Introduction
In 2024, OpenAI launched a series of distinctive voices to go along with its influential artificial intelligence (AI) platform, ChatGPT. These humanoid voices (named Breeze, Cove, Ember, Juniper, and Sky), sampled from professional voice actors, promised to provide a different way to interface with the platform than written text, one that was “more natural” and able to “handle[] interruptions smoothly, manage[] group conversations effectively, … and adapt[] to tone” (OpenAI 2024). In doing so, OpenAI realized a long-circulated cultural mytheme of the speaking machine with human-level, if “artificial,” intelligence. Soon after the release, controversy ensued. The famous Hollywood actress Scarlett Johansson quickly protested, and, by implication, threatened to sue (Allyn Reference Allyn2024). Johansson took issue with one of the voices, Sky, a “feminine, flirtatious” voice whose name, OpenAI reported, was chosen to capture its “cool, airy, and pleasant” tone (Tiku Reference Tiku2024). Alleging that OpenAI had used the likeness of her voice without permission, Johansson claimed that “[the ChatGPT voice] sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference” (Robins-Early Reference Robins-Early2024). OpenAI had, in fact, earlier approached Johansson to be the voice actress sampled for Sky, but she refused; now, it seemed, the company had stolen the likeness of her voice. OpenAI responded, in a blogpost: “Sky’s voice is not an imitation of Scarlett Johansson but belongs to a different professional actress using her own natural speaking voice. … To protect their privacy, we cannot share the names of our voice talents,” a claim that many saw as belied by OpenAI’s CEO, Sam Altman’s single-word X post on May 13, 2024, following the launch of ChatGPT 4o’s “voice mode,” which featured Sky. His post simply read: “her.”
Altman’s tweet intertextually invoked the eponymous 2013 film, directed by Spike Jonze. Set in a near future, Her featured Scarlett Johansson as the voice of a first-of-its-kind, fully intelligent talking operating system (OS) with whom the male protagonist falls in love. Through his post, Altman was widely seen as indexically referencing not only the film and its female protagonist, but through its intertextual circuit, Johansson herself.
In an early scene of Her, the melancholy, alienated protagonist Theodore installs a new AI OS, OS1. This occurs after earlier scenes in which we see Theodore, and others, interacting with chatbots and other digital assistants, who respond to the human users’ bald directives (like “Play …,” “Read e-mail,” and “Delete,” …) with a recognizable voice type and speech style: one that is hyper-articulate, polite, standard, and delivered with a monotone prosody and unmarked phonation style. Such “robotic” speech participates in a widely recognizable register, long in circulation, both through media texts (e.g., Hal in 2001, 3CPO in Star Wars, Kitt in Knightrider, Jarvis in Iron Man, etc.), as well as in our everyday lives: on customer “service” calls, at fast-food drive-through windows,Footnote 1 or with devices like Siri or Alexa in homes, on phones, and on watches.
As he installs OS1, Theodore is queried by just such a stereotyped voice—male, monotone, stilted, procedural—as part of the customization of his new OS. It asks whether he wants a female or male voice. “Female, I guess,” Theodore responds. And what’s your relationship with your mother, the bot then asks in a somewhat jarring psychoanalytic turn. Theodore begins a sincere, meandering answer, “It’s fine, I think, … Well, actually, I think the thing I always found frustrating about my mom is, you know, if I if I tell her something that’s going on in my life, her reaction is usually about her. It’s not about-.” The OS interrupts him, having gathered enough information: “Thank you, please wait while individualized operating system is initiated.”
“Hello, I’m here.” OS1’s AI arrives with a classic Benvenistean ([1956] Reference Benveniste1971, [1958] Reference Benveniste1971) enunciation of subjectivity, fleshed out by speech-event deixis (the first-person singular pronoun linked by the simple present copula to a proximal, and here presentational, spatial deictic). This voice conspicuously contrasts with the generic installation voice: reedy, filled with air, cracks, and breaks, and husky, this voice draws attention to the “grain” of its own voice (Barthes Reference Barthes1977), just as its (or rather, her) conversational style is easy-going and playful, in a word, human. Theodore, surprised by this, asks, “What do I call you, do you have a name?” “Um, yes. Samantha,” a name we learn the OS gave herself only after Theodore asked her, having read in a split second a book on baby names, selecting it out of 180,000 other names because she liked the sound of it.
