Introduction
Norman Cline (a pseudonym) was a 75-year-old white man hospitalized in an American hospital after a severe asthma attack and put on a ventilator, although he was later given a tracheostomy in hopes he could speak again. He did not, and neither did he leave the hospital; after some weeks, he died. An account of this process, included in medical anthropologist Sharon Kaufman’s ground-breaking work on dying in modern American hospitals, And a Time to Die: How American Hospitals Shape the End of Life (2005), provides an example of one paradox of verbal abilities that face people at the end of their lives. This paradox (detailed below) calls out for theorization.
At the time, Kaufman (1949–2022) was a medical anthropologist at the University of California, San Francisco. The book was based on 2 years of fieldwork at three community hospitals in California, which involved interviewing family members, doctors, nurses, other medical staff, chaplains, and hospital bureaucrats about the path toward death of 100 critically ill people. In so doing, Kaufman revealed a number of striking features about modern death. One insight was how hospital bureaucracies, their legal and financial regimes, and the parameters of medical professions kept individuals from the deaths they wanted to have in terms of personal autonomy, dignity, relational presence, and physiological regulation, turning what should have been scenes of typical dying into instances of “problem” dying. Along the way, Kaufman depicted numerous ways in which personhood is measured through and managed by language, not only that from patients but from medical professionals and family members as well.
The paradox facing Norman Cline had this shape: Although his communicative competence and autonomy were used by medical staff and family members to indicate his remaining mental and psychological competence, he was unable to convince any party to take his desires to die seriously because they had already put his competence in question. Modern medicine seems not only to have added one more function, the diagnostic, to Roman Jakobson’s famous six functions of language (Jakobson Reference Jakobson and Sebeok1960), but it has also reduced language to that one function.Footnote 1 As far as linguistic markers of competence go, syntactic complexity and simple phonation were blended into a single folk model of agentive competence, whose loose, impressionistic manner might not have produced the same result in another patient’s situation, depending the medical circumstances, demographic factors, the languages in use, or the type of professional. Cline’s son and daughter wanted him to speak and asked doctors if he “would ever be able to speak again in sentences.” Despite the tracheostomy, Cline was unable to be fully weaned from the ventilator, so he subsequently faced two barriers to speech: the ventilator tube and tracheostomy. (Because a tracheostomy tube is inserted below the larynx, air from the lungs exits the body before passing through the vocal cords.) It is possible to learn how to speak with a tracheostomy, possibly with a tracheoesophogeal voice prosthetic, but this was too arduous for Cline, who was in pain and undergoing other invasive procedures. When he asked nurses to be allowed to die, he mouthed the words: “Will you help me die, please?” (Kaufman Reference Kaufman and Morgan2005, 252). He also directed these questions to his children, who were emotionally unwilling to accept his death, so they requested a psychiatric evaluation.
At this point, Cline’s biological survival was not (yet) at stake; it was his status as an agent with perceived capacity who was able to make decisions about the treatment he received and also as a site for emotional attachments by his offspring. In other words, his survival as a social actor was at stake. As Kaufman relates, the psychiatrist’s evaluation added another layer to the negotiation of his competence as a person, because they acknowledged that Cline was, in fact, fully capable of making decisions about his own health care. As Kaufman explained, such psychiatric evaluations “focus primarily on whether patient have the capacity to make well-considered choices regaring the refusal of life-sustaining therapies. Suffering…gets transformed…in psychiatric evaluations specifically into evaluation of mental or psychological competence and rationality” (Kaufman Reference Kaufman and Morgan2005, 253). Suffering and competence were read through the patient’s linguistic behavior, which seemed to be subjective and impressionistically evaluated. This is often the case in such settings, where the systematicity of language structures and social aspects of their deployment are not appreciated.
Yet, as Kaufman related, no one took Norman Cline’s pleadings seriously. After 5 weeks in the hospital, he told the social worker that he wanted to die. “This is not living. Enough is enough. Please help me” (Kaufman Reference Kaufman and Morgan2005, 253). In an extended quote, Cline’s doctor described the situation in a way that again showed a disconnect between Cline’s persisting linguistic agency and his autonomy, as perceived by the medical caregivers and his family. All of them regarded Cline as less than a competent person. As the doctor later recalled, “So when he first started expressing some of those feelings, I wasn’t quite sure what we were dealing with, you know, whether he was or wasn’t saying that [he wanted to die]. But after a while it became apparent that he was, and he voiced it enough times so that we were able to satisfy ourselves that it was in fact what he was saying” (Kaufman Reference Kaufman and Morgan2005, 253). These professionals in this setting used other diagnostic information to evaluate Cline’s mental status, but it is the role of his language abilities and their mixed effects that I want to highlight here.
