselves as Greimassian or from the “Paris” or “French” school of semiotics. While Metz is not a Paris school semiotician per se, his work on impersonal enunciation is often cited and discussed in further developments of this concept. Needless to say, Greimassian semiotics is a diverse field (Parret 1989), evincing much contestation and development over time (the development of the concept of enunciation is a case in point; see Bertrand 1993), in particular through its dialogues and overlaps with other semiotic traditions (e.g., the work of Hjemslev, Lotman, Eco, and others). My discussion of the relevant texts in the so-called Paris School tradition is further limited given that much of the work is published in French and Italian and not translated into English. The reader familiar with this literature should see my discussion, thus, as a preliminary to a much wider dialogue that would take into account a wider assortment of sources, newer developments, and the like.↩
2. The historical dialogue between the two scholars on this specific issue is important to note. Benveniste’s “La nature des pronoms” was first published in 1956 in a volume in honor of Jakobson’s 60th birthday (For Roman Jakobson, Mouton & Co., La Haye, 1956); Jakobson’s “Shifters, Verbal Categories, and the Russian Verb” was published in 1957, based in part on papers delivered in 1950 (for the Société Genevoise de Linguistique and at the University of Michigan). As Eva Krásová (n.d.) indicates, there is no evidence that Benveniste was at the former meeting, though the two may have met in Paris two weeks earlier. However, from a 04/01/1951 letter from Benveniste to Jakobson (see https://doi.org/10.4000/hel.1284), it is clear that Benveniste knew about that paper. In the 1957 essay, Jakobson uses the distinction of procès de l’énoncé and procès de l’énonciation (p. 45) to distinguish narrated event from speech event. In “La nature des pronoms,” Benveniste talks about processes (processus) of enunciation (énonciation); and Jakobson cites that paper in his 1957 essay (p. 43) to compare his account of the I—as index—with Benveniste’s treatment. John E. Joseph, for his part (“Benveniste and the Origins of Enunciation,” unpublished ms), notes that Benveniste’s full and explicit discussions of énonciation (in distinction to énoncé) emerge in the 1960s, and were predated by Lacan in his 1958 seminar. In any case, and predating Lacan’s 1958 seminar, both Jakobson and Benveniste had already turned to the question of deixis/“shifters” (Benveniste with a focus on questions of subjectivity already in the 1956 essay, Jakobson in relationship to indexicality in language structure and the speech event) and were in dialogue with each other on various related issues in this period.↩
3. Throughout this article, I use the terms token, sinsign, type, legisign, icon, index, and symbol with their technical Peircean senses; in addition, text, (and related terms, entextualization, contextualization), pragmatic, and metapragmatic are used with the technical senses developed in linguistic anthropology unless otherwise noted; these terms (e.g., icon, text) are not always synonymous with their usage in Paris School semiotics. Further, I often use the terms actual and actualityin their non-technical sense, i.e., not in the technical sense developed in accounts of enunciative praxis (e.g., Bertrand 1993; Fontanille 2006[1998], 2017), while using actualization (and actualized) in the technical sense developed in those accounts (see note 11 below). For the process of rendering something actual (i.e., material, embodied, existent) I use the Paris School semiotics verbal forms, realize and realization. ↩
4. Of course, there are many more types of indexical signs in language besides denotational indexes (and thus many more “factors” of the speech event); indeed, most of indexicality in language is non-referential; see Silverstein 1976 for a foundational treatment. As this also implies, there are also many more “functions” to language, perhaps an open-ended set of them, rendering Jakobson’s speech-event model—while important—limited or, at best, suggestive.↩
5. Benveniste rightly saw such signs as key to the actness of discourse, as revealed in his analysis of the then-emergent “ordinary language” philosophy of Austin (1962) and its account of “performatives.” See Benveniste’s (1971:231–39) essay, “Analytic Philosophy and Language,” which comments on the joint conference between French and English scholars that included a presentation by Austin, in French, on his research about constative and performative utterances.↩
6. I would again stress that what I describe in what follows are not the only paths these traditions have explored. For example, one might consider the circuitous connections between Greimas, Fabbri, and Latour with respect to Goffman and Garfinkel, and further, between the latter and their uptake by linguistic anthropologists; or between other “post”-structuralists (Deleuze, de Certeau, Derrida, Foucault, Kristeva) in relationship to both traditions; or to the engagement with Peirce in the work of Umberto Eco and, more recently, Paris School semiotics; and so on. If these are not the only branches and landmarks, then, they are also not the only possible points of reunion.↩
7. This tendency also continues on in much recent work (see, e.g., Lagopoulos and Boklund-Lagopolou 2025b:111), though it also the source of debate and contestation as well (see, e.g., Dondero 2025a in the same volume, and discussion in the main text further below). ↩
