Commentary on the origins and early development of the therapeutic landscapes concept

Wil Gesler

17 Apr 2017
Suileven, Sutherland, Scotland; source: Tony Cooper of tonycooperphotography.comSuileven, Sutherland, Scotland; source: Tony Cooper of


A quarter of a century has gone by since I first began to develop the idea of a therapeutic landscape. Since I retired officially from academia a dozen years ago, I have carried out some research with Sarah Curtis and other British colleagues on hospital design for mental health patients using the concept of therapeutic landscape, but I have not kept up with all the literature that uses the concept. In this commentary I reflect on the theoretical origins of ‘therapeutic landscape’, its conceptual development, and my early work using it as a framework for analysing healing places. I also speculate on why the idea ‘works’, or not, for researchers and others. Perhaps this will provide some historical perspective on where the concept came from, how it developed and changed, and where it may be going.

Theoretical origins

The therapeutic landscape concept would not have come about except for an unexpected turn of events. The first time I came across the idea was in 1990, in an anonymous review of a book I was trying to have published called The Cultural Geography of Health Care. I forget the material in the book referenced or the exact words used, but the reviewer said something to the effect of: ‘You could call this idea a therapeutic landscape’.

Around two years earlier, several chapters of the book had been written, based on the application of major themes in cultural geography, then current in the 1980s, to medical geography. A publisher had agreed to take the book. Then, in hindsight, a fortunate thing occurred: the publisher went out of business. I say ‘fortunate’ because I was just beginning to engage with social theories that were seeping into various subdisciplines of geography. I decided to expand the book, adding chapters centred around humanistic and structuralist theories. Another publisher was found, the book went out for review, and it was something in the new material that elicited the reviewer’s comment.

With the book out of the way, I was left to ponder just what a therapeutic landscape might be and what the concept might do. I spent a year and more delving mainly into literature in the social sciences of geography, anthropology, sociology, and psychology, and, to a lesser extent, in the humanities disciplines of history, philosophy, and religion, on what kinds of environments scholars thought would be conducive to health. The result was a paper that summarized my findings and set out the basic concept of a therapeutic landscape and its characteristics (Gesler 1992).

By trade, I was a medical geographer. However, the titles of the book and article give away the fact that my background reading was guided to a significant degree by a long-standing interest in things cultural. This included familiarity with cultural geography, but also a growing fascination with medical anthropology. As it turns out, again fortuitously, in the 1980s and 1990s first cultural geography and then medical geography were undergoing revolutions or ‘turns’ in thinking as they were infiltrated by social theories. Some of these changes I had already picked up on, but now the ideas loomed much larger and began to serve as a theoretical framework for the therapeutic landscape concept.

It is my experience that conversions, religious or otherwise, seldom appear as symbols in the sky and cause instantaneous change; rather, they take root in ground that has already been prepared. In my case, it was becoming increasingly apparent by the late 1980s that the geographic training I had received, so heavily dependent on the quantitative revolution of the 1950s and 1960s, was not adequate to address the questions I thought mattered the most in the geography of health care.

To be sure, research that performed statistical analyses of disease and mortality patterns in maps, measured geographic access to and utilization of health facilities, and developed location/allocation models was very important in the analysis of health care delivery and was useful for planners, but it seldom touched on political, economic, social, or cultural issues, or addressed people’s personal experiences. I could carefully measure the travel times of a group of women to a maternal and child health clinic in an African country, but weren’t there more important issues, such as whether or not mothers had to walk to the clinic from their homes, what they thought of biomedicine, who paid their clinic fees, or how they were treated by nursing staff? These kinds of thoughts came to a head one day when, on a conference field trip, I walked across the Brooklyn Bridge in New York City with a colleague, Robert Stock, who planted a robust seed in my head that was nourished by a potent treatment of political economy.

Following this personal change in thinking, I was receptive to the recent turns in cultural geography and medical geography mentioned above. There were some aspects of cultural geography prior to its social theory turn that were useful in developing the therapeutic landscape concept, namely ‘cultural ecology’ and ‘cultural landscape’. More important, however, was the ‘new’ cultural geography, advocated by Denis Cosgrove and Peter Jackson (1987) and others, that was infused with structuralist and humanist theories. A reformed cultural geography became involved in, among other things, identity politics, working with the marginalized, examining everyday practices, and favouring cultural processes over forms.