What might be said about this case and others like it? At the end of life, no matter how old they are, people become less communicatively competent for a range of reasons. Their impairments can be adapted to, but they are not remediable; they must have their organic (in)competence in general and specific productions within it reinterpreted by others, something that is an element of all communicative life. In these situations, however, the power relationships can be more stark and the dying person may lack the ability to repair or manage any (mis)interpretations. Such impairments are, in the words of Sterne (Reference Sterne2021), “normal,” meaning that they are expected and normalized. “To find normal impairments,” Sterne wrote, “consider how a context sorts bodies for capacities, look for impairment where it is not normally noticed, and then challenge any explanation that begins with an ideology of ability” (Sterne, Reference Sterne2021, 155). End of life is the context for changes to a range of linguistic phenomena, from changes in production to processing and from patterns of pragmatics and interpretation, which can be examined psycholinguistically (in terms of brain function) and sociolinguistically (through sensitivity toward variation between and within language communities). Only some of these phenomena have been described in academic and popular accounts. However, the point is not that these phenomena are necessarily under-described in themselves (although that is often the case), but that they have not heretofore been conceptualized together. Doing so reveals the fragility of decision-making expertise, assumptions about abledness, and numerous interpretive dilemmas whose study can provide ecologically valid linguistic detail in order to inform clinicians’ evaluative practices and make them encounter their own empirical commitments.
My candidate for such a lens is a concept called “language de-socialization.” This term refers to any process through which the declining social and interactive capacities of a person at the end of life, as well as the loss of personhood as it is defined in a particular cultural setting, are dealt with, guided, interpreted, responded, and adapted to through linguistic processes by varied social actors. This concept is the end-of-the-lifespan corollary to “language socialization,” which has been an enormously influential theory of language development in social and cultural context developed over the last 35 years by linguistic anthropologists, applied linguists, sociolinguists, and others (for overviews, see Garrett and Baquedano-López Reference Garrett and Baquedano-López2002; Berman and Smith Reference Berman and Smith2021, among many others). Definitions of language socialization are varied, but the one I use here is “how young children and others become communicatively and culturally competent within their homes, educational institutions, and other discourse communities, both local and transnational, and how language (in its many varieties and modes) mediates that process” (Duff Reference Duff, Patricia and May2017, ix). My argument is not a critique of this theory except to point to one limitation: it does not adequately explain what happens with language at the end of life in social and cultural context.
Perhaps a new theoretical lens is not necessary; perhaps language socialization will suffice, in fact, for Cline’s story. After all, he had to be socialized to his new status as a patient and the new hospital environment, although that could only be the case if it were his first hospitalization. He would have been instructed about aspects of his care and introduced to the culture of hospitals in general and this hospital by doctors and nurses using language in what would be a culturally patterned event. He is “becoming” a patient, a participant in the health care system. Furthermore, his children had to learn how to interact with their speechless father. Indeed, his case has elements of growth, learning, a “will to plasticity” (Erard Reference Erard2012), and positive regard for the future. His children, and to some extent his doctors, behaved as if this hospital stay was merely temporary and Cline would be able to resume his normal life. Yet, the undeniable reality offered a different set of conclusions, which they had to learn to accept, an acceptance that hinged partly on Cline’s linguistic abilities. Undoubtedly, they were being socialized into being the loved ones of a dying man. Perhaps most importantly, the fact that Cline was dying may have been visible only in retrospect, as dying does not necessarily announce itself; the beginning of “end of life” can be so widely defined in practice as to be ambiguous. (Its terminus also varies, depending on cultural beliefs.) Given all these factors, isn’t the case of Norman Cline made sensible in a language socialization framework?
But Cline’s story—and, I would argue, the matter of dying people more generally—demonstrates that dying people are specific types of linguistic actors whose behaviors and identities are variably consistent with models for linguistic actors over the rest of the lifespan. As Berman and Smith (Reference Berman and Smith2021) point out in a critique of language socialization, the theory assumes that the “‘novice’ is its central analytical object” (Berman and Smith Reference Berman and Smith2021, 124). They define “novice” as an individual “lacking a quality of character or form of skillfulness that is…valuable in some cultural context and that.the agents are expected to eventually mobilize” (Berman and Smith Reference Berman and Smith2021, 124). Norman Cline, and indeed people at the end of their lives, cannot be novices in these terms. They are made into participants in order for them to longer participate in any aspect of life: social, cultural, and biological. I also note that the abled novice is assumed as well. Removing these assumptions about novicehood and abledness is at the center of language de-socialization. These phenomena may not be legible to participants at the time. They are also not specifically related to language and aging, since not all people who die are old. Thus, there needs to be an understanding of language change in the life course that incorporates both biological and social understandings of language and what impairs its use in interaction (such as disease processes, aging, medications, other therapies, etc.), along with better comprehension of the ways that human interaction is constructed in and through ideologies of language, culture, health, and ability more broadly. Such ideologies might link phonation with biological life and even consciousness, for example, or allow a more “competent” communicator to determine the parameters of an interaction and not accommodate to impaired communication practices.
Myriad forces are at play here, and the changing linguistic and cognitive functioning of an ill and dying person is only the beginning. Layered atop them and transforming that functioning are also ideas, beliefs, and practices in any given community about the nature of the dying person and prevailing ideas about mortality, death, and dying; how family members, medical staff, other caregivers, religious guides, and others perform their social roles and meet social expectations; how people will have been socialized to be linguistically agentful aging and dying people through cultural attitudes, religious beliefs, popular media, and perhaps even personal experiences at deathbeds and how they navigate the inability to enact those cultural models; and the affordances of the physical and social environments in which people die (environments which are themselves embedded in, and shaped by, ideologies of life, death, and medical care). The approach begins with the language that dying people produce but is also attentive to how those productions are interpreted and why as well as the purposes to which participants put this language. Understanding this territory would be the project of language de-socialization.
