What Doesn’t Answer Back: Nonrecursive Counterparts in Erving Goffman’s Interactionism

Abstract

Erving Goffman was among sociology’s finest observers of bodily, material, spatial and institutional coordination. Yet his hundreds of concepts appear overwhelmingly to organise these observations through human-human interaction, or through selves, identities and classifications constituted in relation to it. This article treats that pattern as a hypothesis to be investigated and proposes an explanation. The disciplinary object that social theory inherited established a criterion of conceptual admissibility, under which phenomena counted as properly sociological insofar as they could be formulated through human-human interrecursivity. Goffman observed far more than this criterion admitted, and his concepts record the pressure it exerted, through a process the article calls interrecursive conceptual translation. Asylums provides the decisive demonstration, because its most revealing descriptions concern material and symbolic deposits that allow institutions to coordinate conduct without continual interrecursive negotiation, often by making particular encounters unnecessary or unavailable. Living Value Theory’s distinction between nonrecursive, selfrecursive and interrecursive counterparts makes these processes visible without translating them into human-human interaction.

I. Introduction: Nine Hundred Concepts and the Ordinary Words for Things

Few sociologists have watched everyday coordination as closely as Erving Goffman. Almost nothing was beneath his notice: where a person stands when a lift door opens, how the eyes are lowered as a stranger passes, what happens to a room when a performance collapses, who may carry a key and who may not, what the loss of a comb does to a sense of self, how a patient conceals food from an attendant. His books are crowded with bodies, garments, doorways, timetables, case files, wards, gestures and objects, and with the small repairs through which people keep an encounter going when something threatens to break it.

He also named what he saw with unusual profusion. Robin Williams reports that Susan Birrell, preparing a doctoral dissertation at the University of Massachusetts in the 1970s, made an index card for each concept she found in Goffman’s writings and finished with more than nine hundred cards (Williams 1988: 88). The figure needs care. It counts cards made by one reader for terms she judged to be concepts, and it therefore includes borrowed terms, variants of a single idea, near-synonyms and expressions that Goffman used once and never took up again. It therefore establishes less a census of nine hundred independently coined and stable concepts than a density of conceptual naming that few sociologists approach, a habit of introducing, defining, extending and abandoning concepts that Williams elsewhere called Goffman’s “conceptual constructivism” (Williams 2000 [1983]: 106; see also Lenz and Hettlage 2022: xi). The standard explanations are familiar and correct as far as they go: Goffman distrusted grand theory, preferred tools to systems, moved deliberately between metaphors of theatre, ritual, game, territory and frame, and believed that sociology was too young for a settled conceptual framework.

These explanations account for how Goffman worked. They leave open a simpler question. Why did he need so many specialised sociological concepts when many of the activities he observed already had perfectly serviceable ordinary-language descriptions? The inmates of Asylums clean, dress, sleep, eat, carry, store, open, close and move. Ordinary language names such activities with ease, and it is particularly rich in words for embodied and nonrecursive coordination: one wipes a table, folds a shirt, locks a door, hides a cigarette, waits for a bell. The difficulty is that these ordinary descriptions do not satisfy the disciplinary expectations of sociological explanation. Cleaning a surface describes a coordination between a body, a cloth, water and a stain. To become sociology, it has to be redescribed: as backstage work that prepares a front region, as the maintenance of appearances before an audience, as the fulfilment of an institutional role, or as a ward duty within a privilege system. Each redescription is illuminating, and each moves the explanatory weight from the surface and the stain toward the people who will see it. Some activities need no such move because they are already multimediated. Buying and selling involve interrecursivity between parties alongside the material transfer of goods and the symbolic coordination of prices, so their sociological description translates nothing away. The pressure falls on activities whose coordinative centre lies elsewhere.

This article argues that the pressure was disciplinary. It distinguishes three things that are easily run together: observational attention, conceptual construction and ontological explanation. Goffman’s observational attention ranged across the whole mesocosm, the domain in which living beings actually coordinate their lives. His conceptual construction was governed by an inherited criterion of admissibility that favoured human-human interrecursivity. And the ontological explanation that might have held his observations together was unavailable to him. The recognition problem, in other words, was itself produced by a restriction on concept formation, and Goffman’s proliferating vocabulary is one of its clearest records. Asylums (Goffman 1961a) is the decisive case, because its finest observations concern arrangements whose coordinative achievement consists partly in making interaction unnecessary. The article’s central question follows: what happens when a discipline that can observe the entire mesocosm recognises as theoretically central only those phenomena that can be translated into its restricted conception of the social?

II. Goffman Inside the Impossible Object

The Enlightenment division of the world into nature and culture was a split of worlds paired with incommensurable epistemologies, and it distributed the mesocosm across disciplines authorised to know it in incompatible ways (Ecks 2026a). Multisensorial embodiment went to biology, physiology and medicine; multiversal dwelling to natural history and physical geography; multimateriality largely to economics, engineering and the applied sciences; multisymbolization to philosophy, linguistics, theology and the human sciences. Social theory received a fragment of multispecies being-with reduced to human-human relations, and even those chiefly in their institutionalised, normative and symbolically articulated forms (Ecks 2026c).