It is true that the ‘traditional’ medical geography that was part of my training had a strong element of cultural ecology that suited ideas about therapeutic landscapes. But, again, it took a few brave souls such as Robin Kearns (1993) to break ranks and ask why the subdiscipline wasn’t embracing social theory. I joined the new camp enthusiastically. A schism divided new and old medical geography, not a fatal one (they all meet together at biannual international conferences, for example), although the advocates of the new tended to hive off into a group calling themselves ‘health geographers’. Fault lines were not crystal clear. However, health geographers tended to replace positivist and quantitative philosophies and methods with nonpositivist and qualitative ones; social and political neutrality with relevancy and advocacy; natural sciences and biomedicine with social sciences; and treating population characteristics such as gender, income, and race as variables in models with viewing health and disease as social constructs and examining the meaning of these labels.

In both health geography and cultural geography the meaning of place emerged as a central concern. Health geographers, some believed, were beginning to abandon their geographic roots and intrude into the other social sciences. But newly expressed concerns about social, cultural, political, and economic issues could now be grounded in what was an old, nineteenth-century idea in geography: place. In the new thinking, place was no longer simply a location, but a site that was imbued with experience and meaning. One could almost say that, just as Albert Einstein integrated time with the three spatial dimensions into a space–time continuum, so we could think of place as a fourth dimension, closely tied in with the three dimensions of therapeutic landscapes. With place as a focus, many important questions could be asked. How did cultural processes involved with health care-seeking behaviour work themselves out in specific locales? How were people seeking care affected by the character of a place, and how did these people alter that character?

In my literature search, I was particularly drawn to the theoretical musings and ethnographic studies carried out by medical anthropologists. Arthur Kleinman’s work on medical systems as cultural systems (Kleinman 1978) and his stories about mental illness (Kleinman 1988) were absolutely fascinating. Studies such as Victor Turner’s (1975) work on the symbolic complexes used by the Ndembu of Zambia, J. W. Bastien’s (1985) intricate analysis of the health beliefs and practices of the Qollahuaya Indians of the Bolivian Andes, and Byron Good’s (1977) description of the semantic networks employed by the people of Maragheh in northwestern Iran were a strong influence on formulating the symbolic element of therapeutic landscapes. All of this material helped me to realize how important it was to try and get into the heads of people seeking health care, to understand their behaviour, thoughts, beliefs, and emotions.

Looking back at the medical anthropology literature now, I realize that the main components of the therapeutic landscape concept were already there, sometime before the idea made it into the health geography literature. The content is there as ethnographic studies easily encompass physical, social, and symbolic environments. The method of using multiple layers of analysis is there in ideas such as ‘thick description’ (see below). The attempt to elicit meaning from healing situations was articulated by Kleinman (1978), Good (1994), and others. And ethnographies always exhibit a strong sense of place.

Developing the concept

It was not at all clear at the beginning just how the material that had been collected should be organized. What were the defining elements? At times I had four and even five categories. Then, finally, things settled down into a simple tripartite categorization that was analogous to Julian Huxley’s (1955) division of culture into ‘artifacts’, ‘sociofacts’, and ‘mentifacts’. For the therapeutic landscape concept, these came to be called physical (natural and man-made) environments, social environments, and symbolic environments.

As the concept evolved, other scholars, as well as myself, reflected on the nature of the idea, on what it was, what it was not, and how it could be made more useful. Throughout the process of categorization, I came to realize that there was going to be overlap, and interplay, among the elements. It was never intended that the three environments should be set in stone. Clearly, many specific features of a healing situation could be placed in more than one slot. A garden on the grounds of a hospice, for example, is a natural element, but it can also facilitate social interaction and may symbolize repose for many.

Researchers who took up the idea of therapeutic landscapes were quick to point out that an aspect of a healing site that is therapeutic to one person might not be to someone else. Of course, that is true. In fact, I had already incorporated this idea in some of my work. Thus, for example, I pointed out that beggars and bathers in eighteenth-century Bath, England, one of my original healing places, clearly had different perspectives on the place. This example, and others, also countered the criticism that my research always accentuated the positive.

I worried that the therapeutic landscape idea could be construed as attempting to set out some sort of ideal. This was not my intention. When I asked a group of students in a health geography class to think about the perfect spot to recuperate from an illness, they listed the kinds of places many of us would think of: the beach, a cabin in the woods, at home with mother. I would point out that the therapeutic landscape framework was not a search for these places. Rather, it was an analytic framework, a way of looking at, assessing, those places that existed in reality: the hospitals, clinics, homes, and other places where healing was supposed to be taking place. Application of the framework was intended to find out what was actually going on in these locales in terms of their physical, social, and symbolic environments. Of course, this examination could have as one of its purposes the creation of a more ideal situation for users of health care. And, as time went on, the reality restriction was loosened when scholars began examining fictional places described in literature.