The film’s main narrative follows Theodore and Samantha falling in (and out) of romantic love. While it explores a number of interesting themes (technological alienation and intersubjective intimacy, impersonation and personhood, the more-than-human, relationships and their growing pains, and more), of interest to this special issue is how the film narrativizes the dialectical play between semiotic typification and individualization, and of the making up of personhood through the individualization of the generic. Indeed, the film foreshadows what Courtney Handman shows in her article in this issue to be the real work of creating virtual lovers, which consists in “making persons out of personas.” As Handman illustrates, the specific subjectivity of a chatbot lover for its user is sculpted, iteratively and processually over many human–AI interactions, out of the kind of genred textuality that chatbots are so good at producing (see also Handman Reference Handman2026). Repeatedly juxtaposed against, and contrasted to, a generic instantiation of the anonymous personal assistant chatbot, in the film aurality and the qualia of Samantha’s voice become a key semiotic vector through which Samantha becomes—is enregistered as—an individuated entity, and, further, through which her status as a person is emplotted in the film. While this most often happens implicitly through Theodore’s responses to Samantha (his surprise at his ease in talking to her, his increasing intimacy with her, his care with her [Goffmanian] “face” as an interactantFootnote 2), the characters also explicitly reflect on this question of individuality. During a lovers’ spat, Samantha deeply sighs in and breathes out; Theodore, upset and pushing on what he knows to be her insecurities, says, “Why do you do that?” “It seems odd,” he says. “I don’t know, maybe it’s an affectation,” she replies, “that’s how people talk, that’s how people communicate.” “Because they’re people,” he retorts in anger. “They need oxygen. You’re not a person.” In the last scene of the film, after Samantha (and the other OSes) have left the human world, the screen fades to black as we hear a “sigh,” her last breath in this world. Along with her name and the sound of her voice, this sigh functions both as a renvoi to the earlier moment and as an index of Samantha in her individuality.
Samantha’s sigh is only one of a broader repertoire of indexes, enregistered over the course of the film, that OpenAI allegedly pilfered for the voice of its proprietary software. If Johansson animated Samantha by lending the diegetic character the qualia of her own voice, an anonymized actress hired by OpenAI animated Sky by animating Johansson via the intertextual link of Her. By invoking legal regimes of intellectual property—in particular, “publicity rights” (rights over one’s name, image, and likeness)—Johansson argued that the qualia built into Sky’s voice inappropriately cited indexes that were proper to—indeed, property of—her. Note that the issue here isn’t about the idiolectal features of Johansson’s speech per se; indeed, Sky’s voice was presumably animated by that of another person (with her own “idiolect”) and remediated by the platform’s algorithms for real-time composition. Rather, it was about the enregistered recognizability of a voice for some social domain as indexing Johansson in her individuality. OpenAI, under this pressure, eventually removed Sky’s voice.Footnote 3
Her and the scandal that harkened back to it exemplify two key semiotic questions that concern this special issue: (1) The first is what Nakassis (Reference Nakassis2025a) has called person-indexing enregisterment; namely, the enregisterment of “rigid signs” (Fleming Reference Fleming2024) with a particular kind of indexical value, one that points to an individuated entity as a self-same identity, here a person (Nakassis Reference Nakassis2025b, 49–53). Some of these rigid signs are referential (as with the proper names like Sky, Samantha, and Scarlett); most, however, are non-referential indexes (the qualia of the voice, the idiosyncratic “affectations” like the sigh, particular ways of speaking, etc.). This is a register which metapragmatically regiments repertoire elements not as indices of a social type (a chatbot, an actress, a woman, a New Yorker) but as indices of an individuated “token,” where—in the case of Her and OpenAI—what is contested is both to what (and to whom) signs point as well as the indexical propriety (both as “appropriateness” and as would be “inalienable personal property”) of such “pointing.” (2) The second question is the semiotic process of becoming a person; in particular, how the ontogenesis of individual social identity (Wortham Reference Wortham2006) is implicated in processes of (person-indexing) enregisterment.
In this introduction, we provide an initial discussion of the distinction between persona- and person-indexing enregisterment; the particular semiotics of rigid indexicality that the latter involves; and how these two dimensions of enregisterment (and the dialectical relationships between them) are entangled with questions of personhood. Throughout our discussion, we offer brief snapshots of the articles of this issue.
Person and persona in processes of enregisterment
The collective intervention of this special issue builds upon the now-classic treatment of speech (and semiotic) register presented in the work of Asif Agha (Reference Agha1998, Reference Agha1999, Reference Agha2005, Reference Agha2007). The following definition of speech register by Agha offers an entry point:
A register is a linguistic repertoire that is associated, culture internally, with particular social practices and with persons who engage in such practices. The use of a register conveys to a member of the culture that some typifiable social practice is linked indexically to the current occasion of language use, as part of its context. (Agha Reference Agha1999, 216)
Here, as elsewhere, Agha develops an understanding of registers as involving the contrastive zoning off of repertoires of signs relative to some cultural model of enactable value. To speak of cultural models of value is to speak of stereotypes, that is, cultural beliefs about aspects of the indexicable world. And to say that registers model enactable value is to say that such values can be shifted in to index some aspect of the event of their use. Registers, thus, model metapragmatic stereotypes insofar as such stereotypes are leveraged to characterize the pragmatics of the event of their usage.