8. Silverstein (2022:23) critically chided this as a “ball-tossing folk model” of communication (the joke turning on how the diagrams of signs famously published in Saussure’s Cours resemble footballs). ↩
9. Benveniste directs his critique at Jakobson; certainly, Jakobson’s invocations of Shannon and Weaver’s informatic model and its many iterations in “communications” studies encourages this view, though, ultimately, I think it is wrong to see Jakobson’s (1953, 1960[1958]) model of the speech event as suffering from all the ills of the engineering model of communication. The fact that Jakobson’s speech-event model is sourced/lifted from Bühler’s (1990) anti-behaviorist, non-cybernetic organon model should give us pause, just as a close analysis of what Jakobson does with this model reveals that it far exceeds, indeed deconstructs, the referentialist vision of language that it is sometimes wrongly taken as instantiating. See discussion below. (That said, Jürgen Spitzmüller [p.c., 30 XI 2023] reports that Bühler was avidly interested in the new medium of his time—radio—and this, too, may have influenced his uptake of Aristotle’s rhetoric in the organon model.)↩
10. Yet here, too, however, we can detect this ambivalence lingering. Emphasis in the works cited in the main text above has still tended to rest—in practice, if not theory—on understanding the virtual structuring of such practices and less on their actual empirical particularities; hence the tendency to reduce “reception” to the figure of the spectator (or reader) “in” the text or “production” to the traces left by production “in” the text rather than starting from the semiotics of “reception” or “production” themselves. Indeed, these domains still tend, in a not insignificant number of works in Paris School semiotics, to be figured as “extra-semiotic.” In addition to the tendency to reduce such processes to an “extra”-semiotic domain (i.e., what is not “in” the text that the analyst finds themselves analyzing), even when these domains are taken seriously, it is not uncommon that they are posed as autonomous from the semiotics of the text itself; note that terms like reception and production characterize such processes from the perspective of a text that is treated as self-evident and pre-given (see Nakassis 2016:184–87, 218–19, 231–37 for a critique of notions of “reception” and “production”). In this view, also shared with cultural studies approaches to “encoding” and “decoding,” the text, autonomous from but transacting with its contexts, “creates audience responses” and “operates through manipulative procedures to influence the audience’s interpretative processes.” The overall result is that context, events of interpretation, audiences’ “competences,” and the like are super-added to a text—and thus to a semiotics—as exteriorities which we can append to fill out an analysis, but which, arguably, do not fully reconfigure the whole of which they are a part. (There is a further epistemological issue of how we, as analysts, can even discern what is “in” and “out” of the text; and an even further ethnographic question as to how this epistemological question is part and parcel of semiosis itself.) It is precisely this conception that we problematize in an effort to rethink the text/context divide as a process of entextualization and contextualization. Of course, these broad strokes are not meant to characterize all semiotics that draw on the Paris School tradition. See, for example, Dondero 2025a:249, 250, 251, 253–54. Similarly, one might point to works that have used concepts from (post)Greimassian semiotics in empirical research in ways consonant with the approach advocated here. See, e.g., Padoan 2021, 2024; Frisone 2023, as well as the various articles in this theme issue.↩
11. Enunciative praxis (Bertrand 1993, 2000; Fontanille 2017) describes the dialectic movement in between different “modes of existence”: virtualization (cf. langue), actualization (the movement from the virtual to the realized), realization (cf. parole), and potentialization (the movement “back” towards virtuality). This circuit attempts to understand how instances of discourse feedback into and transform the virtualities they call upon. (We can compare this to Silverstein’s [1979] notion of the “total semiotic fact” or Agha’s [2007] discussion of “enregisterment”; in this issue, see Dondero 2025b; Donzelli 2025; D’Armenio 2025 for more discussion.) Yet, even in these accounts, we might note that analysis often first departs from the (“deep”) virtual towards the actual (“surface”) and back again, though less often the other way around. That is, the problem seems to be about how to think about how virtual structures realize themselves (by being “summoned” and “convoked” in enunciative acts from “generative” “deep [narrative] structures”) and then are reconstituted (if differently), rather than the epistemologically complex—and, for the empirical social scientist, methodologically prior—question of how, from the Heraclitan flux of indexical semiosis, virtual forms (Ideas, even) emerge, circulate, are institutionalized, and thus are sensible and knowable to anyone (including the social scientist). ↩
