Readings on the Psychology of Place: Selected Works of David Canter 1032321490, 9781032321493

Citation preview

Readings on the Psychology of Place

In the World Library of Psychologists series, international experts present career-long collections of what they judge to be their most interesting publications – extracts from books, key articles, research findings and practical and theoretical contributions. In this fascinating volume, Professor David Canter reflects on a career that has earned him an international reputation as one of the U.K.’s most eminent applied social psychologists and a pioneer in the field of environmental psychology, through a selection of papers that illustrate one of the foundational themes of his research career: the psychology of place. Split into four parts, each with a new introduction written by the author, the book provides insights into theories, methods and applications of place psychology. Covering a range of publications from early research in the 1960s up to recent explorations, this volume provides the unfolding research that elaborates this seminal theory, ofering rich perspectives on how places gain their significance and meaning. Featuring specially written commentary by the author contextualizing the selections and providing an intimate overview of his career, this collection of key publications ofers a unique and compelling insight into decades of ground-breaking work, making it an essential resource for all those engaged or interested in the study of places. David Canter is one of the U.K.’s most eminent applied social psychologists, being one of the few to be appointed as Honorary Fellow of the British Psychological Society and having been elected as a Fellow of the Academy of Social Sciences, the American Psychological Association and the Royal Society of Medicine. Although he is internationally known for his development of the discipline of Investigative Psychology, bringing scientific precision to ‘ofender profiling’, his earlier and continuing work centres on the development of Architectural/environmental psychology, establishing the well-respected Journal of Environmental Psychology in 1980 and the International Association for People-Environment Studies (IAPS). He has worked as a management consultant to major U.K. companies on risk reduction, amalgamations and briefing for new building complexes and has given advice to government inquiries into disasters. He wrote and presented a six-part documentary series, Mapping Murder, which is also published as a widely read book. He is Emeritus Professor at the University of Liverpool, U.K.

World Library of Psychologists

The World Library of Psychologists series celebrates the important contributions to psychology made by leading experts in their individual fields of study. Each scholar has compiled a career-long collection of what they consider to be their finest pieces: extracts from books, journals, articles, major theoretical and practical contributions, and salient research findings. For the first time ever the work of each contributor is presented in a single volume so readers can follow the themes and progress of their work and identify the contributions made to, and the development of, the fields themselves. Each book in the series features a specially written introduction by the contributor giving an overview of their career, contextualizing their selection within the development of the field, and showing how their thinking developed over time. Studies of Thinking Selected Works of Kenneth Gilhooly Kenneth J. Gilhooly Cognitive Development and the Ageing Process Selected Works of Patrick Rabbitt Patrick Rabbitt Individuals as Producers of Their Own Development The Dynamics of Person-Context Coactions Richard M. Lerner Parenting Selected Writings of Marc H. Bornstein Marc H. Bornstein Readings on the Psychology of Place Selected Works of David Canter David Canter For more information about this series, please visit: www.routledge.com/World-Library-ofPsychologists/book-series/WLP

Readings on the Psychology of Place Selected Works of David Canter

David Canter

Cover image: Courtesy of David Canter First published 2024 by Routledge 4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2024 David Canter The right of David Canter to be identified as author of this work has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-032-32140-0 (hbk) ISBN: 978-1-032-32149-3 (pbk) ISBN: 978-1-003-31305-2 (ebk) DOI: 10.4324/9781003313052 Typeset in Bembo by Apex CoVantage, LLC

Contents

Acknowledgments Preface

viii x

PART 1

Origins of the theory of place 1 Architectural psychology

1 5

DAVID CANTER

2 Ofce size: An example of psychological research in architecture

10

DAVID CANTER

3 Need for a theory of function in architecture

26

DAVID CANTER

4 Should we treat building users as subjects or objects?

35

DAVID CANTER

5 Judgements of people and their rooms

50

D. CANTER, S. WEST AND R. WOOLS

6 Distance estimation in cities

56

DAVID CANTER AND STEPHEN K. TAGG

7 The psychology of place DAVID CANTER

70

vi

Contents

PART 2

Elaborating the theory of place 8 The purposive evaluation of places: A facet approach