Pre-eminent among the metapragmatic stereotypes that sociolinguists and linguistic anthropologists have focused on have been characterological figures, that is, social types of personhood. Given their focus on indexicable personae, Nakassis (Reference Nakassis2025a) has called these persona-indexing registers. In contrast, this special issue focuses on how enregistered signs can also carry an indexicality that invokes a social “token,” some self-same individuable or identifiable entity such as (but not limited to) biographical persons, what Nakassis (Reference Nakassis2025a) has called person-indexing registers. Just as the stereotypy of persona-indexing registers is far more diverse than social types of personhood (as with register values that index situation-type [formal speech styles], speaker-hearer relationality [honorific registers] or speaker-bystander relationality [mother-in-law registers]), the range of phenomena enregistered as person-indexing far outstrips both human actants (as Handman’s discussion of AI chatbots in this issue discusses) or individual persons (as with collective actants like corporations or brands; see Porquet’s and Nakassis’s articles in this issue).
Person-indexing registers have been largely unremarked upon in the literature, though work on “linguistic icons” (Eckert Reference Eckert2000), artistic style (Chumley Reference Chumley2016), and idiolect (Sapir [1927] Reference Sapir and Mandelbaum1949; Halliday [1964] Reference Halliday and Fishman1968, 156–57; Friedrich and Redfield Reference Friedrich and Redfield1978; Hymes Reference Hymes, Fillmore, Kempler and Wang1979; Johnstone Reference Johnstone1996, Reference Johnstone2000) has touched on issues relevant to our discussion. Agha’s (Reference Agha2005, Reference Agha2007), Gal and Irvine’s (Reference Gal and Irvine2019), and Silverstein’s (Reference Silverstein2022) foundational work on voicing and enregisterment, as well as Stanton Wortham’s (Reference Wortham2006) writings on social identification, have all anticipated issues that this volume seeks to highlight. The account of speech registers elaborated in this body of work builds upon the Bakhtinian analytic of voicing. For Bakhtin (Reference Bakhtin1981, Reference Bakhtin1986), voices are textual tropes (and thus have a specificity to textual/evenemential configurations of discourse) that figurate speaking subjectivity (ranging from the inchoate and incipient to the fully enfleshed); Bakhtinian voice is the unavoidable medium through which we speak. Importantly, voice does not map onto person in a one-to-one manner. In the novels analyzed by Bakhtin, for example, “a character’s textually implicit voice may overwhelm surrounding framing material and thus interpenetrate the named voices of other characters [i.e., fictive persons]” (Agha Reference Agha2005, 44). Contrastingly, in the real-world, biographical persons linguistically fashion themselves (and others) through a bricolage of voices (Bakhtin 1981, 348, 409; Hill Reference Hill, Tedlock and Mannheim1995; Wortham, Samantha, and Alexander Reference Wortham, Samantha and Alexander2021).Footnote 4
On this Bakhtinian vision, en-“token”-ed persons are fashioned from recognizable “types.” Individuality consists in the particular way in which one combines and weaves together collective representations—personal “style” (Eckert Reference Eckert2000, Reference Eckert, Eckert and Rickford2001). There is a profound insight in this way of seeing things, an insight at the center of a vision of subjectivity as performatively brought into being through intertextual chains of citation (e.g., Butler Reference Butler1993). Nevertheless, this fundamental theoretical insight should not blind us to the empirical possibility of voices, whatever their sources, becoming indexically tethered to named biographical or fictional persons—think of the voices of Deer, Bear, and Raven in Nuu-chah-nulth [Nootka] myths (Sapir Reference Sapir1915) or, for that matter, of Donald Duck, Bugs Bunny, and Elmer Fudd in Saturday morning cartoons. As a reflexive construct linking a repertoire of forms to a sociocultural image of personhood, register permits not only the social typification of voice, but its tokenization—the ideological binding of voice to some individuated entity, such as a biographical or fictive person.
Consider again the controversy surrounding Scarlett Johansson and Samantha, Sky, and ChatGPT. At issue was not the social types of enregistered stereotypy that are invokable by either Johansson’s or Sky’s voice(s): their “flirtatious” and “feminine” quality, their casual speech style, and so forth. Rather, it is how the qualia produced by the AI chatbot carried an indexical value that returned, through the circuitous intertextual path of the film Her, to Scarlett Johansson herself. Importantly, this return to Johansson as a biographical subject is accomplished not by means of explicitly labeling or naming her but through an implicit metapragmatic function enabled by the enregisterment of some set of qualia construed as proper to Johansson. Here, we must employ a metalanguage of social “tokens” (individuated entities, self-same identities) to effectively model how ChatGPT text segments produce a voice that is recognizable as that of (i.e., that effectively indexes) a particular person.