12. Not only are both histoire and discours enunciated (and, as the product of a process of enunciation, thus have a metapragmatics), both are constituted by the modulation of deictic categories. Histoire, that is, is not characterized by the absence of deictics (PERSON, TENSE, etc.) per se, but by their taking their most unmarked forms (e.g., third-person pronouns; simple present, indicative mood, active voice, non-interrogative/assertorial verb forms, etc.). The absence of indexical grounding, after all, is a deictic modulation (i.e., shifting out is a form of shifting). Of course, every histoire, as an act of discourse will consist of many more indexicalities than those in the (written) narrative, even if they are not deictic (i.e., localizable, segmentable, inherently referential, etc.). Focus on deixis, needless to say, should not blind us to the more fundamental and encompassing indexicality of language (which exceeds the “code”), lest we fall into the seductive ideological trap that we hope to evade through a turn to deixis/enunciation. ↩
13. Metz’s argument upsets, yet reproduces, the problematic assumption that the deictic component of enunciation and subjectivity are reducible to personal pronouns (see Fontanille 2006[1998]:187ff. for such a critique); yet, if enunciation is ultimately indexical, and if indexical grounds far outstrip segmentable, localizable, referential parts of language—which they do—then subjectivity and enunciation has to be studied by attention to the full range of indexical (and meta-indexical, i.e., metapragmatic) sign processes. ↩
14. Another way of saying this, to anticipate some points discussed later, is that enunciation in film (but more generally, we would add, including in discours) is not reducible to highly localizable, segmentable, intrinsically referential, and denotationally transparent (i.e., explicit) signs but is, as a more general rule, distributed, global, non-referential, and implicit (Silverstein 1981). Metz is not wrong to chide a naïve view of subjectivity in film, though he is wrong to assume that such a view of subjectivity works in any medium or modality (after all, it doesn’t work that way in oral communication either; discussed in the main text below). Notice, incidentally, that many of Metz’s (2016[1991]) examples—such as mirrors/frames in the screen that reflexively remind us that we are watching a film—involve iconic resonances between the narrating and narrated frame that are upgraded into an indexical connection (that [in the film] is like this [watching the film]), which is to say, they involve a poetic (metapragmatic) function that does point to and "touch" the viewer in their context of viewing. The difference is that this doesn’t involve a referential index (like a person-indexing deictic) but a diagrammatic (non-referential) indexical icon, which if so entextualized can function “deictically” (in Casetti’s sense). ↩
15. To anticipate some of the technical language introduced below, we would say that the origo of such deictics are shifted-out or reportively or nomically calibrated. ↩
16. Despite my mention here of film production, Metz is uninterested in film production as a semiotic process, reflecting the semiotic concerns of his time with the text and its “reading” by a spectator (see Fontanille 1998; Dondero and Fontanille 2014). ↩
17. Perhaps we should say that when the empirical viewer agrees with the film analyst, it is evidence of the rightness of the latter; and when the viewer doesn’t, their opinions can be dismissed as cinematic dross to be studied by the social scientist. ↩
18. It is, as if, the virtuality of the text (énoncé) replaces the ideality of the “system” (langue) so as to protect semiotics from the vagaries of the actual, the real, the empirical. ↩
19. Claudio Paolucci (2022), for example, has suggested that developmentally, personal enunciation is a late arrival in the linguistic constitution of the speaking subject, a process that is preceded by impersonal enunciative processes. (This echoes, in fact, the context in which Jespersen [1922] introduced the term “shifter”: language acquisition, in which he notes that children have difficulty mastering shifters; it also echoes Lacan’s own formulations of the process of subject formation.) This is only one of Paolucci’s arguments for the impersonality of enunciation. Paolucci’s concerns and arguments are much wider than questions of cognitive development, as I note in the main text; his argument, as I understand it, is primarily a philosophical one within which psychological findings are found to be consonant. Others have also argued that shifting-out and the impersonality of enunciation is a logically prior process of the generative trajectory of semiosis (Bertrand 1993), just as various scholars have productively emphasized that enunciation and subjectivity are the result of “collective assemblages” (Deleuze and Guattari 1987[1980]) or “regimes” (Latour 1999) of enunciation rather than individual sovereign subjects. ↩
20. For Paolucci, subjectivity consists in the capacity of a subject to represent themselves to themselves (i.e., make themselves an object for their own semiosis); here, the fact that “I” and “you” are not only indexes of the speaking or addressed subjects but also parts of denotational texts in which they are objects of reference-and-predication is taken to mean that they also partake of the non-personal category (he, she, it, they). To be an I, thus, is always also to be a denoted he/she/they/it (cf. “I” is who says “I”), though not the reverse. See Urban 1989 on a similar set of ethnographically posed arguments and critiques focused on Benveniste’s conception of subjectivity vis-à-vis reported speech constructions in Shokleng myth. ↩
