[This paper was given at a conference on Erving Goffman and the concept of "total institutions," held in Grenoble in Novermber, 1999.]

The Politics of Presentation: Goffman and Total Institutions

The Problem of Conventional Categories

Erving Goffman's essay, "On the Characteristics of Total Institutions," is a classic example of the relation between methods of presenting research and scholarship and their political content, which is never simple and direct. It exemplifies, in fact, Goffman's presentational solution to a problem which has always plagued social science writing and research: how to avoid the analytic flaws and failures which arise from our unthinking acceptance of the constraints of conventional thinking.

Let me explain the problem. When social scientists study something--a community, an organization, an ethnic group--they are never the first people to have arrived there, newcomers to an unpeopled landscape who can name its features as they like. Every topic they write about is part of the experience of many other kinds of people, all of whom have their own ways of speaking about it, specialized words for the objects and events and people involved in that area of social life. Those special words are never neutral objective signifiers. Rather, they express the perspective and situation of the particular kinds of people who use them. The natives are already there and everything in that terrain has a name, more likely many names.

If we choose to name what we study with words the people involved already use, we acquire, with the words, the attitudes and perspectives the words imply. Since many kinds of people are involved in any social activity, choosing words from any of their vocabularies thus commits us to one or another of the perspectives already in use by one or another of the groups there. Those perspectives invariably take much for granted, making assumptions about what might better be treated, social scientifically, as problematic.

Take the topic of marijuana. People who use it have a language for talking about it. They speak of "getting high." They have many synonyms for marijuana, speaking of it, for instance, as "grass." They might speak of the person they buy marijuana from as a "connection." Other people, whose worlds also contain marijuana--physicians, lawyers, police--will have other words for the same things, perhaps speaking of "addiction" and "cannabis" and "pushers." The language of users suggests that use is voluntary, pleasurable, innocent; the language of some others suggests that it is involuntary, evil, harmful.

What things are called almost always reflects relations of power. People in power call things what they want to and others have to adjust to that, perhaps using their own words in private, but accepting in public what they cannot escape. Whatever I and my friends think, marijuana is defined as a narcotic drug by people who can make that name and perspective stick.

The social scientist's problem, simply, is what to call the things we study. If I study marijuana, do I speak of "marijuana addiction" or, as I chose to do, in a minor linguistic variation that connoted a serious shift in perspective, of "marijuana use"? Do we speak of "getting high on," of "being intoxicated by," or of "being under the influence of" this substance?

If I choose the terms used by the people who "own" the territory, and therefore choose the perspectives associated with those terms, I let my analysis be shaped by conventional social arrangements and the distribution of power and privilege they create. This has both technical and moral consequences.

The technical consequence is that the class of phenomena I want to generalize about is made up of things which have in common only the moral attitudes toward them of powerful people and groups in the society, and the actions that have been taken toward them in consequence. The result of that is a tremendous difficulty in finding anything general to say about the phenomena, other than things associated with those moral attitudes. You can talk about the results of being thought about that way--that's what the labeling theory of deviance did. But you can't find anything to say about how people get that way, underlying causes, or similar matters, because there is nothing related to those matters which all the cases in the class have in common. You can't make science if you can't find anything to generalize about.

The moral consequence of adopting existing language and perspectives toward the phenomena we study is that we accept, willingly or not, all the assumptions about right and wrong contained in those words and ideas. We accept, in the case of drugs, the idea that addicts are people who have lost control of themselves and therefore cannot help doing things which are inherently bad.

This was Goffman's problem as he began to write about the mental hospital he had studied. The existing language for discussing the people confined in such institutions embodied the perspectives of the people who were able to confine others in them--the professional staff who ran them, the legal professionals who assigned people to them, the families who had solved their problem with an unruly family member, the police for whom they were what is sometimes called a public nuisance. How could he avoid taking such categories as "mental illness' and the perspectives associated with them for granted?

The Linguistic Solution

To make clear how Goffman found a workable solution to the problem of conventional categories, I'll begin with a simple observation. No reader of Goffman's essay on total institutions can be unaware of the considerable disparity between the social reality he is talking about and the way he talks about it. He describes and analyzes social practices which are quite common, whose existence and character are known to most adults, if not through their personal experience then through the experience of others they know and through secondhand descriptions in the press, films, drama, and fiction. He describes and analyzes organized social practices of incarceration and degradation which repel and even disgust many readers, and which arouse feelings of shame in us for living in a society in which such things have happened and continue to happen. His detailed and comprehensive descriptions make it impossible to ignore the continued existence of these organized, socially accepted activities, and have on occasion instigated attempts at their reform.