The allocation concerned more than an empirical domain. Social theory also inherited assumptions about what constitutes a legitimate sociological explanation. An account counts as sociological when it shows a phenomenon to be produced, sustained or transformed by relations among humans: by interaction, identity, roles, norms or their symbolic organisation. Bodies, objects and environments may appear in such an account as conditions, means, settings or backdrops, and they enter its explanatory centre when they come to matter to someone else. Two forms of bias need to be distinguished here. Empirical interrecursive bias consists in disproportionately attending to human-human encounters. Conceptual interrecursive bias consists in constructing explanations primarily around human-human interrecursivity, even when the observed processes are mediated through bodies, objects, environmental conditions or selfrecursive coordination. The second can persist in the absence of the first, and Goffman is a particularly revealing case because his observational attention ranges far beyond the encounters that the empirical bias would select, while his concepts repeatedly exhibit the conceptual bias.

The mechanism through which this happens can be called interrecursive conceptual translation: the conversion of observations concerning any of the five mediations, including coordination through nonrecursive or selfrecursive counterparts, into concepts whose principal explanatory significance lies in human-human interrecursivity. It is a tendency, not an absolute prohibition. Goffman’s concepts have to be examined one by one, and some of them resist it. The tendency becomes clear when one asks, for each mediation, what Goffman observed, how he conceptualised it and what remained untheorised.

Embodiment is everywhere in his work. He observed posture, gait, trembling, blushing, the averted eye, the slack face of the long-stay patient, the fatigue of the waiter at the end of a shift. He conceptualised these as demeanour, body idiom, body gloss, flooding out and, where an attribute discredits, stigma (Goffman 1963a; 1963b; 1971). What remained untheorised was the body’s own ongoing coordination with its states, the way a person adjusts to pain, hunger, cold or exhaustion whether or not anyone is watching. Embodiment entered his concepts chiefly as a source of information available to others.

Multimateriality is equally pervasive. He observed furniture, décor, kitchen doors, partitions, uniforms, cosmetics, chairs, newspapers and beach towels. He conceptualised them as setting, the “furniture, décor, physical layout, and other background items which supply the scenery and stage props” (Goffman 1959: 22); as region, “any place that is bounded to some degree by barriers to perception” (Goffman 1959: 106); as involvement shields (Goffman 1963a); and as the stalls, use spaces and possessional territories through which persons press claims against one another (Goffman 1971: 28–61). What remained untheorised was the way objects coordinate conduct through their own nonrecursive persistence, independently of what they communicate: a chair that supports a body for an hour, a partition that blocks a draught as well as a gaze.

Multiversal dwelling, the non-human environmental conditions of coordination, appears less often but does appear, most strikingly in the “cliffs, water, forests, or moors” that seclude total institutions (Goffman 1961a: 4). It is conceptualised as barrier and as symbol. Its standing as a mediation in its own right, the terrain, light, weather and seasonal rhythms within which any institution must coordinate, remained outside the vocabulary.

Multisymbolization receives the most systematic treatment, in frames, keyings, classifications and definitions of the situation (Goffman 1974). Even here the translation operates, since symbols chiefly appear as resources through which interactants define what is going on for one another. What remained comparatively untheorised was the symbolic deposit that coordinates through persistence rather than through present definition, the record that is consulted long after its authors have left the room.

The self undergoes the same translation. At the close of The Presentation of Self in Everyday Life, the self becomes “a product of a scene that comes off”, for which the individual “and his body merely provide the peg on which something of collaborative manufacture will be hung for a time” (Goffman 1959: 252–253). The body here is a peg, a support for something whose real life is interactional. Goffman’s definition of the social situation makes the governing centre explicit: “an environment of mutual monitoring possibilities, anywhere within which an individual will find himself accessible to the naked senses of all others who are ‘present’, and similarly find them accessible to him” (Goffman 1964: 135). This is one of the finest definitions of interrecursive co-presence ever written, and the environment enters it only as the space across which mutual monitoring occurs. Late in his career Goffman described the whole field as “a domain which might be titled, for want of any happy name, the interaction order” (Goffman 1983: 2).

None of this rediscovers a materiality that others have missed. Trevor Pinch (2010) has shown that Goffman’s sociology is thick with technologies, from the merry-go-round of the role-distance essay to the equipment of everyday performance, and that these technologies mediate the situated activity systems Goffman analysed while remaining largely invisible in his theoretical vocabulary. David Alworth has read Goffman as a theorist of social form attentive to sites and settings, both in his account of how Goffman used Melville’s naval fiction as sociological evidence (Alworth 2014) and in a study of postwar fiction and art that draws on Goffman and Latour to treat supermarkets, dumps, roads, ruins and asylums as participants in social life (Alworth 2016). The present argument accepts these findings and asks a different question. The issue is less whether Goffman attended to objects, which he plainly did, than what explanatory standing his concepts allowed them, and why. LVT’s contribution lies in identifying nonrecursivity as an irreducible coordinative property, one that objects share with many other counterparts and that is lost both when objects are absorbed into interaction and when they are elevated into actors.