21. I take this as the most important conclusion of Paolucci’s evenemential approach to enunciation, even if he, too, surprisingly poses a hard divide between “face to face communication with someone” (or “ ‘in person’ locution” “in presence”)—Benveniste’s sense of enunciation—versus a “Latourian” notion of impersonal enunciation (Paolucci 2022:47), where the latter is primary over the former, or where they each have their own domain (Paolucci 2020:125–26). ↩
22. Goffman lurks in the endnotes of Inside the Gaze (Casetti 1998[1986]:153n53, 154n59); Casetti gestures to Goffman as having discussed many of the semiotic issues that Casetti brilliantly analyzes, indicating that a more rigorous engagement with Goffman would yield empirical and conceptual advances for film studies. Here, I hope to cash in on this buried promissory note. ↩
23. This section is a redacted discussion of a fuller analysis in Nakassis 2023a:146–93. ↩
24. Consonant with Metz’s (2016[1991]) discussion of film’s enunciation through metadiscourse, Kaadhal is replete with reflexive references to the viewers’ own looking: characters watch films (and others watching films) in theaters, on television, and in the bus, as well as see each other in mirrors looking (see Nakassis 2023b:162–65, 171–72, 175). Such reflexive gestures were an intentional strategy of the director, designed—as he told me—to tell the viewer that life is not like the cinema (in particular, not like “commercial” mass-hero films with their happy endings) and to point up their own hypocrisy (why won’t you accept in your own life what you accept in film?) (see Nakassis 2023b:171, 297n24). That is, they serve both to reflexively make the film enunciate (as a metadiscourse, as Metz argues) but also to address the viewer (deictically, as Casetti would put it). ↩
25. A certain kind of liberal, elitist subjectivity, as we saw in Balaji Sakthivel’s disappointment in casteist uptakes of his film, would certainly lay responsibility for such performativities as the fault of impressionable subaltern audiences, and see such images as enunciated, in some teleological way, by them. ↩
26. Benveniste typically takes the origo to be the speaker or the moment of utterance, though as Hanks (1990) has rigorously shown, this is only one possibility within a complex range of orīginēs that include addressee, speaker-address relationship, evidential modes of accessibility, and so on. ↩
27. This fantasy isn’t without its own real effects, however. One such subjectivizing/objectivizing effect is securing the subjectivity of the theorist and his object of analysis; indeed, note how this primal scene of discourse exists precisely for Metz to differentiate language from film, and thus to secure the theorist his autonomy—here, from the linguist and the literary scholar—as a film analyst by defining the essential difference of his object of analysis: film. ↩
28. Similarly, in a participation framework, one may be a “hearer” in any number of ways: ratified or unratified addressee, bystander, overhearer, eavesdropper, audience member, and so on. These “footings” are dynamic, shifting from moment to moment, and laminated on top of each other in complex ways. ↩
29. That is, much “oral conversation” is just this: histoire, people dialogically narrating other speakers and their speech events by enunciating their enunciations, shifting-out and speaking through the “voices” of others (Bakhtin 1986), which is what happens when anyone speaks in some speech register or other (a dialect, a professional register, a slang, standard language, etc.): one voices (enunciates) an identity through the virtual personae indexically linked to some repertoire of signs (Agha 2007). And since we always speak in some register or other, all speech has an impersonal component. Enregisterment is the term for the enunciative praxis, the historical process, by which registers emerge, circulate, are troped upon, disappear, perdure, transform, and so on. ↩
30. Here, the claim isn’t the philosophical (or logical or metaphysical) etic claim about the status of enunciation (as personal or impersonal from the analyst’s point of view); rather the claim is that when we look to empirical processes of enunciation (to actual “illeity events”) we find that the seeming impersonality of semiosis (as in film, as in histoire) shades into, maybe appropriated as, and is underwritten by, the personalization of this very process. ↩
31. This reveals the extent to which Benveniste’s notion of histoire as a discourse that effaces its own marks of enunciation (Aumont et al. 1992[1983]:96) cannot be assumed in this context (or maybe any context), precisely since histoire is not a function of effaced, reflexively calibrated deixis but is the outcome of a process of entextualization, one that here involves an explicit intervention by the director’s “I” into the film text, an intervention so that, retroactively, the story can seem to have “told itself.” ↩