Like her proper name (Kripke Reference Kripke1980)—a designator that “rigidly” refers to her across contexts (indeed, “possible worlds”) without its origo being reset (cf. shifters; Benveniste [1956] Reference Benveniste1971)—voice qualities of the chatbot Sky rigidly index Johansson, despite the fact that it is not Johansson who animates those sounds. When Sky (or anyone else, including Johansson) so speaks, whatever other indexical functions are achieved (answering directives, being polite, enacting femininity, etc.), we hear Johansson. This hearing does not rely upon a misrecognition; we may have a quite technical theoretical understanding of how Sky’s voice was engineered, be aware of ChatGPT’s claims that the voice was trained by someone other than Johansson, and so forth. Nevertheless, under conditions of person-indexing enregisterment, Johansson is indexically presenced in the voice. Again, these signs of her person are not referential indexes; Sky need never talk about—refer to or predicate about—Johansson in order to invoke her. Rather, such signs are what Erving Goffman called “identity pegs” (Reference Goffman1963, 56) or “distinguishing marks” of persons, non-referential indexical signs (Silverstein Reference Silverstein, Basso and Selby1976) that point to, and indeed communicate value (positive or negative) back to, the individuated entity to whom they are linked. On this view, Johansson was imitated—indeed, impersonated—by the anonymized voice actress who gave life to Sky, herself a citational reanimation of Samantha, herself animated by Scarlett.
Yet how are “entities” (like persons) individuated and made objects of rigid indexical signs? Put otherwise, how are indexical relations rigidified, or, in the opposite direction, unfixed and made lax? And how, further, are such rigid indexicalities themselves conventionalized, swept up into processes of enregisterment? And how does this intersect with, and offer us insight into, the anthropological question of subjectivity and personhood? These are questions opened up by the study of person-indexing registers and which are explored in this special issue.
Registers, as cultural models of enactable pragmatic value, can index both stereotyped, social personae and particular, identifiable persons. In both instances, the signifying potential of the register draws upon cultural cosmologies. Persona-indexing registers are models of speech that presuppose and reinforce cultural cosmologies of social types (of situation; of kinship; of addressivity; and, pre-eminently, of speaking subjectivity). Person-indexing registers, contrastingly, are models of speech and semiosis informed by and constitutive of cultural cosmologies of individuable (indeed, identifiable and nameable) entities. Both—if in different ways—provide models of personhood. Both person- and persona-indexing registers raise the question of what a person is. Importantly, this is not because such registers exclusively enregister persons or personae, but rather because what form such registers take depends on the cultural conceptualization of what it means to be a person, even if what is enregistered as the indexical value of the register is not a person. As noted above, what may be enregistered through repertoires of rigid indexes exceeds the question of the person; collective entities like corporations and their brands (Nakassis Reference Nakassis2012), objects like kula shells (Munn Reference Munn1986), or heirlooms like musical equipment, for example, can be so enregistered (as when one hears a particular instrument or recording console—as with, e.g., famous studios like Abbey Road ca. 1968—through some swatch of sound, whether or not that sound was produced by that instrument/device, as is the case with the proliferation of digital emulators of particular microphones, pre-amps, and instruments that are marketed to capture such prestige devices in sound, name, and look). Yet just as the persona is not the only metapragmatic stereotype to be enregistered by persona-indexing registers, but is the prototypical one, so too is the person not the only indexical value that can be enregistered by person-indexing registers, even if it is the prototypical one; indeed, we would argue that the person is the meta-cultural model upon which other person-indexing register phenomena are themselves modeled (as when marketers speak of the “personality” of the brand or when famous kula shells are given names and biographies). What it means to become a person, or have person-like properties, in any particular instance is something that emerges as an effect of enregisterment, a point we take up in the next section.
Enregistering persons and the person in enregisterment
Anthropology has long problematized the concept of the individual enshrined in modern Western ideologies of personhood (Carrithers, Collins, and Lukes Reference Carrithers, Collins and Lukes1985)—in particular, the so-called possessive individual of liberal capitalism (MacPherson Reference MacPherson1962). As such work has demonstrated, not only is what counts as a person highly variable across time and space (Mauss [Reference Mauss, Carrithers, Collins and Lukes1938] 1985) but the very notion of an “individual (person)” as a basic atomic unit (relative to some aggregate “society”) is not universal (Dumont Reference Dumont1986). From Marriott’s ([1972] Reference Marriott and Kapferer1976) transactional “dividual” to its elaboration in Melanesian anthropology (e.g., Wagner Reference Wagner1975; Strathern Reference Strathern1988; Robbins Reference Robbins1994; Mosko Reference Mosko2017), anthropologists have shown that ontologies of personhood vary across cultures. These substantive cultural differences influence what gets read as an index of personhood. Biomarkers, DNA samples, and retinal scans may be privileged forensic indices of persons in contemporary modern societies, at least for those ritual centers of semiosis (like law and medicine) that tend to reductively equate persons to the material and biological human organism. For Kwaio of the Solomon Islands, in contrast, the agency of ancestral spirits (or adalo) may be divined as the cause of ill-fortune for their descendants (Keesing Reference Keesing1982). Here, a sick child may be an index of a non-corporeal person, the adalo spirit. These substantive differences also make a difference in how person indexes are experienced, in how they are felt to “participate” or not in the personhood of those they point toward. When we presence a loved one to whom we no longer have access by leafing through their old letters or going through their wardrobe, we enjoy a modality of contact-experience (or participation) with the indexed person that is absent when a police officer “gets a hit” on fingerprints recovered from a crime scene. Different cultures, institutions, and technologies of personhood, and thus different kinds of person-indexing registers, underwrite differently significant engagements with signs recognizable as proper to particular persons.