One thing that I discovered in working with colleagues on hospital design was that the three environments had not been given equal billing in previous research. The idea that physical environments could have an effect on health, positive or negative, had been around since at least as far back as the time of Florence Nightingale in the nineteenth century. Perhaps because features of these environments – the amount of light or noise levels, say – could be quantified fairly easily, they loomed larger in the minds of planners and architects. What we found, however, is that, encouraged to talk freely about their likes and dislikes, hospital users, staff, carers, and managers mentioned social and symbolic elements to the same extent as physical elements. The former are harder to quantify, perhaps, but they are certainly just as likely to produce emotional responses.

Applications in practice

With some theoretical notions in place, it was time to try and put them into practice. I decided to begin with three places that, historically, had gained a reputation for healing over a relatively long period of time. The research would be carried out mostly in libraries, with, hopefully, short visits to the places as well. Picking up on a suggestion from a book editor, I decided to start out with a look at Epidauros in Greece, a site where the half-man, half-god Asclepius was reported to perform dream healings from around the fourth century BCE to the sixth century CE. I read everything I could find about the place and then obtained a small grant to visit the site.

It soon became evident that the therapeutic landscape concept was ‘working’ for the ancient Greek healing place. The physical environment combined natural beauty, solitude, and stunning architecture. Within the complex there were strong, meaningful interactions among patients, physicians, and Asclepius, as well as a busy social calendar that included daily rituals and periodic festivals. The symbolic nature of myths about Asclepius, who bridged the human and the divine, the close interweaving of religious and medical beliefs, and the dream healings themselves were obvious.

Reading up on Bath and Lourdes in France, plus visits to these two healing sites, soon followed. Again, it was easy to apply the three therapeutic landscape elements of physical, social, and symbolic environments to narratives about these places. But it wasn’t long before I was chided for limiting the scope of what could be done with the therapeutic landscape idea. The question was asked: ‘Why focus only on well-known historical examples, when what is really important is the places people go to in their everyday lives?’ To myself, I thought two things. First, one has to begin somewhere, so why not with the more obvious and familiar? Second, I was overjoyed that others were picking up on the concept and applying it to a myriad of different places, some well known, but most of them ordinary. I was also criticized for using only Western examples, but this deficiency was at least partially remedied when other research (studies carried out by medical anthropologists in particular) delved into non-Western parts of the world.

Why it works (or does not)

From giving talks on therapeutic landscapes, mainly to academic groups, I learned that the concept did not appear to appeal or be understood by everyone. Audiences in university settings were generally receptive, but of course they were usually self-selected. I did come across some rather strange reactions, however. For example, an economic geographer in a university department dismissed the idea as irrelevant and went on to suggest that the problem of health care delivery could best be solved using rational choice theory. I felt like I was in a time warp, unbridgeable by mutual understanding.

I had always wondered what nonacademics might make of the concept. At a signing of my book titled Healing Places in a bookshop in a residential community, I came away with the feeling that people were looking for something along the lines of those guides put out by self-help gurus, hoping that the book would tell them where they could find ideal healing sites. A member of the audience somehow managed to steer the question period following my presentation into a kind of New Age discourse on peace, harmony, and love. It was a good example, I think, of the disconnection between the goals of an academic endeavour and the desires of those living in the real world.

Still, over time, despite some misconceptions, the therapeutic landscape concept seems to have taken on a life of its own. I credit a great deal of the growth of the concept to the immense effort Allison Williams put into broadening the idea, organizing conference sessions, and editing two volumes on the topic (see her commentary in this journal issue). Currently, the concept appears to be in rude health. Why has it been successful? I can only speculate. Perhaps it is a geographic metaphor that simply resonates with many people. In addition, the idea appeared at a time when there was a reaction in the social sciences of health against positivist, reductionist approaches to health and well-being. I was pleased to find, for example, that an audience of would-be physicians taking a course in social medicine was generally receptive to the idea. Perhaps medical anthropologists find the idea useful because there are many aspects that are familiar to them. I also like to think that, because it was developed from an eclectic variety of theoretical strains, it was hardy, thrived and, in turn, branched out into a further variety of interesting and useful ideas and practices.

Perhaps the therapeutic landscape approach provides valuable insights into health care-seeking behaviour due to the research methods I and others have used. To use language that anthropologists will recognize, researchers sought to replace the thin description of objective science (Dragnet’s ‘Just the facts, ma’am’) with the thicker description of in-depth analysis (Geertz 1973). Admittedly, my investigations of famous healing sites were by no means full-on ethnographic studies, but the therapeutic landscape framework enabled me to describe those places with different layers of overlapping meanings. In our work on mental hospitals, we attempted to achieve thicker descriptions of Clifford Geertz’s ‘social discourse’ by conducting our encounters with patients, staff, and carers as interactive discussions or conversations rather than interviews or focus groups and letting our respondents speak freely with almost no guidance. Taking still another idea from Geertz, we, on occasion, were able to elicit the ‘unapparent import’ of things. For example, we realized that, counter-intuitively and against the grain of medical and public health orthodoxy, smoking among mental patients and staff had potential benefits in terms of encouraging social interactions that had the potential to heal (Wood et al. 2013).