It is striking that one of the medical interventions designed to prolong Norman Cline’s life took his ability to phonate from him, while his inability to phonate and re-learn speaking were clearly linked to judgments about his capacity to make decisions about his own quality of life. What is at stake for dying people like Norman Cline can be demonstrated by slightly altering a key quote about language socialization. As the pioneers of language socialization studies, Bambi Schieffelin and Elinor Ochs, wrote in 1986, “The process of acquiring language is deeply affected by the process of becoming a competent member of society” (Schieffelin and Ochs Reference Schieffelin and Ochs1986, 168). In Cline’s case can be seen how “the process of losing language at the end of life is deeply affected by the process of becoming a less competent member of society” (my modifications in italics). His own language, one vector of his agency, becomes the vehicle by which he participates in his non-participant status. This depends on which aspect of “language” is salient for whom and how such a definition is deployed in practice. Furthermore, in this context, the notion of “competence” needs to be viewed through new perspectives on the body in linguistics that come from “crip” linguistics and disability studies (discussed below). This would highlight the ways in which “competence” is based on problematic, impressionistic assumptions about abled bodyminds. It would also expose how language socialization itself and the quantitatively normed developmental trajectories of infants and children that it presupposes are also built partly atop ableist assumptions.
In the rest of this article, I describe some salient similarities and differences between language de-socialization and language socialization. I point to the potential for language de-socialization to serve as a useful theoretical frame for considerations of a range of social phenomena, three of which are briefly described in terms of their relevance for a de-socialization approach: moaning, the diagnostic status of linguistic behavior, and elderspeak. Seeing these side by side makes visible how participants in a cultural sphere become non-participants through what someone else has judged to be the remnants of their linguistic agency. I will draw again from Kaufman’s work with another case from the North American, Anglophone hospital setting. Taken all together, I mean for them to suggest the rich promise of this theoretical frame for understanding language at the end of life and provide a direction for future research on the full arc of linguistic agency, both individual and collective, over the lifespan.
Similarities between language socialization and de-socialization
Most prominent among the parallels between language socialization and de-socialization is the way that intertwined nature and culture are implicated in processes of change. At the end of life, individuals bring a lifelong socialization process to their organic decline, which is shaped by many factors, including contexts that are designed to preserve socialization processes and deal with organic declines. Here, it may be worthwhile to consider which linguistic and communicative abilities are preserved and how, within a cultural framework that authorizes and legitimates which abilities to preserve in the first place. As Penelope Brown writes in an essay in the Handbook of Language Socialization, various cultures enculturate parents and other caregivers to give joint attention to infants in different ways (Brown Reference Brown, Duranti, Ochs and Schieffelin2014). This is also the case at the end of life; indeed, one wonders the degree to which people bring models of joint attention with infants to their encounters with the dying. It may be tempting to dismiss this and other end of life processes as too variegated to study. However, this assumption of heterogeneity may not withstand scientific scrutiny, in the same way that an assumption of homogeneity of developmental processes has not withstood a comparative cultural approach.
For tracing the intertwinings of nature and culture, the tools and perspectives of linguistic anthropology are critical for understanding the stages of language development (Schieffelin and Ochs Reference Schieffelin and Ochs1986), as is work in the ethnography of speaking/communication that situates conceptions of the speaker and meaningful acts of language production in cultural context (Bauman and Sherzer Reference Bauman and Sherzer1975) along with an emphasis on language as a form of joint social action (Goodwin Reference Goodwin2018). Goodwin’s studies of a man (his father) with a severe non-fluent aphasia as a result of a stroke who could nevertheless participate in conversations as a competent speaker, even with a lexicon of three words, stand in contrast to the linguistic paradox of Norman Cline (Goodwin et al. Reference Goodwin, Harness Goodwin, Olsher, Ford, Fox and Thompson2002; Goodwin Reference Goodwin2004). Language socialization approaches have the ambition to encompass the entire lifespan. As language socialization pioneers Bambi Schieffelin and Elinor Ochs wrote, “language socialization begins at the developmental point at which members of a community recognize that a person enters into existence and continues throughout the life course until a person is viewed as no longer a living social being” (Schieffelin and Ochs Reference Schieffelin, Ochs, Duranti, Ochs and Schieffelin2014, 3). This sentence appears in an introductory essay to the landmark Handbook of Language Socialization, which provides a comprehensive view of the many forms, places, and processes of language socialization, although none are consistent with end-of-life realities.
However, as medical anthropology demonstrates, being “a living social being” is not a binary affair. In fact, personhood is a continuum on which an individual may occupy many points in a nonlinear fashion until they are deceased (and even then they may still be regarded as socially alive though biologically dead) (Kaufman and Morgan Reference Kaufman and Morgan2005). This is clearly the case with Norman Cline, who was enough of a person to be a site of emotional attachment but not enough of a person for his wishes to have sufficient force. Although Ochs and Schieffelin appear to be open in their essay to an approach that covers the lifespan, the rest of the paragraph in which the quoted sentence appears uses examples that are exclusively related to fetuses and babies. Other essays in the Handbook similarly focus on children. Elsewhere in their introductory essay, Ochs and Schieffelin refer to “persons across the life cycle” as research subjects of interest, but the other chapters focus almost exclusively on child language. This is an empirical and theoretical gap which, I argue, the concept of language de-socialization is exceptionally well-suited to fill.