32. Here are a list of related propositions from Metz I refute: “enunciator and addressee ... do not exchange their marks along the way; and the addressee does not change by his reactions either the propositions or the proposal of the enunciation” (1995[1987]:145); “this ‘You’ will never be able to respond” (ibid.); “the two poles of enunciation ... cannot be exchanged, nor can they touch each other” (ibid.); “this encounter is always desired, always missed, sometimes approached; it constitutes the fundamentals of film” (ibid); “film is not interactive, it does not accept any response” (ibid.); “Film does not contain any deictic equivalents” (ibid.); “Reflexive rather than deictic, it does not give us any information about the outside of the text, but about a text that carries in itself its source and its destination” (ibid.:153), “the film ‘speaks’ alone all the time, it does not allow me to say anything and it cannot get out of itself (it was made before, once and for all)” (ibid.); “the so-called outside of the text can therefore be only text, reduplicated text, metatext” (ibid.:154); “it seldom happens that the reception instance of the film is solicited in such an explicit way, that is, that the spectator is directly addressed by a diegetic intervention” (ibid.:155). ↩
33. As Hansen (1994) and others have noted, rather than a long-narrative form involving psychological identification and absorption, early cinema in the US and Europe offered up spectacles, with particular films as one part of a bricolage of numerous (co-)texts: from live dance and vaudeville acts to short films, the latter of which would have been accompanied by live music and a lecturer—an enunciator (among many) who would co-textualize the image with commentary and explanation. This was a cinema where the entextualization of theatrical experience and participation was distributed, non-standardized, and improvised. But not just a mode of engagement, to speak of the institutionality of enunciation at this moment is also to speak of a particular aesthetics of textuality: films at this time presupposed in their texture this state of exhibition. Such films were largely frontal in their addressivity, often short one-shot, static moving images with a theatrical proscenium, exploring events and actions with minimally elaborate narratives. While my comments here are limited to US cinema, they have a wider import insofar as it is in this period that American cinema became a dominant, hegemonic form with influence reaching far wider.↩
34. Other such dualisms that fall out of this founding split (all of which need to be deconstructed) in Metz’s (1995[1987]) text are: textual stage versus personal stage (p. 158), semiotics vs. sociology, textual vs. empirical (p. 157), “true” deixis versus anaphora (p. 142), “real” vs. unreal/“mimicked” (p. 142 on “real feedback”; p. 144 on “mimicked deictics”), image versus speech, oral versus written, “immediate interaction” vs. mediated interaction (p. 151). ↩
35. Well, not anyone. Metz’s view seems to be that actual spectators’ experience, interpretations, and views are far less sophisticated that the analyst (Metz 1995[1987]:152, 157), a gesture of dismissing the reflexive views of parties to film (producers, viewers, etc.) that, again, shores up the unique access the semiotician has to the truths of the text, with one negative effect (among others) being the inability to theorize how local semiotic ideologies are constitutively part of the enunciative process. ↩
36. Metz (1995[1987]) appeals, without embarrassment, to seemingly self-evident propositions, each of which are actual empirical claims (that lack any evidence, or any attempt to produce it): “this is the general impression” (p. 145; whose "general impression"?) “not everyone feels it clearly” (p. 145; who is "not everyone"?), “in most cases, this spectator does not think of the ‘Imager’. ... he does not, of course believe that things reveal themselves: he simply sees images” (p. 145; who so believes, thinks, or sees?), “It is also the common feeling” (150, whose "feeling"?), “The film is the enunciator ... This is how people think” (p. 150, which "people"?). Are any of these “plausible suppositions” plausible? And if so, to whom and under what conditions? And does a certain normativity lurk underneath them? ↩
37. Not only Metz, of course (and if I focus on Metz it is because he is the smartest and clearest-eyed in posing this problematic); it is commonplace in film studies (see, e.g., Dudley Andrew’s introduction to Casetti’s Inside the Gaze [1998(1986)]). This cut, as we already noted, is the founding structuralist gesture of Saussure’s Cours as handed down to us. One continues to find it in works in some works in continental semiotics, as indicated in the introduction to this article, even if there are many within this tradition that have pushed beyond such false divides (e.g., Latour 1999; Padoan 2021, 2024). ↩
38. This is not to say that we must take as our only empirical material observations of singular events of semiosis; certainly, the material discussed related to Kaadhal is not primarily of this type, but based, in addition to my own ethnography (Nakassis 2023b), on second-hand accounts such as interviews, journalistic reportage, and the like. Rather, what we aim to theorize and study is semiosis not as an abstract object to be reconstructed from decontextualized texts but as an actual, situated process borne in, as, and across contextualized events and with some demonstrable reality for a social domain beyond that of the analyst. ↩
Interviews Cited
Balaji Sakthivel (director). 2008. Interview with the author. December 10, 2008, Chennai, India.
Balaji Sakthivel (director). 2018. Interview with the author (phone interview). September 30, 2018, Chennai, India.