III. Nine Hundred Concepts: The Interrecursive Criterion of Sociological Admissibility

If the restricted disciplinary object shaped what could count as a sociological concept, it should leave a signature in concept formation, and Goffman’s unusually large conceptual inventory is the obvious place to look for it. The prediction is specific. Concepts should cluster around human-human interrecursivity, and concepts that refer extensively to nonrecursive or selfrecursive coordination should define their theoretical significance through it. The nine hundred cards thereby become a possible empirical test of the impossible object thesis.

No classification of all nine hundred has been carried out, and none is claimed here. What can be proposed is a systematic audit of representative concepts across Goffman’s principal publications, conducted on definitions and explanatory uses rather than on mere references to objects or persons. Counting how often chairs or doors appear would establish only what Goffman noticed, which is not in dispute. For each selected concept, the audit asks four questions: what phenomenon Goffman observes; what the concept explains; which counterpart types are involved; and whether nonrecursive coordination receives independent theoretical standing. The audit must also search actively for counterexamples, concepts that might theorise nonrecursive or selfrecursive coordination without making interaction their explanatory centre. Table 1 applies the procedure to twelve concepts as an illustration of the method, not as a census.

Table 1. An illustrative audit of twelve Goffman concepts

Dates refer to the works in which each concept is principally developed.

Three groups emerge. The first comprises concepts that directly theorise human-human interrecursivity: face-work, civil inattention, working consensus. Nothing is translated here, because the phenomena are themselves interrecursive, and these concepts are Goffman’s home ground. The second comprises concepts describing selfrecursion embedded in interrecursivity: impression management and role distance, in which a person adjusts to their own conduct because, or in the course of, its being available to others. The third group is the revealing one. Setting, region, territory, identity kit and institutional equipment refer extensively to nonrecursive counterparts, yet each defines their significance through human-human interaction. The chair matters as a stall that can be claimed or violated, the partition as a barrier to perception, the comb as an item of personal front.

The counterexamples are instructive precisely because they are partial. Frame Analysis admits the nonrecursive as a category, distinguishing natural frameworks for occurrences seen as “undirected, unoriented, unanimated, unguided, ‘purely physical’” from social frameworks that incorporate “the will, aim, and controlling effort of an intelligence, a live agency” (Goffman 1974: 22). But the distinction lives in the frames of participants, as a matter of how events are construed. The essay on normal appearances comes closer, borrowing the idea of an Umwelt to describe how individuals monitor their surround for signs of alarm, so that a creak, a shadow or an unfamiliar sound can break the ease of attention (Goffman 1971: 238–333). Here selfrecursive vigilance toward nonrecursive signs receives sustained analysis, although the alarm is referred predominantly to the designs of other persons. The strongest counterexample is the make-do, the use of available artefacts for ends not officially intended (Goffman 1961a: 207–208), which names a coordination between inmate and object directly. Even here, the concept’s defining contrast is with the official intentions of staff.

The result is a defensible qualitative pattern, not a quantitative finding. Nonrecursive and selfrecursive coordination are observed throughout Goffman’s work, are sometimes named, and are rarely granted standing independent of their interactional significance. The pattern also explains the proliferation. When every configuration of bodies, objects and others must be made sociological by referring it to interaction, each new configuration requires its own concept, because there is no general concept for the counterpart type that it shares with other configurations. The local concepts are precise and usable. What they lack is a general counterpart ontology within which they could be related. A full audit could justify a study of its own; here the pattern frames the reading of Asylums.

IV. Three Types of Recursive Counterparts

Living Value Theory distinguishes three types of recursive counterparts (Ecks 2026e). They describe relations in which a recursive being engages something, and they are neither three kinds of substance nor three kinds of person or institution. A nonrecursive counterpart is one toward which a recursive being orients and which does not orient back: a wall, a clock, a printed form, a pair of trousers, a locked door. A selfrecursive counterpart obtains when a recursive being engages some aspect of its own ongoing state and alters in relation to it. An interrecursive counterpart is another recursive being that responds to one’s responses, so that each party’s next move is shaped by its discernment of the other’s discernment.

Nonrecursive counterparts can change, fail and surprise. A lock can jam, a form can be lost, a building can burn. Their distinguishing feature is the absence of recursive responsiveness to an appeal, and it has nothing to do with predictability. One can rely on a nonrecursive counterpart or be ruined by it, but one cannot persuade it.