Finally, like language socialization, language de-socialization is concerned with a range of linguistic phenomena. People note features of behaviors in early language that are also present at the end of life. For instance, Akira Takada notes how language socialization begins before speech, involving “gestures such as facial expression, gaze direction, back channel response, and pointing are effectively used in caregiver child interactions long before children start speaking” (Takada Reference Takada, Duranti, Ochs and Schieffelin2014, 59). This also happens at the end of life, widely reported anecdotally but not systematically studied. How could it possibly pattern, given the diversity of ways that people die and given communicative diversity among dying people? Is this an evolutionary-conditioned mirroring, or do speakers assume that dying people are a sort of linguistic actor similar to young children? To resolve this, a language de-socialization approach would note that dying adults, more so than young children, do interact with many people, not just medical professionals. As Brownlee and Bruening (Reference Brownlee and Bruening2012) documented in a retrospective survey of surviving partners of deceased amyotrophic lateral sclerosis (ALS) patients, the choice of communication mode depended partly on who the communication partner is.
As in language socialization, language de-socialization allows for the consideration of the possible phenomena, including spoken, written, and sign language; symbol use; moans, groans, and other “liminal signs” (Keevallik & Ogden Reference Keevallik and Ogden2020; Dingemanse Reference Dingemanse and Enfield2017, Reference Dingemanse2018, Reference Dingemanse2020); formulaic language (Wallesch and Blanken Reference Wallesch and Blanken2000; Wray Reference Wray2012; Wray & Perkins Reference Wray and Perkins2000; Bridges Reference Bridges and Sidtis2013; Sidtis Reference Sidtis2010; Sidtis & Postman Reference Sidtis and Anne Postman2006); religious language (Coppens Reference Coppens and Ghaly2023; Rambachan Reference Rambachan, Coward and Stajduhar2012); delirious language, both structurally and its interpretation (Duppils & Wikblad, Reference Duppils and Wikblad2007; Namba et al. Reference Namba, Tatsuya, Chizuru, Emi, Sayuri and Kei2007; Wright Reference Wright, Brajtman, Cragg and Macdonald2015) in its social context; emotional expressions; interjections; names; curses; silence; and patternings of linguistic accommodation by age, gender, race, culture of origin, and other demographic variables. These phenomena occur across physical settings, institutional environments, and cultural contexts, and they involve a range of language users: not only the dying person, but also religious figures, medical professionals (doctors, nurses, therapists), social workers, spiritual workers (such as chaplains), informal carers, helpers (such as doulas), and of course family members.
Finally, like language socialization, language de-socialization as a scholarly enterprise can describe cultural beliefs, including religious ones, without confirming or critiquing them. There is scholarly literature on aspects of dying, such as “deathbed communications” and “end of life dreams and visions,” that proceeds from the view that these phenomena are present cross-culturally and across history (Hession et al. Reference Hession, Luckett, Chang, Currow and Barbato2023). Similar are popular discussions in the United States of journey metaphors by dying people, often taken as evidence of a destination in the afterlife. Such a claim to universality is the one that anthropological sensitivities of language de-socialization are prepared to critique. In the same way that language socialization perspectives can show that Western-style first words, such as “mama,” are not shared in other cultures (Ochs Reference Ochs1982; Schieffelin Reference Schieffelin1990), language de-socialization makes it possible to understand both the linguistics of the interaction window and the variability of what happens in that window at the end of life.
Differences between language de-socialization and socialization
Nevertheless, language at the end of life is a more complicated matter than language at other times of life. A main reason is because the social, cognitive, biological, and cultural variables are more numerous and interact in complex ways. Approaches to language socialization are founded in the biological certainty that language development may proceed in a variegated way yet produces a uniform outcome for nearly all the individuals who undergo it: they become competent users of their language, whose use varies socially within their communities. These organic processes are assumed to provide a uniform substrate of experience in early life, one that is mediated by culture but which proceeds with more or less the same milestones across social groups. These milestones include pointing, first adult-like lexical productions, and first two-word syntactic structures. For instance, hearing young children around the world are able to produce their first phonetically distinct, semantically consistent utterances at about 12 months of age, whether or not adult caregivers engage in joint attention with them, whether or not a first word has cultural salience, and no matter how the particular language constructs words.
However, the biological inputs to language decline and communicative impairment at the end of life are more diverse while less well understood. This relative paucity of knowledge has numerous causes. Given that the developmental focus in the language sciences targets processes by the young, it is not surprising that the language socialization literature does not address processes across the whole lifespan. Also, medical research has a tendency to turn people into patients, categorizing them according to their medical condition rather than some other attributes, such as communicative functioning. (The only linguistic categorization, which is mainly an administrative one, would be people who have exercised their legal right to interpreting services. In that sense, one could not, at a glance, find out how many verbally non-responsive patients are in a hospital at any given time.) This, in turn, is a matter of the requirement that patient-related research return direct benefits in terms of care to those patients, whereas research on the linguistic and cognitive development of children can occur regardless of any practical benefit. Additionally, ethics approvals are more difficult to get for vulnerable populations like the dying, although it has been done (Blundon et al. Reference Blundon, Gallagher and Ward2020, Reference Blundon, Gallagher and Ward2022).