Balaji Sakthivel (director). 2018. Interview with the author. October 16, 2018, Chennai, India.
References
Agha, Asif. 2007. Language and Social Relations. Cambridge: Cambridge University Press.
Agha, Asif and Stanton Wortham, eds. 2005. Discourse across Speech Events: Intertextuality and Interdiscursivity in Social Life. Special Issue of Journal of Linguistic Anthropology 15(1):1–150.
Aumont, Jacques, Alain Bergala, Michel Marie, and Marc Vernet. 1992[1983]. Aesthetics of Film. Translated and revised by Richard Neupert. Austin: University of Texas Press.
Austin, J. L. 1962. How to Do Things with Words. Oxford: Clarendon.
Bakhtin, Mikhail. 1982. The Dialogic Imagination. Austin: University of Texas Press.
Bakhtin, Mikhail. 1986. The Problem of Speech Genres. In The Problem of Speech Genres, pp. 60–102. Austin: University of Texas Press.
Benveniste, Émile. 1971. Problems in General Linguistics. Coral Gables, FL: University of Miami Press.
Benveniste, Émile. 1970 [2014]. The Formal Apparatus of Enunciation. In J. Angermuller, D. Maingueneau and R. Wodak, eds. The Discourse Studies Reader: Main Currents in Theory and Analysis, pp. 141–45. Amsterdam: John Benjamins.
Bertrand, Denis. 1989[1984]. Narrativity and Discursivity. In P. Perron and F. Collins, eds. Paris School Semiotics, I. Theory, pp. 105–40. Amsterdam: John Benjamins.
Bertrand, Denis. 1993. L’impersonnel de l’énonciation; Praxis énonciative: conversion, convocation, usage. Protée 21(1):25–32.
Bertrand, Denis. 2000. Précis de sémiotique littéraire. Paris: Édition Nathan HER.
Branigan, Edward. 1984. Point of View in the Cinema. Berlin: Mouton Publishers.
Buckland, Warren. 2000. Cognitive Semiotics of Film. Cambridge: Cambridge University Press.
Bühler, Karl. 1990[1934]. Theory of Language. Trans. D. Goodwin. Amsterdam: Benjamins.
Carr, E. Summerson and Michael Lempert, eds. 2016. Scale: Dimension and Discourse in Social Life. Berkeley: University of California Press.
Casetti, Francesco. 1995[1983]. Face to Face. In W. Buckland, ed. The Film Spectator, pp. 118–40. Amsterdam: Amsterdam University Press.
Casetti, Francesco. 1998[1986]. Inside the Gaze. Bloomington: Indiana University Press.
D'Armenio, Enzo. 2025. The Case of The Last of Us: Analyzing Identity from the Perspectives of Continental Semiotics and Linguistic Anthropology. Semiotic Review, 12. https://doi.org/10.71743/tq8dwq65.
Deleuze, Gilles and Felix Guattari. 1987[1980]. A Thousand Plateaus. Minneapolis: University of Minnesota Press.
Derrida, Jacques. 1976. Of Grammatology. Baltimore, MD: Johns Hopkins University Press.
Derrida, Jacques. 1988. Limited Inc. Evanston, IL: Northwestern University Press.
Dondero, Maria Giulia and Jacques Fontanille. 2014. Semiotic Challenge of Scientific Images: A Test Case for Visual Meaning. Ottowa: Legas.
Dondero, Maria Giulia. 2025a. The Greimasian and Post-Greimasian Semiotics of Photography. In A. Lagopoulos et al., Semiotics of Images: The Analysis of Pictorial Texts, pp. 226–55. Berlin: De Gruyter.
Dondero, Maria Giulia. 2025b. Enunciative Praxis and Enregisterment in the Domain of Generative Artificial Intelligence. Semiotic Review, 12. https://doi.org/10.71743/bwf80g96.
Donzelli, Aurora. 2025. On the Making of Signboards: Corporeal Inscriptions and Material Transpositions of the Writing into the Written. Semiotic Review, 12. https://doi.org/10.71743/eq18s297.
Edwards, Terra. 2017. Sign-Creation in the Seattle DeafBlind Community: A Regeneration of Obviousness. Gesture 16(2):307–32.
Edwards, Terra and Diane Brentari. 2021. The Grammatical Incorporation of Demonstratives in an Emerging Tactile Language. Frontiers in Psychology 11. https://doi.org/10.3389/fpsyg.2020.579992.
Fontanille, Jacques. 1989. Les espaces subjectif. Paris: Hachette
Fontanille, Jacques. 1998. Décoratif, iconicité et écriture. Geste, rythme et figurativité: à propos de la poterie berbère. Visio 3(2):33–46.