The disparity I mentioned exists, first, in the language he uses to make these descriptions. For, despite the repellent nature of many of the activities he describes, he never uses judgmental language. He does not explicitly denounce the practices his descriptions make us want to denounce, nor does he use adjectives and adverbs which betray a negative assessment of them. He might just as well be describing an ant hill or a bee hive as a common form of social institution which treats some people (never forget, with the complicity of the rest of society, and that does mean us) in such a way that their lives resemble those of members of those insect societies: regimented in an inflexible and humiliating caste system without regard for their own feelings or wishes. His detailed description of what we might find in such places brings us to that sort of conclusion, though he never says anything like that himself. Here are some of the ways he uses language to avoid built-in judgments.

He uses the word "echelon" (instead of, for instance, "domination") to describe the typical authority system of a total institution: "any member of the staff class has certain rights to discipline any member of the inmate class, thereby markedly increasing the probability of sanction." (p. 42) The word is neutral. Since it is not commonly used for this purpose, it has no immediately negative connotations, such as a term like "domination" would have. It simply describes one way among many of organizing authority relations, just as Weber's distinction between charismatic, bureaucratic, and traditional describes three other ways.

I will note, without providing argument or examples, that it is far easier to find examples of "echelon control" than of "domination." The former simply requires demonstration of an observable fact--who gives orders to who--while the latter includes, scarcely beneath the surface, a judgment as to the moral suitability of the order-giving arrangement, which is always more arguable.

Some further examples of this kind of neutral language used to describe matters about which readers such as ourselves would probably have strong negative feelings are:

ï "role dispossession" to describe how new recruits are prevented from being who they were in the world they previously inhabited

ï"trimming" and "programming" to describe how "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." (p. 16)

ï"identity kit," to indicate the paraphernalia people ordinarily have with which to indicate who they are but which is routinely denied inmates in total institutions

ï"contaminative exposure," to indicate ways inmates are humiliated and mortified in public

ï"looping," to indicate how an inmate's attempt to fight mortification lead to more mortification

ï"privilege system," to indicate the way withheld ordinary rights become privileges used to coerce conformity

ï"secondary adjustments," to refer to "practices that do not directly challenge staff but allow inmates to obtain forbidden satisfactions or to obtain permitted ones by forbidden means" (p. 54)

ïa variety of "personal adjustments," such as "situational withdrawal," which (he notes) psychiatrists might call "regression"

He also uses words that have negative overtones, but uses them in a neutral way, so that they lose their negative charge. For instance, he talks of new recruits being "mortified," but examples of this include officer candidates in military organizations.

He discusses the staff by treating what they do as a kind of work (thus showing himself the student of Hughes he often claimed to be), to be seen as part of a series that includes many other kinds of work. He emphasizes that the work of the staff of a total institution deals with people, rather than inanimate stuff, and that creates distinctive problems.

The multiplicity of ways in which inmates must be considered ends in themselves, and the large number of inmates, forces upon the staff some of the classic dilemmas that must be faced by those who govern men. Since a total institution functions somewhat as a state, its staff suffers somewhat from the tribulations that beset governors. (P. 77) Here too he uses the linguistic devices I have discussed, speaking "objectively" of staff work as dealing with "human objects" or "human material."

The Comparative Solution

The disparity I spoke of also exists in the comparative procedure he uses to arrive at the ideal type of the total institution. He creates this type, remember, by comparing a variety of organizations found in modern societies which have an important distinguishing characteristic, and abstracting from them their common features. He first defines the general class of "social establishments" as consisting of "places such as rooms, suites of rooms, buildings or plants in which activity of a particular kind regularly goes on," and speaks of the difficulty of classifying members of this class. Nothing could be more "neutral" or "scientific." He then classifies establishments, roughly, by their relations to the lives of the individuals who participate in them. Some institutions will not accept people of certain kinds at all. Many institutions have a changing population of customers or workers. Others, like families, change their personnel less frequently. Some institutions house activities their participants take seriously; others are for more frivolous activities.