A few scholars have approached medical interactions and communicatively impaired populations from a language socialization perspective. In an extensive overview of dementia and communication research, Saunders (Reference Saunders, Patricia and May2005) argues for a strong rationale for applying language socialization to people with dementia, as they navigate identity changes and learn new communication strategies and as others adjust to their impairments. Overall, the focus is an optimistic one and does not treat what happens when strategies begin to fail. Stivers et al. (Reference Stivers, Nicholas, Enfield, Brown, Christina, Makoto, Trine, Gertie, Federico, Jan Peter, Kyung-Eun and Stephen2009) analyzed patterns in doctor-pediatric patient interactions in the healthcare context, noting particularly how pediatric visits shape children to be future adult patients.
On this point, various institutions serve as contexts for socialization as well as de-socialization, some of which overlap, such as the family and the medical setting. However, the underlying logics of critical interest in those institutions differ for language socialization and de-socialization. In regard to the former, the society and the state have significant interest in the resource-intensive processes involved in language development, which lead to citizenship activities and economic productivity. Language socialization provides a critical framework for understanding these institutions and their interests. However, there is little to no corresponding economic benefit or political interest in understanding language at the end of life outside the economics of health systems. Because the subject of language de-socialization is the parameters of life, which inevitably involves decrescence and decline, it draws heavily from work in medical anthropology (Kaufman and Morgan Reference Kaufman and Morgan2005) where “the delineation of cultural forms and structural sources of subject-making are central” and “stresses how scientific practice, together with discursive power arrangements, shapes understandings of the parameters of life, death, and the person and creates particular desires and needs” (Kaufman and Morgan Reference Kaufman and Morgan2005, 319).
Furthermore, because it deals with these “parameters of life, death, and the person,” language de-socialization must be concerned with the physical body in ways that language socialization does not. To address this, it draws from the nascent subfield of “crip linguistics,” first articulated by two linguists, the late Henner and Robinson (Reference Henner and Robinson2023) and was originally conceived as a theoretical intervention to critique linguistics and related fields through a disability lens. Crip linguistics can be foundational for language de-socialization because it highlights the role of ableist assumptions about linguistic competence and incorporates the “bodymind” into conceptions of that competence. As linguists have pointed out, the body’s role in language has been underappreciated (Buchholz and Hall Reference Buchholz, Hall and Coupland2016). One result, among other things, has been to privilege abledness as a norm and “disability” as forms of atypical language (Henner and Robinson Reference Henner and Robinson2023). Just as groups of people with lifelong disabilities have been marginalized in linguistic studies, so too are the abled bodies that eventually stop working. Central to language de-socialization is the acknowledgment that in the periods involving marked linguistic and cognitive decline—no matter how brief or extended—language is “cripped” for everyone, where “cripping” “highlights the linguistic adaptions used by disabled people, including their relations and world-making, and illuminates structures of ableism that govern how we perceive language” (Henner and Robinson Reference Henner and Robinson2023, 11). I point back to Norman Cline’s story and forward to my discussions of moaning, language as diagnostic marker of dying, and elderspeak.
Crip linguistics usefully highlights the impact of ableist expectations about discourse as a temporal phenomenon, offering the co-construction of meaning as a form of care work, in which “the time taken in being patient, in supporting and providing semiotic resources, in seeking, expanding, and claiming our own semiotic resources, in calibrating to each other in seeking mutual understanding” (Henner and Robinson Reference Henner and Robinson2023, 27). One is struck, reading stories of deathbed interactions as well as research by speech-language pathologists working in palliative care (Chahda et al. Reference Chahda, Mathisen and Carey2017; Pollens Reference Pollens2012, Reference Pollens2020), by the intense forms of linguistic care work that occur, for instance, by waiting for a person to respond or modifying one’s turn-taking behavior to allow the other participant to use alternative and augmentative modes of communication. Relevant to end-of-life experiences is the way that crip linguistics foregrounds what Mingus (Reference Mingus2011) called “access intimacy” or the connection between people when needs for various sorts of access can be anticipated and delivered. “Access” here means the ability to use services (medical, educational, and legal). Language-related examples might include allowing a written word or gesture to serve as legally binding affirmation when a spoken one is not possible; the proactive availability of medical interpreters in medical contexts; and licensing language users from marginalized groups to use non-standardized language forms without sanction or comment. Medical professionals seem to be variably competent in access intimacy for their patients, although speech-language pathologists play a growing role in palliative care settings, where professionalizing communication access intimacy is a goal (Chahda et al. Reference Chahda, Mathisen and Carey2017; Pollens Reference Pollens2012, Reference Pollens2020).