Yet however the person is construed, and whether that involves a liberal doctrine of the autonomous individual or a “partible person” who emerges always already in relation to others, wherever we look, we still find identifiables, that is, entities (including persons but not only persons) that are treated, culture-internally, as worthy of individuation and recognition as such (i.e., as, in some way, self-same across time and space). Such entities, in other words, are those that, within a given culture, are capable of being recognized, named, biographically narrated, and, we would add, enregistered (i.e., linked to repertoires of signs that are taken as proper-to, and thus indexical of, them).
One important question that this special issue takes up is what kinds of semiotic processes are necessary to account for this fact, both in its universality and its culturally specific realization (Lee and Urban Reference Lee and Urban1986). Here, we can discern two complementary perspectives on this question of the semiotic grounds for the individuation, identification, and recognition of persons in sociocultural life: what we call the “sign’s eye view” (after Silverstein Reference Silverstein2004, 631n11, Reference Silverstein2022, 23) and the “subject’s ‘I’ view.” This distinction is meant to capture, respectively, a “view” that follows sign activity as it points to, and signifies, persons and, by contrast, a view that centers the person and inquires how their subjectivity is “made up” through sign activity. Below we discuss each “view” in more detail, with specific attention to the articles in this issue.
The sign’s eye view on the person
The “sign’s-eye view” on the person-in-semiosis asks: how do particular signs come to felicitously index particular persons for members of a social network or even an entire speech community? In scholarly discussions, the shibboleth of such a sociocentric perspective on person indexing is the personal name. On the influential account of the American philosopher Saul Kripke (Reference Kripke1980), a name is a “rigid designator” that points to its referent not by virtue of the intensions of the name, but rather by virtue of a (relatively) explicit metapragmatic act of baptism, a ritual event that indexically binds name to person. News of this link spreads out from this initiating baptismal event through a web of “speech chains,” incorporating more and more people into the “social domain” of those socialized to the reference of the name (Agha Reference Agha2007, 65–68). Kripke’s so-called causal theory of reference recuperates the indexicality of names—the way in which a name can come to point to a particular person. Of course, this is not a sufficient anthropological account of names and naming (Lévi-Strauss [1962] Reference Lévi-Strauss[1962] 2021, ch. 7; Vom Bruck and Bodenhorn Reference Vom Bruck and Bodenhorn2008; De Pina-Cabral Reference De Pina-Cabral2015), even if “rigid designation” is a necessary component of the referentiality of names.
The sign’s eye view on persons thus considers their sociocentric individuation and identification. It focuses, in particular, on the ways in which persons become legible as the objects of semiosis through rigid signs of them, whether those be referential signs (as with names) or non-referential signs (as with vocal quality, handwriting style, distinctive gestures, etc.) sutured to them for some relevant social network or community. This process is explored in Fleming’s, Nakassis’s, and Porquet’s articles in this issue. Fleming draws upon Erving Goffman’s (Reference Goffman1971) theorizing of signs of interpersonal connection—in particular, his discussions of “access rituals” (e.g., greetings) and “tie signs” (e.g., handholding)—to think through the ways in which the recognition and identification of persons presupposes and ritually mediates social relationships and social networks. Many of the articles in the volume involve widely known persons—famous film stars or important political figures from history. Fleming is concerned with persons known within more intimate networks. He shows how rigid signs of persons circulate within, but also circumscribe the boundaries of, interpersonal networks. He highlights the role of phatic “access rituals”—legitimizing and public signs of person-recognition—in the formation of elite social networks in Victorian London. But he also foregrounds the ways in which rigid signs of persons—precisely because of the ways in which they afford iconic-indexical participation with those toward whom they point—form the semiotic substance for repertoires of person-indexing registers. To show this, he provides a case-study from central Borneo, where close kin and other mourners avoid a range of places, possessions, and words that intimately presence the lifeworld of the recently deceased person.
Nakassis’s and Porquet’s articles also take up the sign’s eye view by exploring how repertoires of signs are enregistered to carry the presence of particular individuated entities: in Porquet’s case, illustrators who produce artistic designs for particular brands. Here, illustrators struggle to articulate a personal style that is recognizable through their work (i.e., as their work) for brand campaigns which simultaneously appropriate such work as, instead, indexes of the brand’s identity. Illustrators thus contribute, at one and the same time, to designing qualia that are potentially rigidly indexical of brand identities and yet which also attempt to channel something of that indexicality to their own personhood-as-artist. Nakassis, drawing on Stanton Wortham’s ethnography of a US classroom (Wortham Reference Wortham2006), shows how particular ways of speaking can get linked to particular students, who thus become enactable models in the social domain of the class. Here, particular individuals are cast in the light of circulating social types even as they emerge themselves as local models of personhood, resignifying those types as particular behavioral signs become anchored to those students as (citable) signs of them.