Fontanille, Jacques. 2006[1998]. Semiotics of Discourse. New York: Peter Lang.
Fontanille, Jacques. 2017. Praxis and enunciation. Sign Systems Studies 45(1–2):54–69.
Frisone, Marilena. 2023. Eight Dusts: Healing Rituals and Metaphysics of Dust among Nepalese Followers of a Japanese New Religion. E/C, Rivista dell’Associazione Italiana di Studi Semiotici (AISS) 17(39):27–40.
Gaudreault, André. 2009[1999]. From Plato to Lumière. Toronto: University of Toronto Press.
Goffman, Erving. 1974. Frame Analysis. Cambridge, MA: Harvard University Press.
Goffman, Erving. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press.
Gunning, Tom. 1986. The Cinema of Attractions. Wide Angle 8(3–4):63–70.
Hanks, William. 1989. Text and Textuality. Annual Review of Anthropology 18:95–127.
Hanks, William. 1992. The Indexical Ground of Deictic Reference. In A. Duranti and C. Goodwin, eds. Rethinking Context, Language as an Interactive Phenomenon, pp. 43-77. Cambridge: Cambridge University Press.
Hanks, William. 1990. Referential Practice. Chicago: University of Chicago Press.
Hanks, William. 2005. Explorations in the Deictic Field. Current Anthropology 46(2):191–220.
Hansen, Miriam. 1994. Babel and Babylon. Cambridge, MA: Harvard University Press.
Hymes, Dell. 1995[1962]. The Ethnography of Speaking. In B. Blount, ed. Language, Culture, and Society, 2nd edition, pp. 248–82. Prospect Heights, IL: Waveland Press.
Irvine, Judith. 1996. Shadow Conversations: The Indeterminacy of Participant Roles. In M. Silverstein and G. Urban, eds. Natural Histories of Discourse, pp. 131–59. Chicago: University of Chicago Press.
Jakobson, Roman. 1953. Chapter 2. In Results of the Conference of Anthropologists and Linguists (Memoir 8 of the International Journal of American Linguistics), pp. 11–21. Baltimore, MD: Waverly Press.
Jakobson, Roman. 1981[1957]. Shifters, Verbal Categories, and the Russian Verb. In L. Waugh and M. Halle, eds. Russian and Slavic Grammar: Studies, 1931–1981, pp. 41–58. Berlin: Mouton.
Jakobson, Roman. 1960[1958]. Closing Statement: Linguistics and Poetics. In T. Sebeok, ed. Style in Language, pp. 350–77. Cambridge, MA: MIT Press.
Jespersen, Otto. 1922. Language. New York: H. Holt.
Krásová, Eva. N.d. Words Referring to Their Own Utterance: Jakobson and Benveniste on Shifters. https://www.academia.edu/5984378/.
Lagopoulos, Alexandros and Karin Boklund-Lagopolou. 2025a. The Semiotics of the Static Image: The Structuralist Tradition. In A. Lagopoulos et al., : The Analysis of Pictorial Texts, pp. 46–87. Berlin: De Gruyter.
Lagopoulos, Alexandros and Karin Boklund-Lagopolou. 2025b. The Semiotic of the Dynamic Image. In A. Lagopoulos et al., Semiotics of Images: The Analysis of Pictorial Texts, pp. 88–122. Berlin: De Gruyter.
Landowski, Eric. 1989[1983]. Pragmatics and Semiotics: Some Semiotic Conditions of Interaction. In P. Perron and F. Collins, eds. Paris School Semiotics, I. Theory, pp. 95–103. Amsterdam: John Benjamins Publishing Company.
Latour, Bruno. 1999. Petite philosophie de l’énonciation. In P. Basso and L. Corrain, eds. Eloquio del senso. Dialoghi semiotici per Paolo Fabbri, pp. 71–94. Milano: Costa & Nolan.
Lee, Ben. 1997. Talking Heads. Durham, NC: Duke University Press.
Metz, Christian. 1974[1971]. Language and Cinema. The Hague: Mouton.
Metz, Christian. 1991[1971]. Film Language. Chicago: University of Chicago.
Metz, Christian. 1982[1975]. Story / Discourse (A Note on Two Kinds of Voyeurism). The Imaginary Signifier, pp. 91–98. Bloomington: Indiana University Press.
Metz, Christian. 1995[1987]. Impersonal Enunciation, Or the Place of Film (In the Margin of Recent Works on the Enunciation in Cinema). In W. Buckland, ed. The Film Spectator, pp. 140–63. Amsterdam: Amsterdam University Press.
Metz, Christian. 2016[1986]. Théorie de la Communication versus Structuralisme: Sur la Notion de Communication. Mise au Point 8. https://doi.org/10.4000/map.2121.