77 79

DAVID CANTER

9 Putting situations in their place

105

DAVID CANTER

10 Intention, meaning and structure: Social action in its physical context

132

DAVID CANTER

11 Action and place: An existential dialectic

145

DAVID CANTER

PART 3

Methodology – Studying place experience

159

12 A non-reactive study of room usage in modern Japanese apartments

161

DAVID CANTER AND KYUNG HOI LEE

13 Picture or place? A multiple sorting study of landscape

172

M. J. SCOTT AND DAVID CANTER

14 Revealing the conceptual systems of places

196

DAVID CANTER

PART 4

Applications – The theory of place in action

219

15 Intentionality and fatality during the King’s Cross underground fire

221

IAN DONALD AND DAVID CANTER

16 The environmental range of serial rapists

238

DAVID CANTER AND PAUL LARKIN

17 Why do we leave it so late? Response to environmental threat and the rules of place DAVID CANTER

249

Contents 18 Creating places without designers

vii 266

DAVID CANTER

Bibliography – Selection of David Canter’s Architectural/Environmental Psychology Publications Index

276 280

Acknowledgments

Chapter 1. Canter, D. Architectural Psychology in D. Marchand, E. Pol, and K. Weiss (eds) 100 Key Concepts in Environmental Psychology. Paris: Dunod Chapter 2. Canter, D. (1968). Ofce size: An example of psychological research in architecture. Architects’ Journal, April 24, 881–888. Chapter 3. Canter, D. (1970). Need for a theory of function in architecture. Architects’ Journal, February 4, 299–302. Chapter 4. Canter, D. (1969). Should we treat building users as subjects or objects? In D. Canter (Eds.), Architectural psychology (pp.  11–17). London: RIBA Publications. Chapter 5. Canter, D., West, S. and Wools, R. (1974) Judgements of People and Their Rooms British Journal of Social and Clinical Psychology 13, pp. u3–u8. Chapter 6. Canter, D., and Tagg, S. (1975). Distance estimation in cities. Environment and Behaviour, 7(1), 59–80. Chapter 7. Canter, D. (1977). The psychology of place. London: Architectural Press. Chapter 8. Canter, D. (1983). The purposive evaluation of places: A facet approach. Environment and Behaviour, 15(6), 659–698. Chapter 9. Canter, D. (1986). Putting situations in their place. In A. Furnham (Ed.), Social behaviour in context (pp. 208–239). Boston: Allyn & Bacon. Chapter 10. Canter, D. (1985). Intention, meaning and structure: Social action in its physical context. In G. P. Ginsburg, Marylin Brenner & Mario von Cranach (eds.), Discovery Strategies in the Psychology of Action. Academic Press. Pp. 35–171. Chapter 11. Canter, D. (1988). Action and place: An existential dialectic. In D. Canter, M. Krampen, and D. Stea (Eds.), Environmental perspectives (pp.  1–17). Aldershot, England: Avebury. Chapter 12. Canter, D., and Lee, K.H. (1974) A non-reactive study of room usage in modern Japanese apartments. In: D. Canter and T. Lee (eds.) Psychology and the Built Environment. Architectural Press, pp. 48–55. Chapter 13. Canter, D.V., and Scott, M.J. (1997) Picture Or Place? A Multiple Sorting Study Of Landscape Journal Of Environmental Psychology 17, pp. 263–281. Chapter 14. Canter, D. (2016) Revealing the Conceptual Systems of Places In R. Gifford (Ed.) Research Methods for Environmental Psychology, pp  137–160 https://doi.org/10.1002/9781119162124.ch8.

Acknowledgments

ix

Chapter 15. Donald, I., and Canter, D. (1992) “Intentionality and Fatality during the King’s Cross Underground Fire”. European Journal of Social Psychology Volume 22, 203–2 18. Chapter 16. Canter, D., and Larkin, P. (1993) The Environmental Range of Serial Rapists The Journal of Environmental Psychology, 13 63–39. Chapter 17. Canter, D. (2013) Why do we leave it so late? Response to Environmental Threat and the Rules of Place, J Earth Sci Clim Change, 5:1 http://dx.doi. org/10.4172/2157-7617.1000169. Chapter 18. Donald, I., and Canter, D. (1992) Intentionality and Fatality During the King’s Cross Underground Fire, European Journal of Social Psychology, 22 (3) 203–218.