Crip linguistics highlights how models of abledness in medical cultures construct the cognitive and linguistic competence of patients. For instance, sociologist John MacKay writes of his experiences after a stroke with a doctor who “built on my communicative competence to make me appear as an incompetent person” (Mackay Reference Mackay2003, 815). He goes on: “The doctor was oblivious to my ability to express my thoughts non-verbally, as well as through the signal specific system I developed, and that allowed interaction” (Mackay Reference Mackay2003, 815). Another illustration comes from nursing researcher David Wright, who described how delirium is explained to family members in hospice in order to reduce their distress. “Rather than communicate to families that these conscious and cognitive alterations were disturbances, hospice caregivers normalized them, weaving them into a coherent system of meaning that preserved the integrity of the good death idea” (Wright et al. Reference Wright, Brajtman, Cragg and Macdonald2015, 4). Such contestations over personhood on the grounds of assessed linguistic competence lies deeply within the territory of language de-socialization.
Crip linguistics aids with the boundary problem mentioned earlier, in which there may not be a clear trigger by which participants sense that they are in fact operating at the end of life. No one may know for sure that an individual is dying; they may be denying it. (This points to another difference with language socialization, in which participants can be aware at the time that language socialization processes are operative.) Crip linguistics helps because it insists that linguistic competence is always present, endangered (if not erased) by ableist assumptions about “normal” language, and socially constructed in terms of those unexamined assumptions. Where language de-socialization comes into play is where competence is becoming impaired. Thus, whether or not they are dying, a person might be simultaneously marginalized by ableist definitions and be linguistically less competent organically, as in the case of Norman Cline. Certainly what they are not is a novice, but an expert becoming less able to deploy their expertise. However, where crip linguistics and language de-socialization do not necessarily articulate with each other regards the issues of time and identity. A dying person faces a cascade of changes over a relatively brief period of time, too short to build a stable identity around their dyingness. Moreover, as previously able-bodied people, they may not believe that their identities have shifted. On the other hand, crip linguistics is concerned with the implications for attributions of personhood by linguistic difference over longer periods and is organized around benefiting a collective, say of disabled people or deaf communities in their struggles for access and justice.
To be clear, I am not saying that a dying person who cannot verbalize but can still respond with hand squeezes belongs to the same category as a lifelong signer and member of a deaf community, except to the degree that the communicative repertoires of both will be regarded through ableist assumptions and, on the basis of this, their social competence and status as a person regarded and managed. In fact, the critique of benchmarks of language development by crip linguistics because they do not account for ableist assumptions about other aspects of development (for instance, aligning motor control over speech with motor control necessary for walking) may be useful for explaining why language at the end of life has been disregarded by linguistics: in the absence of developmental benchmarks, there can be no “normal” against which a trajectory of dysfunction can be mapped, only idealizations of linguistic agency within cultural ideas of the “good death.”
Applying the language de-socialization frame
Language de-socialization is more than a flipped version of language socialization, because there are sufficient differences in underlying assumptions, focus, theoretical background, and the phenomena under question. Here I examine three disparate phenomena and address them through a language de-socialization lens. I also return to Kaufman’s book for another example that can be elucidated through language de-socialization. Although aspects of each topic may touch on socialization, I argue that the dynamics of the phenomenon are more fully mapped through de-socialization, that is, through an approach focused on the dynamics involving the negotiation of “less” personhood at the end of life in an impaired body, rather than one that assumes an operative “novice” in an abled body.
Moaning
Moaning is a vocal, non-linguistic behavior often encountered in clinical care settings and produced by a range of patient populations: the dying (Frade et al. Reference Frade, Carvalho Santos and Galriça Neto2021), the elderly, those with dementia and other neurological diseases (Samuelsson and Hyden Reference Samuelsson and Hyden2011), those with transient forms of post-operative delirium, and anyone expressing pain or discomfort. The linguistic aspects of the phenomenon do not lie in the semantic content of the moan. Rather, it lies sometimes in perceptions of what the moaner intends, to the degree that these intentions are accessible, and how it may pattern with other aspects of the moaner’s spoken language, including other sounds at the margin of language, such as clicks, sniffs, and sighs and expressions like ugh, argh!, pfft, and phew (Keevallik and Ogden Reference Keevallik and Ogden2020). A listener may interpret the moan according to previous interactions with the person, previous encounters with moans, and cultural norm-invoking ideas of how dying people should behave. For instance, vocalizations such as “repetitive vocalization, verbal or nonverbal utterances, presented inappropriate language, repeated and insistent demands, repeated calling out, shouting, complaining, or moaning” are called “disruptive” if they do not “pertain to [the person’s] circumstances or environment” (Palese et al. Reference Palese, Menegazzo, Baulino, Pistrino and Papparotto2009, 191). There appear to be no aspects of learning or accommodation around the issue of disruptive vocalizers; in fact, nurses report distancing themselves from disruptive vocalizers with severe dementia (Palese et al. Reference Palese, Menegazzo, Baulino, Pistrino and Papparotto2009). Also linguistically relevant are the cultural backgrounds of moaners, as expression of pain differs across cultures (Bouchard Reference Bouchard, Incayawar and Knox2014). Some moans appear to be intentional attempts to gain the attention of others; other moans are involuntary vocalizations, which may or may not be expressions of pain. Thus, the status of a moan as initiating an interaction or amounting to a conversational turn is unclear. Health professionals often must instruct relatives about the meaning of moans, assuring them that the moan is not necessarily a signal of suffering (Lichter and Hunt Reference Lichter and Hunt1990). However, relatives are described as often disturbed by the moaning and urge medical staff to sedate the patient, perhaps because they are socialized to respond and frustrated because no intervention is possible.