Metz, Christian. 2016[1991]. Impersonal Enunciation, Or the Place of Film. New York: Columbia University Press.
Nakassis, Constantine V. 2016. Doing Style: Youth and Mass Mediation in South India. Chicago: University of Chicago Press.
Nakassis, Constantine V. 2017. Rajini’s Finger, Indexicality, and the Metapragmatics of Presence. Signs and Society 5(2):201–42.
Nakassis, Constantine V. 2020. Deixis and the Linguistic Anthropology of Cinema. Semiotic Review 9. https://doi.org/10.4000/map.2121.
Nakassis, Constantine V. 2023a. Onscreen/Offscreen. Toronto: University of Toronto Press.
Nakassis, Constantine V. 2023b. A Linguistic Anthropology of Images. Annual Review of Anthropology 52:73–91.
Nambath, Suresh. 2003. The Politics of Art. The Hindu, 28 December 2003.
Padoan, Tatsuma. 2021. Pilgrimage and Spirit Possession: Reconnecting Senses, Discourse and Subjectivity on Mt Kiso Ontake. Irish Journal of Asian Studies 7:39–57.
Padoan, Tatsuma. 2024. Walking the Sutra: A Semiotic Theory of Pilgrimage. In J. Eade and M. Katic, eds. Approaching Pilgrimage: Methodological Issues Involved in Researching Routes, Sites, and Practices, pp. 59–76. London: Routledge.
Pandian, M. S. S. 1992. The Image Trap. New Delhi: Sage Publications.
Paolucci, Claudio. 2020. Wish you were here. Dai Pink Floyd all'enunciazione impersonale. E|C 29:117–27.
Paolucci, Claudio. 2021. Cognitive Semiotics: Integrating Signs, Minds, Meaning and Cognition. Cham, Switzerland: Springer.
Paolucci, Claudio. 2022. Face and Mask: “Person” and “Subjectivity” in Language and through Signs. International Journal for the Semiotics of Law 35:1257–74.
Parret, Herman. 1989. Introduction. In P. Perron and F. Collins, eds. Paris School Semiotics, I. Theory, pp. vii–xxvi. Amsterdam: John Benjamins.
Prasad, M. Madhava. 2014. Cine-Politics. Chennai: Orient Blackswan.
Rajangam, Stalin. 2016. Tamil Cinima [‘Tamil Cinema’]. Chennai: Pirakñai.
Silverstein, Michael. 1976. Shifters, Linguistic Categories, and Cultural Description. In K. Basso and H. Selby, eds. Meaning in Anthropology, pp. 11–55. Albuquerque: University of New Mexico Press.
Silverstein, Michael. 1981. Limits of Awareness. In Working Papers in Sociolinguistics 84, pp. 2–31. Austin, TX: Southwest Educational Research Laboratory.
Silverstein, Michael. 1993. Metapragmatic Discourse and Metapragmatic Function. In J. Lucy, ed. Reflexive Language, pp. 33–57. Cambridge: Cambridge University Press.
Silverstein, Michael. 2003. Indexical Order and the Dialectics of Sociolinguistic Life. Language & Communication 23(3–4):193–229.
Silverstein, Michael. 2004. “Cultural” Concepts and the Language-Culture Nexus. Current Anthropology 45(5):621–52.
Silverstein, Michael. 2017. Forty Years of Speaking (of) the Same (Object) Language—Sans le Savoir. Langage et société 160–161(2–3):93–110.
Silverstein, Michael. 2021. The Dialectics of Indexical Semiosis: Scaling Up and Out from the “Actual” to the “Virtual.” International Journal of the Sociology of Language 272:13–45.
Silverstein, Michael. 2022. Language in Culture: Lectures in the Social Semiotics of Language. Cambridge: Cambridge University Press.
Silverstein, Michael and Greg Urban. 1996. The Natural History of Discourse. In M. Silverstein and G. Urban, eds. Natural Histories of Discourse, pp. 1–18. Chicago: University of Chicago Press.
Srinivas, S. V. 2009. Megastar. Delhi: University of Oxford Press.
Srinivasan, Perundevi. forthcoming. Love Meets Death: “Honour,” Violence, and Inter-caste Marriages in Tamil Nadu. In C. Nakassis, E. Annamalai, and F. Cody, eds. The Radiance of Tamil: Ten Years of the Chicago Tamil Forum. Hyderabad: Orient Blackswan.
Urban, Greg. 1989. The “I” of Discourse. In B. Lee and G. Urban, eds. Semiotics, Self, and Society, pp. 27–51. Berlin: Mouton de Gruyter.