Selfrecursion deserves more attention than the sociological tradition has given it, because the translation problem affects it as much as it affects nonrecursive coordination. A person can coordinate with their own fatigue, posture, attention, sensation or ongoing action without that coordination being primarily oriented toward anyone else. A cleaner shifts weight as the back stiffens; a reader returns to a sentence that slipped past; a patient paces the ward until a restlessness eases; a sleeper turns over without waking. Goffman registers such moments with great precision, but they tend to enter his concepts when they become visible to others, as body gloss or flooding out, and to drop from conceptual view when they do not.

Where Goffman does theorise selfrecursion, his contribution is considerable. Role distance, illustrated by the older child who rides the merry-go-round with ostentatious nonchalance and by the surgeon who jokes through an operation, describes an actor recursively positioning themselves toward a role while communicating that positioning to others (Goffman 1961b). The concept combines selfrecursion and interrecursivity without making them identical. One can feel distance from a role without displaying it, and display a distance one does not feel. Role distance names the configuration in which the two are coupled, and it shows that selfrecursion can be organised for an audience without being reducible to that audience.

Civil inattention shows a different achievement. The brief glance that registers a passing stranger, followed by a withdrawal of attention that signals they are not a target (Goffman 1963a: 84), is interrecursivity organised to prevent further recursive engagement. Two recursive beings read one another precisely in order to agree, without words, not to go on reading.

These distinctions make it possible to state what Goffman was right about and what he could not say. He was right that the interaction order is analytically irreducible: its regularities cannot be derived from social structure, psychology or material conditions (Goffman 1983). But analytical irreducibility does not entail self-sufficiency. Every encounter is mesocosmically dependent, carried by bodies that are rested or exhausted, rooms that are lit or dark, doors that are open or locked, schedules that bring people together or keep them apart, records that precede them into the room. The interaction order is irreducible and dependent at once, and the three counterpart types allow both properties to be held in view without translating the second into the first.

V. Asylums: How Institutions Remove Interaction

A. Why Asylums Is the Decisive Case

Asylums grew out of Goffman’s fieldwork at St Elizabeths Hospital in Washington, D.C., in 1955 and 1956, a federal institution then holding some seven thousand patients (Ayaß 2022: 204). He worked nominally as an assistant to the athletic director and passed the day with patients, “avoiding sociable contact with the staff and the carrying of a key” (Goffman 1961a: ix–x). The key was the material threshold of the division between staff and inmates, and his refusal to carry one was the condition of his access to the inmate world. Goffman treated the total institution as an ideal type, conceding that “none of the elements I will describe seems peculiar to total institutions, and none seems to be shared by every one of them” (Goffman 1961a: 5), and critics have long noted that a model built from one hospital was asked to carry the weight of prisons, barracks, monasteries and camps (Ayaß 2022: 203–204). The book matters here for a different reason. Its most revealing descriptions concern arrangements that coordinate conduct by removing interaction, and so they confront an interaction-centred sociology with its own limit.

B. Prevention, Substitution, Restriction and Delegation

The total institution is “a place of residence and work where a large number of like-situated individuals, cut off from the wider society for an appreciable period of time, together lead an enclosed, formally administered round of life” (Goffman 1961a: xiii). Whereas in ordinary life “the individual tends to sleep, play, and work in different places, with different co-participants, under different authorities, and without an over-all rational plan”, the total institution breaks down these separations (Goffman 1961a: 5–6): the phases of the day are “tightly scheduled”, and the sequence of activities is “imposed from above by a system of explicit formal rulings and a body of officials” (Goffman 1961a: 6).

The organising question of this section is how institutions use nonrecursive counterparts to prevent, bypass, restrict or replace particular interrecursive encounters. Four mechanisms can be distinguished, and they should not be collapsed into one undifferentiated process of control.

Prevention makes an encounter unavailable. The locked door does more than symbolise staff authority: it prevents the encounter in which a patient might otherwise negotiate departure. Goffman noted that the encompassing character of total institutions is “symbolized by the barrier to social intercourse with the outside and to departure that is often built right into the physical plant, such as locked doors, high walls, barbed wire, cliffs, water, forests, or moors” (Goffman 1961a: 4). The list mixes multimaterial deposits with features of dwelling recruited as barriers, and it calls the whole assemblage “symbolized”. A high wall does symbolise, but its coordinative work lies first in the fact that it cannot be walked through, so that the question of leaving never reaches anyone who could answer it.

Substitution makes an encounter unnecessary. The timetable does more than express institutional control: it eliminates the repeated negotiations over when activities occur that would otherwise take place, many times a day, between every patient and every attendant. Rising, eating, bathing, medication and lights-out are settled once and then carried by the schedule, the bell and the clock.

Restriction narrows what can be coordinated. The issued uniform does more than mortify identity: it eliminates the choices and material practices through which the wearer might otherwise coordinate their appearance, from the selection of a garment in the morning to its mending, cleaning and storage. Restriction leaves interaction possible while removing the material resources from which some of its forms are composed.