This dispassionate sorting of social organizations in the essay's first paragraph--treating families, leisure time activities, and workplaces as all equal, simply establishments which vary along one or more dimensions--warns us that Goffman will not be engaging in social science as usual. Ordinary social science, unlike Goffman, typically uses as classificatory categories the words, and their associated judgments of moral and social worth, common in the organizations being analyzed. The distinction, for instance, between "deviant' and "normal" activities contains just such judgments, common in the legal and therapeutic organizations which deal with the matters conventionally so classified. So do classifications of organizations and activities as "functional" or, more clearly, "dysfunctional." And these are categories that intend to be scientific and dispassionate. The judgmental character of social science categories is clearer yet in more politically and ethically engaged research and writing, which routinely use terms like "repressive" or "corrupt" to describe the phenomena they analyze.

He treats theories in social science and related areas as raw material, whose analysis will reveal the basic character of the institutions which use them rather than as "science," as in his offhand discussion of psychiatric thinking:

Mental hospitals stand out here because the staff pointedly establish themselves as specialists in the knowledge of human nature, who diagnose and prescribe on the basis of this intelligence. Hence in the standard psychiatric textbooks there are chapters on "psychodynamics" and "psychopathology" which provide charmingly explicit formulations of the "nature" of human nature. (P. 89) Needles to say, he explains that the purpose of these theories is to validate the methods used to accomplish the end of managing large numbers of people under the conditions of a total institution.

Having defined social establishments, Goffman immediately proposes yet another principle for their classification, one which will separate out a group whose "members appear to have so much in common . . . that to learn about one of [them} we would be well advised to look at the others." He then isolates the defining characteristic of this class this way:

Every institution captures something of the time and interest of its members and provides something of a world for them; in brief, every institution has encompassing tendencies. When we review the different institutions in our Western society, we find that some are encompassing to a degree discontinuously greater than the ones next in line. Their encompassing or total character 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. These establishments I am calling total institutions, and it is their general characteristics I want to explore. So: institutions take up varying amounts of the time and interest of the people who participate in them, from a little to a lot. Some take up so much of their participants' time and lives that they are "discontinuous" with others in this array. They are "total institutions." He distinguishes among the institutions this single criterion isolates by whether people are confined in them because they can't take care of themselves, because they are a danger to others, or both, or whether they are so isolated in order to better accomplish some important work or as a retreat from the world for religious or similar purposes. His analysis will look for the other features which commonly accompany this sort of total control over the lives of the people in them, who he soon starts calling "inmates," thus adopting for the whole class (including nuns, priests, soldiers, and others not usually thought of as being incarcerated) the demeaning term typically used in mental hospitals.

Goffman's analytic tack emphasizes the disparity between the kind of place he is talking about and the way he talks about it. Though he will, through most of the essay, be discussing places about which we routinely make strongly negative judgments--mental hospitals, concentration camps, prisons--he treats them as members of the same class as organizations about which we usually make no such simple negative judgments--military establishments, ships at sea, or religious retreats. This creates what seems to be a moral confusion at the heart of his method, for we are confronted with a classification that combines and treats as equivalent things which, as morally competent members of our society, class, and profession, we "know" are morally quite disparate. We may be anti-militarist, but most of us do not think that army camps are concentration camps. We may have little sympathy for organized religion, yet not be ready to say that monasteries or convents are prisons.

The comparative method works by establishing, as we have seen, a common dimension along which a variety of cases can be ranged. So there is a dimension of how much of the person's time an establishment controls, and organizations vary widely in this respect. Some--a tennis club that one belongs to, for instance--control very little while others--a family--control more. There is a general problem or question of how people's time is divided among the groups they participate in and the total institution takes its place as providing one of the many possible resolutions of this question. The total institution no longer stands out as aberrant--as though the social world was divided into institutions and practices which are "ordinary" or "normal" and do not ask for an abnormal commitment from a person, and then there is this strange one, completely different, which requires total control. Instead of being different and strange it is now just a different reading on a dial, another of the possible positions on this scale. This is not a trivial result.