Why does moaning belong to language de-socialization? One reason is because the interpretation of the moan depends on the perceived status of the patient as a person (their agency, their autonomy, and their social position), which may also depend on the perceiver’s social relationship to the utterer. Depending on those perceptions, a vocalization with the same phonetic shape could be interpreted either as an intentional, cortically organized signal with communicative meaning or as an involuntary, limbically organized and organic behavior with no communicative intent. It could be a socially valid and permitted production, or it could be “disruptive.” From a language de-socialization perspective, not only would the meaning of a moan be considered but also to whom it matters that a moan has meaning and why. Unlike other vocal performances, the moan does not socialize either producers or receivers, nor is one socialized to produce it; this is what puts this in the realm of language de-socialization. Another reason is that moaning may be a function of “separate, phylogenetically older neural networks responsible for the production of non-linguistic vocalizations” (Anikin et al. Reference Anikin, Baath and Persson2018, 55), but there needs to be a critical apparatus to handle the potential claim that moaning at the end of life marks a return to a phylogenetically older status. In this way, language de-socialization potentially offers evidence relevant for theories of language evolution.
Prognostic conversations/conversation as prognostic
The ability to establish how soon a dying person might die—to “diagnose dying”—is important for clinical and health systems reasons, because it allows administrators to optimize the use of staff and facilities. For this reason, the diagnostic and prognostic value of a range of candidate factors has been tested. Most of these are physiological in nature (such as drooping of the nasal folds), while a few are cognitive and social, including linguistic abilities, vocalizations, and the appetite for social interaction. For example, “withdrawal from conversation with staff and a new tendency to decline social interaction with other residents or family members was… identified” (Kennedy et al. Reference Kennedy, Brooks-Young, Gray, Larkin, Connolly, Wilde-Larsson, Larsson, Smith and Chater2014, 4) as one marker of imminent dying among nursing home patients. Among patients with advanced cancer, a “highly diagnostic” bedside sign of impeding death (within 3 days) was “decreased response to verbal stimuli,” as was “grunting of vocal cords” (also known as the “death rattle”) along with six other signs (Hui et al. Reference Hui, Dos Santos, Chisholm, Bansal, Souza Crovador and Bruera2015). “Highly diagnostic” meant that decreased response to verbal stimuli had a prevalence of 69% within the last 3 days of life. In a small (n = 50) sample of non-cancer patients, a decreased response of verbal stimuli had a slightly higher prevalence of 76% in the last 3 days of life, the highest prevalence among 10 other markers that were observed as well (Hosoi et al. Reference Hosoi, Ozone, Hamano, Maruo and Maeno2021). Twenty-nine of 50 patients who showed this symptom died within 72 hours of its appearance; 18 died within 24 hours. In another study, an audit of charts of 185 patients in a long-term care facility found significant cognitive impairment in the last 48 hours of life; interestingly, however, the only evaluation of communication skills was that of physicians and allied health professionals, not of patients (Hall et al. Reference Hall, Schroder and Weaver2002).
The critical language de-socialization frame is relevant for examining how such studies conceive of “language” and “communication.” In the diagnosing dying literature, for example, language and communication are sometimes grouped as cognitive behaviors and not mentioned specifically. Elsewhere, language production is assumed in descriptions of delirium but has no explicit mention, particularly not types of utterances and their quality. Also, although verbal behavior (such as verbal stimuli by relatives or caregivers) is noted, types and frequencies of nonverbal behaviors are not. In some studies, language, interaction, or communication are not mentioned at all. Thus, one rich area for potential research is determining what sorts of linguistic features, behaviors, and structures are used by whom for what purposes in medical settings, along with what these markers are determinative of, in terms of the person-ness of the individual who is dying. As we will see below, several behaviors are used as ad hoc diagnostic markers which may not necessarily be admissible as formal consent for medical procedures or the withdrawal of care.
Elderspeak
Elderspeak is speech addressed to older adults that is marked by “exaggerated pitch and intonation, simplified grammar, limited vocabulary, and slow rate of delivery” (Kemper et al. Reference Kemper, Ferrell, Harden, Finter-Urczyk and Billington1998, 43). Older adults consider it inappropriate and patronizing, as the features of this linguistic register do not actually aid their comprehension. One of its effects is to “lead to negative self-perceptions in older adults and challenging behaviors in persons with dementia” (Shaw and Gordon Reference Shaw and Gordon2021, 13). This phenomenon is relevant for language de-socialization because it stems from implicit beliefs about the reduced agency and personhood of the elderly and also works to reinforce elderly participants’ negative self-perception.
Does its use count as language socialization? It depends on who is being described and who is the target for the socialization. In fact, caregivers are de-socializing their patients or charges, less so bringing them into competent personhood but ushering them to the periphery of competence. If language socialization is how “one is socialized into and through language forms and practices,” then it is.
Personhood and competence at the end of life
Another end-of-life profile in Kaufman’s And a Time to Die… tells of a 45-year-old acquired immunodeficiency syndrome (AIDS) patient, Walter Cole (also a pseudonym). Here, Kaufman related how a doctor uses multimodal communication to determine Cole’s wishes. “…The critical care physician arrives immediately and goes to Mr Cole’s bedside. ‘Squeeze my finger if you want to go on the ventilator.’ Mr Cole does not squeeze. The doctor repeats his request and again there is no response. ‘What about CPR?’ Mr Cole shakes his head no” (Kaufman Reference Kaufman and Morgan2005, 261).