Delegation transfers a decision across encounters. The case record does more than classify a patient: it carries decisions forward without requiring the original participants to explain or renegotiate them. Goffman’s account of the moral career shows the record selecting and fixing episodes that justify hospitalisation and travelling with the patient through a ward system whose grades of privilege materially enact a verdict about the self (Goffman 1961a: 146–159). Staff who never witnessed the recorded events read present conduct in their light.

The key claim is that institutions frequently organise coordination by making specific interrecursive encounters unnecessary or unavailable. Totality can be redefined accordingly: an institution becomes total when more of a person’s ordinarily distributed mesocosmic coordination is subjected to the same architecture of prevention, substitution, restriction and delegation. This exposes a conceptual irony. A sociology centred on interaction redescribes, in the vocabulary of interaction, arrangements whose coordinative achievement consists partly in the removal of interaction.

C. The Identity Kit: Observation and Conceptual Translation

Goffman’s account of admission shows the contrast between observation and translation with special clarity. He describes the procedures as “taking a life history, photographing, weighing, fingerprinting, assigning numbers, searching, listing personal possessions for storage, undressing, bathing, disinfecting, haircutting, issuing institutional clothing, instructing as to rules, and assigning to quarters” (Goffman 1961a: 16). The new arrival “allows himself to be shaped and coded into an object that can be fed into the administrative machinery of the establishment, to be worked on smoothly by routine operations” (Goffman 1961a: 16). Personal possessions are removed, among them the “identity kit” for the management of one’s personal front, the cosmetic and clothing supplies, the tools for adjusting and repairing them, and a secure place to store them (Goffman 1961a: 20). Standard issue replaces them, a number sometimes replaces the name, and the result is “personal defacement” (Goffman 1961a: 20). The recruit undergoes “a series of abasements, degradations, humiliations, and profanations of self”, so that “his self is systematically, if often unintentionally, mortified” (Goffman 1961a: 14).

As observation, this is meticulous. Goffman documents the removal and replacement of bodily and material coordinative resources item by item: weighing, fingerprinting, undressing, bathing, disinfecting and haircutting act on the body; searching, listing and storing possessions, issuing clothing and assigning quarters reorganise the material counterparts through which daily life had been coordinated; the life history, the number and the rules stabilise the transformation symbolically. As conceptual construction, the process is gathered under mortification, defacement and identity. These interpretations are valid. They concern consequences of a more extensive mesocosmic reorganisation, which they do not exhaust.

The point should not be put as a sequence in which bodies and things come first and identity follows, since the mediations are co-present and irreducible and the life history is taken in the same hour as the hair is cut. The analytical point is that the material and embodied reorganisation cannot be exhausted by its identity consequences. What the inmate loses with the identity kit includes the ability to present a self, and it also includes the capacity to coordinate with one’s own body, one’s own things and one’s own schedule in ways that were never addressed to anyone. Goffman came close to this when he wrote that the recruit arrives “with a conception of himself made possible by certain stable social arrangements in his home world” and is “immediately stripped of the support provided by these arrangements” (Goffman 1961a: 14). The phrase “stable social arrangements” carries the weight of the ontology he lacked, for those arrangements were the room with its own door, the garments chosen and mended, the habitual routes through a neighbourhood and the hours of sleep, as well as the people who answered to one’s name. What Goffman calls the mortification of the self is also a forced reorganisation of the coordinative conditions from which a self can continue to be recursively sustained, a reading his own essay on moral careers supports when it locates the self in “the pattern of social control” and concludes that the institutional arrangement “does not so much support the self as constitute it” (Goffman 1961a: 168).

D. Institutional Deposits

The formulation can now be stated precisely. Institutional power is frequently the power to replace a recursive counterpart with a deposit. This means replacing the need for a particular recursive negotiation with a durable coordinative arrangement. The institutional answer to “Can we negotiate this?” is often “It has already been decided”, and the decision is carried by a lock, a schedule, a category or a file that will not hear the question again. Hence the article’s title: at the points where it has deposited its decisions, the institution does not answer back.

Two clarifications are needed. First, an institution is a metastable architecture coordinating recursive beings through all five mediations and through all three types of counterpart relation, and never itself a nonrecursive counterpart (Ecks 2026d). Institutional stabilisation as a whole includes trained bodies, relational histories of trust and fear, environmental conditions and symbolic codes. Within that process, deposits are the particularly durable multimaterial and multisymbolic arrangements that carry previous decisions beyond the encounters in which they were made. Roles occupy a middle position. Inmate, patient, physician, nurse and attendant work as anticipatory compression devices, allowing recursive beings to treat one another as if their conduct were sufficiently predictable in advance, while the persons occupying them remain fully recursive. Goffman’s attendants are drawn into an “involvement cycle” of sympathy and withdrawal (Goffman 1961a: 82), and they carry the double task of enforcing obedience while upholding humane standards (Goffman 1961a: 92). The architecture never converts them into deposits.