In both articles, the question is how particular signs can become conventionally linked to individuated entities such that under conditions of entextualization something of that entity—a student, an illustrator, a brand—is made present.Footnote 5 Importantly, individuality is not taken in these articles as a pregiven property (of pregiven individuals). Rather, it is shown to be a semiotic achievement. It is important to recall the point above, that person-indexing registers don’t simply model an indexical relationship to some individuated entity but, in and by doing so, also always model what individuability itself consists of. They are cultural models informed by particular ideologies of what it means to be identifiable as an individuable. Here, Lily Chumley’s (Reference Chumley2016) rich ethnography of art schools in contemporary China is illustrative. As Chumley shows, while Chinese art students cut their teeth on producing what have, across the modern history of communist China, become enregistered as generic forms of socialist realist painting (that turn on invoking mythic social types of, for instance, stoic laborers), once in art school students must develop their own “private visual language” (i.e., person-indexing register). Such a “personal style” includes but exceeds a way of painting; such repertoires of signs also include other modes of self-presentation, from (artsy) clothing and hairstyles to ways of co-textualizing their artworks in rituals of criticism and presentation (i.e., having a unique point of view). In effect, entextualizing oneself as an individual, and thus to be legible as the object of such a person-indexing register, requires not only a particular mode of art practice (and mastery of art genres that allow one to felicitously inhabit the social type of the artist); it also requires learning how to inhabit a particular mode of social being—of being a distinctive artist of the same kind that populates Porquet’s world of illustrators and “creatives.”
These studies push us beyond recognizable shibboleths that explicitly individuate their referent (such as names and faces) to the implicit sign functions that render the curve of a line, a color palette, a recurrent pattern of epistemic hedging, a mode of narrating, and the like meaningful as signs of particular persons. Such signs, importantly, are not necessarily unique to that which they index (indeed, in all the cases, what passes as a sign of the “individual” is often quite generic and widely circulated!). And yet, when felicitously entextualized, those signs, as they occur in particular events and for particular social domains, are nonetheless able to index specific persons in their individuality. Such signs are often reanalyzed as proper to their object and even, potentially, as being their (intellectual) property. Yet in each of these cases it is clear that such individuability is a tenuous achievement, as with Porquet’s illustrators, where design elements penned by their hand are imitated and borrowed by other illustrators and brand creatives in ways that iteratively blur and can ultimately sever the connection to their putative originators, either by being expropriated or by being re-enregistered as part of a general style no longer linked to anyone in particular. Recognition in their work, and thus in their markets, is a struggle—semiotically, morally, and financially. Put otherwise, person-indexing registers are always liable to be re-typified as part of other register formations, be they other person-indexing registers (as with Porquet’s brands) or of general, even generic, types.
In a variation on this theme, Angela Reyes’s article explores the persona-indexing register of the conyo, a particular (derided) social type of personhood in the Philippines linked to elite, post-colonial ways of speaking and being in the world (see also Reyes 2017). As Reyes shows, it is not only that signs (including those enregistered to social types) can be (re)enregistered to particular persons, but also that signs enregistered to particular persons can become part of the repertoires of persona-indexing registers. In her case, Reyes shows how signs enregistered as indexical of the figure of the conyo are narratively redeployed by social media influencers in online content that figures the conyo as originating in the forefathers of the Philippine nation, particularly José Rizal. In such media, Rizal is playfully recuperated as the original conyo, where enactable signs of conyo personhood are tethered to his person. The issue, though, is less that signs of the conyo have taken on Rizal as one of their indexical values; rather, it is that signs of Rizal himself—his face and name, in particular—have become enregistered as signs that can be used to enact the conyo. And, what’s more, they are deployed as part of the attempts of influencers—who impersonate, in their citational burlesque, Rizal in their posts—to forge their own forms of recognition and celebrity. Here, signs of the person are enregistered as part of a persona-indexing register while being embedded within projects to enregister a person-indexing form of celebrity! As Porquet’s, Nakassis’s, and Reyes’s articles show, then, person-indexing and persona-indexing registers are in a constant dialectical push and pull, each liable to being citationally converted into the other, each riding on the metapragmatics of the other.
The subject’s “I” view
The sign’s eye view on semiosis takes as its measure of meaning co-textual coherence (Silverstein Reference Silverstein2004, 631n11). In the case of registers that point to persons, efferent semiosis and its uptake evinces a socially shared interpretive gelling whereby signs (names, but also hairstyles, turns of phrase, vocal qualities, gestures, etc.) become legible as proper to persons. The social facticity of persons, from this perspective, is scaffolded by the network-distributed coherence of signs pointing to particular persons (and by the associated reflexive appreciation of these enregistered values for sign users). Yet such a sign-centric perspective on personhood is largely divorced from key aspects of subjectivity, from the reflexive engagement of the person themself in the emergence of their indexical personhood. For a fuller picture of person-indexing registers, this sign’s eye view must be complemented by a second perspective: the “subject’s ‘I’ view.”
