The permeable institution: An ethnographic study of three acute psychiatric wards in London

Abstract

In Asylums, Goffman [1961. Asylums. London: Penguin] identified some permeable features of the old mental hospitals but presented them as exceptions to the rule and focused on their impermeable aspects. We argue that this emphasis is no longer valid and offer an alternative ideal type that better represents the reality of everyday life in contemporary ‘bricks and mortar’ psychiatric institutions. We call this the “permeable institution”. The research involved participant observation of between 3 and 4 months and interviews with patients, patient advocates and staff on 3 psychiatric wards.

Evidence for permeability includes that ward membership is temporary and changes rapidly (patients tend to have very short stays and staff turnover is high); patients maintain contact with the outside world during their stay; and institutional identities are blurred to the point where visitors or new patients can easily mistake staff and patients for one another. Permeability has both positive consequences (e.g., reduced risk of institutionalism), and negative consequences (e.g., unwanted people coming into hospital to cause trouble, and illicit drug use among patients). Staff employ various methods to regulate their ward's permeability, within certain parameters.

The metaphor of the total/closed institution remains valuable, but it fails to capture the highly permeable nature of the psychiatric institutions we studied. Analysts may therefore find the permeable institution a more helpful reference point or ideal type against which to examine and compare empirical cases. Perhaps most helpful is to conceptualise a continuum of institutional permeability with total and permeable institutions at each extreme.

Introduction

Institutions like mental hospitals are of the ‘total’ kind, in the sense that the inmate lives all aspects of his life on the premises in the close company of others who are similarly cut off from the wider world. These institutions tend to contain two broad and quite differently situated categories of participants, staff and inmates (Goffman, 1961, pp. 183–184).

Erving Goffman was one of the first sociologists to examine the experiences of patients in order to understand how mental hospitals work. Like other researchers of that time, he found those institutions to be disturbingly easy to be admitted into, extraordinarily difficult to get out of, and damaging to in-patients who were isolated from the outside world (Caudhill, 1958; Strauss, Schatzman, Bucher, et al., 1964; Rosenhan, 1973). Goffman's Asylums (1961) was particularly influential, and his total institution model or metaphor remains firmly planted in the minds of sociologists, psychiatrists and service user advocates (Weinstein, 1994). Even so, some have argued that the model was out of date even at its inception (Weinstein, 1994). This is not our view. Rather, in this paper we will argue that the total institution firstly needs to be understood for what it was; namely an ideal type against which empirical cases can be examined and compared. Using findings from previous research and our own ethnographic study of life on three acute psychiatric (admission) wards in London, we shall argue that there has been a trend over time towards an increased degree of permeability on such wards and the hospitals in which they are located. Further, we offer an alternative ideal type to the total institution, which we believe better represents the reality of everyday life in contemporary ‘bricks and mortar’ psychiatric institutions. We call this the “permeable institution”.

In his social anthropological study of a psychiatric hospital, Caudhill (1958) argued convincingly that the old psychiatric hospitals were closed social systems that affected the behaviour of people who made them up. He observed a formal social structure with a sharply defined status hierarchy of physicians, psychologists, social workers, administrators, clerical workers, attendants and patients, and noted that these groups developed separate values and perceptions of hospital life. By looking beyond interaction within or between social groups to transaction—the processes going on throughout the hospital—Caudhill was able to track and account for phenomena such as sudden ‘mood sweeps’ and collective disturbances (Caudhill, 1958). But the relationship between life in hospital and the outside world was largely unexplored.

The process of deinstitutionalization has brought most of these old mental hospitals to a close in many developed countries (Fakhoury & Priebe, 2002; Lamb & Bachrach, 2001). In the UK, the number of hospital beds has been dramatically reduced (Becker et al., 2002), and those that remain have been relocated to psychiatric units in general hospitals. Although very few patients now live in a psychiatric hospital, the mental health care system retains the function of admitting people briefly in order to treat their mental disorder or to prevent them from causing harm to themselves or others. Indeed in many European countries, and elsewhere, ‘acute’ psychiatric (admission) wards are an important component of a system comprising a diverse range of residential, day and community services (Becker et al., 2002). It is not known how many of the estimated 1.85 million psychiatric beds worldwide, equating to 4.36 per 10,000 population (WHO, 2001), are in acute psychiatric wards. The ratio of acute to long-stay beds is likely to vary greatly from country to country and be highest in those countries that have undergone deinstitutionalization.

This process was well underway when Baruch and Treacher (1978) examined the workings of a general hospital psychiatric unit. However, they found a geographically isolated institution that experienced problems due to its impermeability to the outside world:

[The unit's] relationship with the community was so weak that many of the specific problems that psychiatric units were meant to avoid were depressingly still evident. The unit's staff members were effectively ‘institutionalised’ … so they could never develop an understanding of the patients’ way of life or devise methods for using community resources to help the patients (Baruch & Treacher, 1978, p. 223).