Second, deposits do not simply reduce unpredictability. They can stabilise one domain while producing severe uncertainty in another. A patient may know to the minute when lights go out and have no idea when, or whether, discharge will come. The record is fixed, but how it will be read at the next case conference is not. The privilege system makes small rewards dependable while leaving the patient unable to predict which act will be entered against them (Goffman 1961a: 48–51). Deposits redistribute predictability, and its redistribution is itself a form of institutional power.

E. Looping

Goffman’s concept of looping identifies a distinctive mechanism. In the total institution, he observed, the agency that provokes a defensive response from an inmate “takes this very response as the target of its next attack” (Goffman 1961a: 35–36). The clinician or attendant who receives the response remains recursive, and may even be sympathetic. What changes is the interpretive arrangement through which the response is received. Looping can therefore be defined as interrecursivity routed through a prior symbolic deposit that converts contestation into confirmation. A patient’s objection to a diagnosis becomes evidence for the diagnosis; insolence confirms the need for a stricter ward; withdrawal confirms the illness. The response is heard, but it cannot readily reopen the classification through which it is heard.

Looping is thus a particularly revealing mechanism of institutional recursive capture: the inmate’s recursive capacity is engaged, provoked and then absorbed into the deposit it contests. Goffman’s usage should be distinguished from Ian Hacking’s later account of the looping effects of human kinds, in which classifications change the people classified, whose changed conduct in turn obliges the classifications to change (Hacking 1995; see also Hacking 2004). Hacking’s loop runs across populations and over time and tends to reopen classifications. Goffman’s loop runs within an establishment and an encounter and tends to close them.

F. Secondary Adjustments and Underlife

Against the architecture of deposits stands the underlife. Secondary adjustments are “practices that do not directly challenge staff but allow inmates to obtain forbidden satisfactions or to obtain permitted ones by forbidden means” (Goffman 1961a: 54), and “The Underlife of a Public Institution” catalogues them with extraordinary density: make-do’s, ways of working the system, free places, group and personal territories, stashes and informal systems of transport, communication and exchange (Goffman 1961a: 207–248). Its insight is compressed into one sentence: “To prescribe activity is to prescribe a world; to dodge a prescription can be to dodge an identity” (Goffman 1961a: 187).

Secondary adjustments come in several kinds. Some repurpose materials: a radiator becomes a place to dry a washed shirt, a coat lining becomes a stash. Some exploit gaps in institutional rules or in the reach of surveillance. Some create alternative relationships, among patients or between patients and particular attendants, for which the official classification has no place. Some reopen classifications themselves, as when a patient succeeds in being treated, in a corner of the hospital, as a competent worker rather than a case. And some become routinised, sedimenting into stable informal arrangements that new inmates learn as they learn the official rules. Not every one of them is a recursive opening in the strict sense, a reopening of a question the institution had settled.

Goffman’s own distinction between “disruptive” secondary adjustments, which aim to abandon the organisation or radically alter its structure, and “contained” ones, which fit into existing structures without introducing pressure for radical change (Goffman 1961a: 199), warns against romanticising any of them. Contained adjustments can help stabilise the institution they evade, and a ward may tolerate a stash because the tolerated stash keeps the ward quiet. The theoretical contribution of the underlife lies elsewhere than in resistance. It shows that institutional compression never abolishes the possibility of alternative coordination. The deposits remain, but recursive beings find in them possibilities that the deposits were never arranged to close, and some of these possibilities, though by no means all, become openings.

G. An Extreme, Not a Different Species

Total institutions are analytically valuable because they expose mechanisms that every institution employs. A university prevents, substitutes, restricts and delegates with locked offices, timetables, submission formats and transcripts; a bank with counters, forms and ledgers; a household with a house, a calendar and an inheritance. The total institution extends these mechanisms unusually far across time, space, body, possessions, relationships and classification, until the process becomes visible as a process. Goffman called total institutions “the forcing houses for changing persons; each is a natural experiment on what can be done to the self” (Goffman 1961a: 12). Each is also a natural experiment on how far coordination can be withdrawn from interrecursive encounters and carried by deposits before recursive beings find alternative ways to coordinate.

VI. What Living Value Theory Inherits from Goffman

The diagnosis of conceptual translation should not obscure what Goffman discovered. Several of his concepts identify empirical distinctions that LVT would otherwise be less well equipped to describe, and they enter LVT as findings rather than as terms awaiting translation.

Civil inattention shows that successful interrecursivity may require mutual restraint rather than further engagement. An ontology that defined interrecursivity as intensifying mutual responsiveness would miss the large part of urban and institutional life in which recursive beings coordinate by agreeing, through a glance, not to engage. Working consensus, the provisional agreement “as to whose claims concerning what issues will be temporarily honored” rather than “a real agreement as to what exists” (Goffman 1959: 9–10), shows how coordination can continue without shared beliefs or full agreement. It is an indispensable corrective to any theory that grounds coordination in shared meaning. Face-work specifies the local processes of relational recognition and repair through which encounters survive threats to the value participants claim (Goffman 1967: 5–45). LVT can locate these processes in being-with at L2 and L3, but it could not have specified their fine mechanics, the avoidance, the corrective interchange, the offering and the acceptance, without Goffman.