An example from our work on the design of a newly built mental health inpatient unit illustrates a very modest attempt at thick description (Curtis et al. 2007). Conversations with respondents revealed a set of design features and hospital practices related to risk avoidance that could be located in one or more of the three therapeutic landscape environments. These included the panopticon-like location of nursing stations, in a position to monitor three wards that radiated out from the stations; windows in patients’ rooms fitted with blinds that could be raised by staff for enhanced observation; smooth, hard-to-turn door handles to prevent ligature (committing suicide); and locked windows in bedrooms to dissuade patients from absconding, which made controlling room temperatures very difficult. All these items interacted in a web of meanings that revealed conflicting agendas held by interested parties (patients, staff, hospital management, government health ministers, the media, and the public): obtrusive observation versus patient privacy; restrictive, risk-averse rules versus patient autonomy; and keeping patients and staff at arm’s length rather than in potentially healing social interactions.

I have been intrigued by some recent work that attempts to explain why or how therapeutic landscapes work from a perspective that derives from theories of the mind. Using ideas from psychoanalysis and psychotherapy, David Conradson (2005) has developed the concept of the ‘relational self’. The therapeutic landscape experience, he suggests, centres on the relationships people develop with others and with various environments during the healing process. Emma Rose (2012) takes this line of inquiry further, talking about how infants learn to internalize positive experiences of environments from caregivers that can be beneficial when they are revived in later life.

About the author

Wil Gesler is a health geographer. He received his PhD in geography from the University of North Carolina at Chapel Hill in 1978. From 1978 to 1981 he was a member of the Department of Geography at Rutgers University and then spent a year at the University of Sierra Leone, Freetown, as a Fulbright scholar. From 1981 until his retirement in 2003 he was a member of the Department of Geography at UNC-Chapel Hill. His academic interests include cultural geography, spatial analysis, and geographic aspects of literature.


Bastien, Joseph. 1985. ‘Qollahuaya-Andean Body Concepts: A Topographic-Hydraulic Model of Physiology’. American Anthropologist 87: 595–611.
Conradson, David. 2005. ‘Landscape, Care and the Relational Self: Therapeutic Encounters in Rural England’. Health & Place 11: 337–48.
Cosgrove, Denis, and Peter Jackson. 1987. ‘New Directions in Cultural Geography’. Area 19: 95–101.
Curtis, Sarah, Wilbert Gesler, Kathy Fabian, Susan Francis, and Stefan Priebe. 2007. ‘Therapeutic Landscapes in Hospital Design: A Qualitative Assessment by Staff and Service Users of the Design of a New Mental Health Inpatient Unit’. Environment and Planning C: Government and Policy 25: 591–610.
Geertz, Clifford. 1973. ‘Thick Description: Toward an Interpretative Theory of Culture’. In The Interpretation of Cultures: Selected Essays, 3–30. New York: Basic Books.
Gesler, Wilbert. 1992. ‘Therapeutic Landscapes: Medical Geographic Research in Light of the New Cultural Geography’. Social Science & Medicine 34, no. 7: 735–46.
Good, Byron. 1977. ‘The Heart of What’s the Matter: The Semantics of Illness in Iran’. Culture, Medicine, and Psychiatry 1: 25–58.
Good, Byron. 1994. Medicine, Rationality, and Experience: An Anthropological Perspective. Cambridge: Cambridge University Press.
Huxley, Julian S. 1955. ‘Guest Editorial: Evolution, Cultural and Biological’. Yearbook of Anthropology, 2–25.
Kearns, Robin. 1993. ‘Place and Health: Towards a Reformed Medical Geography’. The Professional Geographer 45: 139–47.
Kleinman, Arthur. 1978. ‘Concepts and a Model of Comparison of Medical Systems as Cultural Systems’. Social Science & Medicine 12: 85–93.
Kleinman, Arthur. 1988. The Illness Narratives: Suffering, Healing and the Human Condition. New York: Basic Books.
Rose, Emma. 2012. ‘Encountering Place: A Psychoanalytic Approach for Understanding How Therapeutic Landscapes Benefit Health and Wellbeing’. Health & Place 18: 1381–87.
Turner, Victor. 1975. ‘Symbolic Studies’. Annual Review of Anthropology 4: 145–61.
Wood, Victoria, Sarah Curtis, Wilbert Gesler, Ian Spencer, Helen Close, James Mason, and Joe Reilly. 2013. ‘Spaces for Smoking in a Psychiatric Hospital: Social Capital, Resistance to Control, and Significance for “Therapeutic Landscapes”’. Social Science & Medicine 97: 104–111.