A different picture emerges when researchers have chosen to look for ‘deviant cases’. Prior (1995), for example, offers a fascinating account of a single patient, ‘Samuel’, who managed to maintain a strong personal identity throughout his 36-year stay in a large mental hospital in Northern Ireland. For the last 10 years of his ‘compulsory’ hospital treatment Samuel received no treatment whatsoever and refused to be discharged on the grounds that he was happy with his life. For him, the hospital functioned as a lodging house while he held down a job as a labourer and played an active role in the local church. He thus “kept his links with the outside world, through family, work and church activities, which indicates that some sections of the community ignored the ‘mental patient’ identity and did not engage in social distancing” (Prior, 1995, p. 650). This account of a single case therefore successfully challenges the notion of the all-pervading power of the total institution, giving us a better understanding of how some people survive their negative effects.

Much more critically, Weinstein (1994) argues the total institution model was out of date soon after Asylums was published, because deinstitutionalization was already underway and mental hospitals were experiencing changes that ameliorated their totalitarian characteristics. That noted, Weinstein (1994) concedes the metaphor's value in sensitising psychiatrists and public officials to the anti-therapeutic consequences of hospital treatment, and, crucially, he reminds us that Goffman constructed an ideal type—a conceptual device to be used as a tool for examining formal organisations. While the type or model was constructed from observations of reality, Goffman (1961) acknowledged that it was not intended to correspond exactly to any single case: what is distinctive about total institutions is that each exhibits to an intense degree many of the same general features. Criticism of the total institution model for corresponding poorly with reality therefore misses the point. However, it has to be said that this ‘naïve’ reading of Goffman is not helped by his cavalier use of definitions (Williams, 1988) and a prose style that encourages readers to forget his qualifiers and caveats rather quickly (Weinstein, 1994). Contemporaries spotted such literary/theoretical sleights of hand, claiming that the model was overdrawn and illusory (Levinson & Gallagher, 1964) and that inmates in different types of total institution define their situations quite differently (Lin, 1968).

So is the total institution model still relevant? Or is now the time to think about introducing a better ideal type that is more recognisable to the people who live or work on today's psychiatric wards? The picture emerging from recent UK health service research literature suggests that this may be the case. Indicators of a much greater degree of permeability on today's wards include:

• The comparatively short length of psychiatric in-patient stays (Thompson et al., 2004). This makes it questionable whether people remain in hospital long enough to fully experience the difficulties described in earlier studies.

• The ease by which drugs and other illicit substances are finding their way onto the wards (Dolan & Kirwan, 2001; Phillips & Johnson, 2003), e.g., via visitors or people's regular dealers (Phillips & Johnson, 2003). This mirrors what has been found in other ostensibly ‘total’ institutions, such as prisons (e.g., Cope, 2000).

• The extension of nurses’ responsibilities beyond the boundaries of the ward's spatial environment and into the outside world. For example, community mental health professionals often call upon the knowledge of hospital nurses when dealing with patients who have been discharged into their care (Deacon, 2003).

In this paper we will examine the nature of the modern, acute psychiatric ward, focusing on its permeability to the outside world. Positive and negative consequences of a high degree of permeability will be identified, as will some of the methods used by social actors (staff and patients) to regulate or resist the level of permeability of their ward.

Previous authors have been aware of this permeability, though to a limited degree. Caudhill, for example, acknowledged the “question of the influence of the culture in general upon events in hospital”, but for practical purposes chose to limit the scope of his analysis (1958, p. 27). Goffman (1961, pp. 111–114) also discussed “permeability” directly, regarding it as one of a number of dimensions of variation among total institutions. He defines permeability as the degree to which the social standards maintained within the institution and the social standards maintained in the surrounding society have influenced one another. However, the emphasis in his account remains firmly on the impermeable aspects of the institution and minimal “role carry-over” from the outside world:

In examining the admission procedures of total institutions, one tends to be struck by the impermeable aspects of the establishment, since the stripping and levelling processes which occur at this time directly cut across the various social distinctions with which the recruits enter (Goffman, 1961, p. 112).

Research undertaken more recently indicates that there is plenty of analytic mileage left in the total institution model (for example, McCorkel, 1998). However, even when the model is assumed, the analysis can nevertheless direct the reader to permeable aspects of the institution. For example, Leyser's (2003) ethnographic study of life in a US mental hospital found male residents accessing masculine power in similar ways to outside populations (e.g., using sexualised talk to turn women into ‘props’). Leyser (2003) concludes that gendered norms found in the larger society were very much part of life in the hospital because residents brought into the institution their experiences of a gendered world, and were exposed to the “outside world” during their stay through media stories, family visits and interactions with staff.

Previous research accounts have thus paid some attention to the relationship between the organisation and its environment, but the issue of permeability has generally been downplayed. Our position, then, is that: (1) previous sociological accounts of psychiatric hospitals in the 1950s and 1960s portray them as comparatively closed to the outside world, which seems to be a fair representation of what they were like at the time; (2) the ‘bracketing’ of permeability as a side issue in those accounts has led to its importance being overlooked; (3) the subsequent trend towards a greater degree of permeability in psychiatric hospitals has made those accounts appear dated in some respects; and (4) this makes it important to turn the analytical spotlight onto the permeable membrane between the hospital and its environment.