An example, He describes how three classes of total institutions give differing rationales for "assaults on the self": religious institutions say they are good for people, assisting them to reach a goal they aspire to; prisons and concentration camps do it for the sake of mortification itself; others excuse themselves on the grounds that it is necessary for some other important purpose to be achieved (e.g., military readiness or security). Then he says that, in all three classes, these rationales are rationalizations "generated by efforts to manage the daily activity of a large number of persons in a restricted space with a small expenditure of resources," (pp. 46-7)

The Technical and Moral Result

The avoidance of built-in judgment is not evidence of a moral confusion on Goffman's part. He was not a moral dope (to adapt Garfinkel's famous description of the homunculus in most sociological theorizing as a "cultural dope)". Far from it. Any careful reader feels, beneath the cool, unemotional language of Goffman's essays in this volume, the beating heart of a passionate civil libertarian. By adopting this method, which entailed both antiseptic "scientific" language and a non-judgmental comparison of cases, Goffman found a solution to the problem of the assumptions built in to conventional thinking.

If you accept the conventional categorizations built into ordinary language and the ordinary way that institutions and practices are sorted out in conventional thought, if you unthinkingly refer to people who drink a lot of alcohol as alcoholics, if you refer to people who smoke marihuana as addicts, then you accept the ideas that those words more or less oblige you to accept, ideas built in to the words themselves and into the perspectives associated with them. If a person who smokes marihuana is an "addict," then that person will smoke it uncontrollably, will be a "slave" to the practice, will engage in crimes in order to pay for the drug, and so on. If you use these words to define the class you are studying, as I have argued above, you will not find empirical regularities to make scientific generalizations about.

By using the neutral language he constructs to discuss total institutions, Goffman isolates a class of social objects which have well-defined characteristics in common, characteristics which are empirically observable and can be connected to one another in verifiable patterns. He can make science.

You may wonder why I speak so insistently of "making science." It is not often appreciated to what degree Goffman was a serious empiricist, even perhaps what might be called (in some meaning of the term) a positivist. (In this he resembled, I might say in passing, Margaret Mead.) He believed that there was an empirical reality and was wary of anything that smacked of the supernormal, could not be verified empirically, or was overly speculative.

(Perhaps I will be permitted a personal reminiscence here. Many years ago, when he was teaching at Berkeley, Goffman asked me to come to his seminar to hear a student, Marvin Scott, present his research on horse racing. This excellent research dealt with the way the social organization of what he called "The Racing Game" made it reasonable for some trainers, owners, and jockeys to want their horse to lose, rather than win. However, in the course of his presentation, Scott suggested in passing that gamblers, including horse players, sometimes had "winning streaks" or "losing streaks." Goffman, who had been listening appreciatively until that point, interrupted to say that of course Scott meant that they thought they had such streaks of good or bad luck. But Scott said no, these were observable "facts." Goffman, unwilling to accept such supernatural talk, persisted, appealing to the laws of probability to assure Scott that such "streaks" were natural occurrences in any long run of tries in such a game as blackjack or craps. And finally exploded in anger at Scott's "unscientific" insistence on luck as a natural phenomenon.)

Goffman used his linguistic inventiveness to name things in ways that evaded conventional moral judgments and thereby to make scientific work possible. Instead of pointing with scorn at the "inhuman practices" of mental hospitals, or defending them as honest professionals doing the best they could with a difficult job, he situated their activities in a context of organizational necessity they shared with other organizations with widely varying degrees of moral repute. The resulting generalizations made possible a deeper understanding of these phenomena than either denunciation or defense ever had.

His generalizations about total institutions simultaneously made possible a far more serious moral evaluation of those practices, since the judgment was now based on a more than superficial understanding of what the moral choices actors had to make actually were. This leads, inevitably, to blaming organizations rather than individuals, and not even to blaming them for doing what they have to do under the circumstances in which they exist. It is never easy to assign blame for what a whole society, in all its parts, is responsible for. As Goffman explains:

I have defined total institutions denotatively by listing them and then have tried to suggest some of their common characteristics. . . . the similarities obtrude so glaringly and persistently that we have a right to suspect that there are good functional reasons for these features being present and that it will be possible to fit these features together and grasp them by means of a functional explanation. When we have done this, I feel we will give less praise and blame to particular superintendents, commandants, wardens, and abbotts, and tend more to understand the social problems and issues in total institutions by appealing to the underlying structural design common to them all. (Pp. 123-4)
References

Goffman, Erving. 1961. Asylums. New York:Anchor.

Scott. Marvin B. 1968. The Racing Game. Chicago: Aldine Publishing Co.



El diseño de la página y las imágenes son
© 1996-2000 Grupo "mmm"
Comentarios a: © Dr.Vicent Fores
València  15th September 2000