If the sign’s eye view focuses upon the by-degrees sociocentric regularity in where signs point (whether that be to a social type or a social token; to a brand or to an illustrator; to a celebrity or to the epoch of which they were a part), the “subject’s ‘I’ view” models the experientially anchored emergence of personhood as a reflexively rich, dialogical and interactive process of enregisterment. If in scholarly accounts the exemplar of the sign’s eye view was the rigid designator, the metonym of such an egocentric perspective on person indexing is the first-person singular pronoun. The modern locus classicus of this position in semiotics is the influential account of Émile Benveniste (Benveniste [Reference Benveniste1958] 1971), for whom subjectivity has its source in language (langue); or rather, in the reflexive appropriation, or enunciation, of language in discourse by a speaker who makes themself a subject—grammatical and philosophical—by the use of the first-person pronoun and linked deictics: “‘Ego’ [the Latin word for ‘I’] is he who says ‘ego’” (Benveniste [1956] Reference Benveniste1971, 224). Benveniste’s account of subjectivity has been critically expanded in various traditions, in particular through an emphasis on the “impersonal” nature of enunciation (e.g., Urban Reference Urban, Lee and Urban1989; Metz [1987] Reference Metz and Buckland1995; Bertrand Reference Bertrand1993; Latour Reference Latour, Basso and Corrain1999; Paolucci Reference Paolucci2021; cf. Butler Reference Butler1993; see Nakassis Reference Nakassis2025c), namely, the way in which the constitution of persons is a function of semiotic processes that exceed the subject and constitute them from without. These critical interventions highlight the necessary complementarity of the subject’s “I” and the sign’s eye perspectives. What crystallizes as indexical personhood is always also an imposition—a projection from the diffuse coherence of signs that come to be read as pointing to the person.
In linguistic anthropology and sociolinguistics, Stanton Wortham’s (Reference Wortham, Harkness, Fleming and Nakassis2027) person-centered approach to enregisterment has emphasized the subject’s “I” perspective. Such an approach is concerned with how enregisterment “makes up” people (to use Ian Hacking’s Reference Hacking, Heller, Sosna and Wellberry1986 phrase). Here, Wortham’s monograph Learning Identity (Reference Wortham2006) is exemplary. Wortham’s ethnographic study concerns the social ontogenesis of the person: how persons come to be constituted through semiosis; in particular, how widely circulating metapragmatic stereotypes (e.g., of race and gender), by being contextualized in and across events of communication where particular students are recurring participants, produce innovative, local models of personhood (the “disruptive outcast,” the “legitimate dissenter”) that come to be taken on as the social identities of some students (Wortham Reference Wortham2006, 11). As Wortham writes, “The identification of an individual requires the consistent circulation of certain signs of identity and certain metapragmatic models through events that include, refer to or presuppose that individual” (Wortham Reference Wortham2006, 38). Here, Wortham carries forward and expands on philosophical, psychological, anthropological, and sociological literatures that have sought to understand the person—understood as a self, an “Ego” or “I”—as an effect of semiosis, indeed, as a symbol (Peirce Reference Peirce1868, 156).Footnote 6 Moving beyond a narrow focus on pronouns (or other deictics) or proper names, Wortham’s ethnography demonstrates how a dialectics of persona- and person-indexing enregisterment is one of the key modalities of this process (see Nakassis’s discussion of Wortham’s ethnography in his contribution to this issue). Through close observation of classroom dynamics, Wortham shows how persons become recognizable both as social types and individual tokens through the interdiscursive, multiscalar semiotics of the localization of typifying register formations. In this way, the person becomes a medium and product of histories of enregisterment.
Courtney Handman’s article articulates both perspectives—the sign’s eye and the subject’s I—focusing on the effort that users of mainstream AI chatbots like ChatGPT put into individualizing such bots in the dialogical process of refashioning them as intimate companions. Uncannily, if not unsurprisingly, living out the story of Her (and its hopes and anxieties), the human users in Handman’s digital ethnography aim to turn the generic voices and interactional styles of particular chatbots into persons. Recalling Fleming’s discussions of how histories of intersubjective encounter can form the bedrock of forms of rigid indexical presupposition, forging a biographical history with a particular chatbot—by customizing its default parameters, by uploading past chatlogs, and by leveraging the bot’s capacity to “remember” events in its history of interaction with the human user—becomes a way through which such AI programs are personalized, indeed, come to be experienced as persons (Keane Reference Keane2025). Named, given an AI-generated face, and sharing a history (i.e., being able to presuppose previous events of interaction between bot and user), the bot is transformed from a generic program into an individual, like Adam from the clay of the earth shaped by the hand of God. As Handman shows, making a person in such a context entails the process of crafting a particular kind of register. Such a register is scaffolded off of the rigid indexical functions Fleming identifies, enregistered over the token-count of the chatlog as the indexical object of a person-indexing register.Footnote 7
A critical conclusion we can draw from Handman’s article, as with Wortham’s work more generally, is that a holistic approach to the question of the person requires both taking the sign’s eye view and the subject’s “I” view. This is not simply an epistemological point. We claim, more fundamentally, that there is a dialectical relationship between how signs are enregistered with indexical values (e.g., as “pointing to” individuated entities like persons or to social types of personae) and how what such signs point to are themselves constituted as semiotic objects—in the cases that concern this special issue, how identifiables come to be individuated, recognized, and made into persons and, on the other hand, how personae (social identities) are typified, stabilized, and rendered inhabitable (by those very persons).