Normal appearances help explain how recursive certainty is maintained without constant checking (Goffman 1971: 238–333). A person can move through a world of nonrecursive and interrecursive counterparts in ease because most of the surround is monitored at a level below articulation, and alarm is triggered by departures from what is expected. This is an analysis of L1 coordination and its disruption at L2 that LVT inherits almost directly. Role distance, as shown above, identifies the coupling of selfrecursion and interrecursivity without their identity. Looping identifies a distinctive mechanism through which institutional classifications absorb challenges, a mechanism that recurs wherever symbolic deposits stand between recursive beings: in courts, welfare offices, schools and clinics. Secondary adjustments and underlife show that compression never exhausts the possibilities of coordination within a compressed field, and that the alternatives are as varied as the deposits they work around.

The encounter is therefore two-way. LVT supplies the counterpart ontology that Goffman’s concepts lacked, and Goffman supplies distinctions that a counterpart ontology alone would not have generated.

VII. Why Goffman Cannot Complete the Theory of Institutions

The classic question about Goffman asks how micro-interaction relates to macrostructure. From the standpoint developed here, the question is an artefact of the impossible object. Once social theory had been confined to human-human being-with, the large-scale persistence of coordination had to be explained either as an aggregate of interactions or as a structure hovering above them. Institutional durability, built from sedimented coordinative processes operating across all five mediations and all three counterpart types, can be explained as neither, and the opposition between micro and macro begins after the ontological mistake has been made.

Goffman’s late formulation shows how close he came. He described the relation between interaction and social structure as “a nonexclusive linkage” and a “‘loose coupling’”, in which interaction works as “a set of transformation rules, or a membrane selecting how various externally relevant social distinctions will be managed within the interaction” (Goffman 1983: 11). LVT can explain why the coupling is loose. The larger institutional order includes accumulated deposits that endure beyond any encounter. A sympathetic nurse may treat a patient with warmth while the diagnostic code, the locked ward, the staffing rota and the order of commitment remain as they were. Interaction is far from powerless, since interrecursive action can reopen and transform deposits, as reforms of commitment law, patients’ movements and the closure of the great asylums showed in the decades after Asylums. But particular encounters may lack the authority or the practical means to do so, and the conditions under which an encounter can reach a deposit are themselves distributed by the institution.

Niklas Luhmann’s systems theory encounters a related restriction despite its more explicit theory of stabilisation. His double contingency is close to interrecursivity, and his account of institutions as expectations that are maintained even when disappointed is close to recursive compression (Luhmann 1995). Yet in making communication the constitutive element of the social, Luhmann placed bodies, buildings, documents as physical objects and environmental conditions in the environment of social systems, outside the operations that perform stabilisation. Goffman sees the process of institutional removal without a theory of the compression it achieves; Luhmann theorises compression while excluding much of the architecture through which it is achieved.

VIII. From Goffman to Sofsky: When Compression Becomes Destructive

Goffman included the concentration camp among the establishments to which the total-institution model could be extended, drawing on survivors’ accounts to illustrate admission, mortification and the privilege system. Wolfgang Sofsky’s The Order of Terror (1997) can be read as the most rigorous inheritor of this microsociological attention, and as a demonstration that the camp cannot be understood as a more extreme psychiatric institution.

Institutional compression can be used to sustain coordination, to constrain it or to destroy the conditions of viable life. These uses must not be mapped neatly onto institutional types. Psychiatric institutions were by no means necessarily protective or nonlethal, and their formal purposes of care and cure need to be distinguished from their actual practices and consequences, which in many mid-century hospitals included neglect, injury, coercive treatment and premature death. The difference lies in the relation between deposits and the continuation of life. However badly, the hospital’s architecture was organised around its inmates’ continued existence. The camp’s architecture was organised so that compression attacked the conditions of viable coordination themselves: food, sleep, warmth, movement, relationship and time.

Sofsky’s analysis also shows something that bears directly on the theory of deposits. In the camp, stable deposits coexisted with arbitrary application. The roll call, the block, the wire, the categories marked on clothing and the schedule of labour were fixed, while their application by guards and functionaries was made unpredictable, so that institutional organisation produced terror rather than dependable coordination. Prisoners could neither negotiate with the architecture nor rely on it. This is important evidence against equating nonrecursivity with predictability. The camp’s deposits did not answer back, and they did not make the next hour knowable either. Goffman and Sofsky both observed the transfer of coordination from interrecursive encounters to imposed deposits; neither possessed an ontology that identified this transfer as a general mechanism whose effects depend on how it is used.