Of note is the status of different modes of communication (a head shake, a silence). In contrast to Norman Cline’s experience, Cole’s inability to speak is not diagnostic; what matters here are answers to the question. Also of note is the confusing structure of the doctor’s queries. The first is not completely suitable for a yes/no response (because the meaning of the silence is vague), and neither is the second. It is obvious that the physician is not adequately equipped for what has been called “conversation as praxis,” which is defined as a situational adaptation to the usefulness of certain linguistic modalities and the shifting among them (Bateson Reference Bateson1975). His queries do not match Cole’s pragmatic capacity; his unimpaired communicative abilities are defining the terms of the interaction. (It is an example of what “access intimacy” does not look like.) Nevertheless, this interaction convinces the doctor that Cole has changed his mind. As a result, a “Do Not Resuscitate” order is put on his chart (Kaufman Reference Kaufman and Morgan2005, 261). To what degree is the doctor’s contextual knowledge about the patient’s wishes increasing the pragmatic force of the absence of the squeeze?
Later in Kaufman’s account, a different doctor reflects on how Cole’s decision allowed himself to maintain some quality of life, which he described via language behavior: “He was able to respond to people. He was able to nod, to talk a little bit, and he was able to mouth words” (Kaufman Reference Kaufman and Morgan2005, 261). In a language de-socialization-oriented medical ethnography, we would pay attention to the modes of communication that have import and to whom. What counts as an utterance? What counts as a meaningful utterance that is legible to institutional actors, and when does “conversation as praxis” cede to legal definitions? What is the indexicality of such multimodal utterances, and how is access to these indexical meanings distributed across interaction partners? How stable is the hierarchy of utterance types across medical situations? Cole’s hand squeeze was indicative, but its status as a speech act was probably situational. What seemed to have more weight in a later conversation was his ability to vocalize, when he said “It’s time to be comfortable” (Kaufman Reference Kaufman and Morgan2005, 264). (This meant that he wanted to stop treatment for an infection and begin palliative care.) In an ethnography, we would pay attention to who directs language to whom and how. For example, Kaufman depicts how a doctor gives an extended speech to Cole, who is weak and unresponsive, about his prognosis and care options. We would also pay attention to which utterances and other behaviors have solely diagnostic status, only communicative status, or both, and when and why their status shifts.
Conclusion
What does the concept of language de-socialization offer? Its contributions are manifold. Most importantly, it extends language socialization, helping it achieve its goal of understanding language across the lifespan. It tests the limits of assumptions around the nature of “competence” as well as “personhood” as simultaneously culturally constructed, socially performed, and biologically bounded. It dovetails with critiques of language socialization, such as the insight that novicehood is actually a sociopolitical category, not a natural kind (Berman and Smith Reference Berman and Smith2021). It lines up with the critique of the assumption that this novice is able-bodied, leaving disability, impairment, and injury all unaccounted for. It has other potential other impacts, as well. The concept not only illuminates the fact that socialization is not, in fact, a lifelong linear process, but can decline, sometimes precipitously; it also points to the implications of this decline and demands they be dealt with. In this way, it can connect linguistic research at the end of life to goals that directly benefit patients. Such research can produce material facts that inform ethical evaluations, and it can aid in understanding how personhood at the end of life is understood (and by whom) in medical contexts. As such, it can provide critical recommendations for care practices, such as the importance of medical interpretation and cultural translation services even after curative therapies have been halted and the critical role for speech-language pathologists in palliative care. Further research from a language de-socialization perspective would help show how verbal responsiveness may not necessarily serve as a reliably universal diagnostic of cognitive competence or how expressions of pain are as much culturally shaped as organically determined. All these constitute knowledge that could be fed back into the system of education for physicians, nurses, and allied health professionals, including speech pathologists.
Secondly, it provides a conceptual frame for understanding a range of phenomena that have not heretofore been considered together. In so doing, it joins with a disability-inflected linguistics, or “crip linguistics,” to connect the realities of bodies with broader conceptions of language as a phenomenon and various language sciences as disciplines. It opens the possibility of studying language, interaction, and communication in their sociocultural frame across the true span of the life, across all the experiences that affect the bodymind, rather than the relatively narrow and unacknowledged default of experience of youth, health, ability, and novicehood. One result may be that dying people would be viewed more positively and less fearfully.
Finally, it emphasizes that the matter of the beginning of language, for all of its complexity, is far better understood than its end. In the case of language socialization, the biological boundedness of language development and acquisition is understood to be a matter of a developmental course defined by evidence-based norms, inflected by individual differences that stem from genetic variety or disability. It is also bounded by the notion that most children end up in the same place: as competent users of their native languages. A similarly unified story is also assured in the case of language at the end of life, as everyone ends up deceased. However, the biological and linguistic elements along this path add unique complexities, including the effects of aging, disease, and therapies, all of which are in interaction with able-bodied counterparts who hold various types of power and control over them. It is critical to understand more systematically the articulations of power and language ideologies at this stage. We may even live long enough to benefit from such understandings.
Open access