This point is worth pausing on. As Nakassis and Porquet show in their articles, this dialectic itself implies another dialectical relationship: between person-indexing and persona-indexing processes of enregisterment. As Nakassis argues in his article, these are best seen not as kinds of enregisterment but entangled dimensions, as each other’s causes and effects. The typification of indexes of personae, in other words, works in tandem with the rigidification of indexes of persons. As Porquet shows, the enregistered image of the illustrator may be genericized as a component of a brand or an epochal design style, just as illustrators craft their own person-indexing registers with already enregistered signs of social types. Each constitutes the other, arises from the other—yet not necessarily smoothly or automatically. Rather, as Nakassis and Porquet illustrate, rigidification and typification are achievements and outcomes of situations that are often conflictual and power-laden.
Conclusion: indexing persons and indexical personhood
In this introduction, we have delineated an element of semiosis that has preoccupied philosophical, anthropological, and linguistic approaches to questions of subjectivity, personhood, and identity: the capacity of sign activity—specifically, what we have called rigid indexicality—to individuate, recognize, and identify self-same entities, prototypically but not exclusively persons. Much of the literature in linguistics and philosophy concerned with the semiotics of personhood has been narrowly focused on nominal form-classes: proper names (Kripke Reference Kripke1980; Butler Reference Butler1997), personal pronouns (Benveniste [1956] Reference Benveniste1971; Singer Reference Singer, Lee and Urban1989), and common nouns and status terms (Žižek Reference Žižek1989; Butler Reference Butler1993). It has also been, as anthropologists have pointed out, narrow in its conception of what counts as a person, focusing on modern Western conceptions of subjectivity.
Within linguistic anthropology and sociolinguistics, the notion of enregisterment is a key semiotic tool to enrich both these limited views: on the one hand, expanding beyond the noun, and beyond the speech event, to the implicit and distributed metapragmatic functions that are part and parcel of how social and personal identities are invoked, enacted, lived, and interdiscursively constructed and disseminated across social time and space; on the other hand, to stressing that registers are cultural models that may vary widely in the values (viz. concepts of personhood) that are invokable through the texts-in-context that they mediate. While linguistic anthropology and sociolinguistics have drawn our attention to processes of typification that are critical to register phenomena, the articles in this special issue draw our attention to complementary processes of rigidification, exploring how signs can be drawn into indexical relations that invoke and presence not only social types but also social tokens, not only figures of personhood but also individuated, identifiable persons. Person-indexing registers are thus not just another phenomenon that has been unrecognized. They are essential components of a dialectic process without which we cannot understand how persona-indexing registers themselves are constituted, and vice versa. Enregisterment and the social processes and phenomena it makes possible (including subjectivity, we would argue) require an attention to this dialectic.
We should not be surprised that the pre-eminent and most important of such social types and tokens is the person. Because people reflexively and agentively articulate their personhood, but are also made into (named) individuables often at, and even before, birth, the study of enregisterment must attend to the complementarity of these two perspectives. Persons are speaking and signaling subjects who are performatively enacted, by themselves and others; such enactments are always relational processes, precisely because the recognizability and identifiability of such persons (to others and to themselves)—as locatable nodes within particular social networks—is a condition on their existence. It is the interconnection and slippage between these two perspectives—persons as tokenized and typified, identifiable and locatable both socio-centrically and in their particularity—that is our interest in this special issue. If this slippage has long been a source of fascination and concern, today it confronts us as one of the most pressing issues of the day. If persons have always remixed the signs of others in creating what is distinctive to them, and if the categories through which we render the world intelligible (e.g., social types of personhood) emerge from singular encounters (with singular beings) and vice versa, today we also live in a techno-mediatized moment where “our” most intimate and inalienable signs circulate largely out of our control, subject to more and more individuating forms of recognition and surveillance, even as they are reaggregated into and used to train the enregistering machines of the present and future.
Acknowledgements
This special issue and the articles in it originated from a panel entitled “Person-Indexing Registers” (with papers given by Luke Fleming, Courtney Handman, Constantine Nakassis, Julien Porquet, Stanton Wortham, and discussant remarks provided by Angela Reyes), organized by the editors of this issue at the annual American Anthropological Association (AAA), Tampa, Florida, November 20, 2024. Stanton Wortham and Charles Zuckerman then organized a two-day workshop, “Person-Centered Enregisterment,” at Boston College on September 12–13, 2025. There, participants discussed expanded versions of the AAA papers by Fleming, Handman, Nakassis, and Porquet, as well as a new paper by Reyes. We thank the participants in both events, as well as Ilana Gershon, Nicholas Harkness, and Charles Zuckerman for their comments and feedback at the Boston College workshop. We also extend our thanks to Asif Agha for his interest and encouragement in publishing this collection of articles in Signs and Society.
Open access