IX. Conclusion: The Solid Buildings and the Cracks

The argument began with nine hundred index cards. Goffman was an exceptionally attentive observer of the entire mesocosm: of bodies and their states, of objects and their persistence, of rooms, doors and grounds, of records and categories, and of the encounters in which all of these are at stake. Yet the conceptual expectations of sociology repeatedly drew his explanations toward human-human interrecursivity, and each time an observation had to be made sociological it acquired a concept of its own. His conceptual proliferation is evidence of the extraordinary pressure that a restricted disciplinary object exerted upon an exceptionally observant theorist. It is also evidence of his resistance to that pressure, since he kept recording what the criterion of admissibility would have allowed him to leave out.

His descriptions of institutions reveal the limitation with special clarity, because institutions frequently stabilise coordination by preventing further interrecursive negotiation. Locks, timetables, uniforms and records do their work by making certain encounters unnecessary or unavailable, and a vocabulary built for encounters can describe them only by translating them back into what they remove. The closing sentences of Asylums nonetheless state the relation between deposits and recursive possibility more exactly than any later commentary: “Our sense of being a person can come from being drawn into a wider social unit; our sense of selfhood can arise through the little ways in which we resist the pull. Our status is backed by the solid buildings of the world, while our sense of personal identity often resides in the cracks” (Goffman 1961a: 320). The solid buildings of the world are the deposits, the walls, wards, schedules and files that carry earlier decisions and no longer answer back. The cracks are the places where recursive beings, in coordination with their own bodies, with things and with one another, continue to find ways of living that the buildings were never arranged to close.

References

Alworth, David J. 2014. “Melville in the Asylum: Literature, Sociology, Reading.” American Literary History 26 (2): 234–261.

Alworth, David J. 2016. Site Reading: Fiction, Art, Social Form. Princeton: Princeton University Press.

Ayaß, Ruth. 2022. “Totale Institutionen.” In Goffman-Handbuch: Leben, Werk, Wirkung, edited by Karl Lenz and Robert Hettlage, 201–205. Berlin: J. B. Metzler.

Goffman, Erving. 1959. The Presentation of Self in Everyday Life. Garden City, NY: Doubleday Anchor.

Goffman, Erving. 1961a. Asylums: Essays on the Social Situation of Mental Patients and Other Inmates. Garden City, NY: Doubleday Anchor.

Goffman, Erving. 1961b. Encounters: Two Studies in the Sociology of Interaction. Indianapolis: Bobbs-Merrill.

Goffman, Erving. 1963a. Behavior in Public Places: Notes on the Social Organization of Gatherings. New York: Free Press.

Goffman, Erving. 1963b. Stigma: Notes on the Management of Spoiled Identity. Englewood Cliffs, NJ: Prentice-Hall.

Goffman, Erving. 1964. “The Neglected Situation.” American Anthropologist 66 (6, part 2): 133–136.

Goffman, Erving. 1967. Interaction Ritual: Essays on Face-to-Face Behavior. Garden City, NY: Doubleday Anchor.

Goffman, Erving. 1971. Relations in Public: Microstudies of the Public Order. New York: Basic Books.

Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. New York: Harper and Row.

Goffman, Erving. 1983. “The Interaction Order.” American Sociological Review 48 (1): 1–17.

Hacking, Ian. 1995. “The Looping Effects of Human Kinds.” In Causal Cognition: A Multidisciplinary Debate, edited by Dan Sperber, David Premack and Ann James Premack, 351–383. Oxford: Clarendon Press.

Hacking, Ian. 2004. “Between Michel Foucault and Erving Goffman: Between Discourse in the Abstract and Face-to-Face Interaction.” Economy and Society 33 (3): 277–302.

Lenz, Karl, and Robert Hettlage, eds. 2022. Goffman-Handbuch: Leben, Werk, Wirkung. Berlin: J. B. Metzler.

Luhmann, Niklas. 1995 [1984]. Social Systems. Translated by John Bednarz Jr. with Dirk Baecker. Stanford: Stanford University Press.

Pinch, Trevor. 2010. “The Invisible Technologies of Goffman’s Sociology: From the Merry-Go-Round to the Internet.” Technology and Culture 51 (2): 409–424.

Sofsky, Wolfgang. 1997 [1993]. The Order of Terror: The Concentration Camp. Translated by William Templer. Princeton: Princeton University Press.

Williams, Robin. 1988. “Understanding Goffman’s Methods.” In Erving Goffman: Exploring the Interaction Order, edited by Paul Drew and Anthony Wootton, 64–88. Cambridge: Polity Press.

Williams, Robin. 2000 [1983]. “Sociological Tropes: A Tribute to Erving Goffman.” In Erving Goffman, vol. 2, edited by Gary Alan Fine and Gregory W. H. Smith, 105–110. London: Sage. Originally published in Theory, Culture and Society 2 (1).