Preface

By the 1980s, a small but dedicated community of academics and some practitioners were studying psychological aspects of people’s interactions with their surroundings. As explored in my article (reproduced here) describing architectural psychology, the origins of these studies had been the practical concerns about designing buildings so that they were more efective for their users. But by the 1980s, the interest in human experience and use of their physical contexts had broadened. Architectural psychology had been subsumed under the wide-ranging umbrella of environmental psychology. This latter area is evolving to focus on the important issues of human response to climate change and how people may be persuaded to modify their environmentally related activities. The emerging focus on this ‘green environmental psychology’ led to a reduced interest in the earlier research, carried out during the last three decades of the twentieth century. Those studies, and contributions to many aspects of design, were overshadowed by the pressing need to connect with behavioural aspects of the many environmental challenges. However, in recent years, there has been a reawakening of interest in those earlier studies. This is demonstrated, for example, by the book by Lily Bernheimer (2017) and, most recently, the book by my former student and later colleague, Ian Donald (2022). That reawakening has encouraged a reappraisal of the significance of the earlier research. It is contributing to understanding how those psychological processes underlying human transactions with the physical environment are at the heart of any climate-relevant behaviour. One particular aspect of the early research has not been ignored. This is the study of ‘place’ as an aspect of experience, as distinct from a mere space or location. This was first elaborated in my book The Psychology of Place (Canter, 1977), the introductory summary of which is reproduced here. There has since been a great outpouring of studies and publications around the theory and applications of this way of thinking about places. It is, therefore, appropriate to bring together in the present volume a set of earlier publications that elaborate the details of this seminal theory. These earlier works are complemented by more recent studies which build directly on that initial research. Most of these publications emerged in a pre-digital age. Consequently, they are not as readily available as subsequent publications. Furthermore, many of them were published in books of readings or relatively obscure journals because that was where

Preface

xi

the interest was in the topics. This also make them more difcult to access. Therefore, bringing them together here provides a valuable resource for those who wish to understand the origins and engage directly with the many strands of The Psychology of Place. References Bernheimer, L. (2017) The Shaping of Us London: Robinson Canter, D. (1977) The Psychology of Place London: Architectural press (Also available on Kindle) Donald, I. (2022) Environmental and Architectural Psychology: the basics New York: Routledge

Part 1

Origins of the theory of place

At the age of 21, I embarked on a PhD with the Pilkington Research Unit in the Department of Building Science at the University of Liverpool (although my PhD was registered and ofcially supervised in the Department of Psychology). The topic of the PhD, awarded in 1969, was ‘The Psychological Implications of Ofce Size’.1 This was formulated within a simple-minded experimental psychology tradition. The assumption was that the larger the ofce, the more distractions, so the poorer the performance of ofce workers in those rooms. This was a ‘determinist’ assumption about the impact of the environment on behaviour. As illustrated in the 1968 article, reproduced here from the Architects’ Journal, the results challenged that assumption. In the tradition of field experiments, I got people to complete clerical tests in diferent-sized ofces. These varied from rooms with only a few people in to those with over 100 in. Some of the ofce workers were tested in the ofce in which they normally worked. Others were tested in ofces of a diferent size from their usual room. To my confusion, I found no change on the test scores for people tested in different ofces but a clear indication that the larger the ofce in which they normally worked, the poorer their scores. After much head scratching, I realized that I was identifying something about the sorts of people who worked in ofces of diferent sizes. I was not measuring the direct impact of the environment on their performance. This turned my determinist assumptions on their head. For one reason or another, people of difering clerical abilities seemed to be drifting towards ofces that housed diferent numbers of workers. In the early 1960s, when my data were collected, there were high levels of employment. People could choose where they wanted to work. Therefore, one assumption could be that people who liked working in large open ofces would accept jobs there. Those who did not like that possibility would not. There were also likely to be status issues about who ended up in which ofces. These considerations led the awareness that people make use of buildings in relation to their understanding of the possibilities these ofer. This understanding relies to some extent on the users’, or potential users’, interpretation of any particular physical context – in a word, the ‘meaning’ of that environment. Remarkably (I think, for the young researcher I was then), I did indicate in this first paper, which was presented as an illustration of the possibilities for research in architecture, that there was a need ‘to build theoretical models which explain, in DOI: 10.4324/9781003313052-1

2

Origins of the theory of place

detail, the relationships between the disciplines of architecture and that of psychology’ (page 881). The subsequent quarter of a century of my research struggled with aspects of this possibility. The starting point for developing ‘theoretical models’ was to explore what architecture is for. This may initially seem obvious, but there is a crucial distinction between civil engineering and architecture. I did develop the consideration of this distinction a few years later (Canter, 1977), but the initial idea was to explore the functions of architecture. The need for an account of that exploration was also published in the Architects’ Journal and reprinted in the present volume. It draws attention to three ways in which buildings function: • as a filter – keeping nasty things (like noise) out and letting good things (such as daylight) in; • as a social facilitator – providing the spatial opportunities to house various activities and the potential relationship between those activities; and • as a signifier – providing possibilities for the interpretation of the meaning of buildings and the spaces they house. This latter point was rather contentious at the time because there was still the myth enshrined in the slogan ‘form follows function’, somehow indicating that the design of a building was merely a reflection of the uses to which it was to be put. However, once it was acknowledged that one function of a building is to indicate what those purposes are with all the cultural and status overtones that implies, then the idea that a building was just an engineering project, ‘a machine for living in’ as Le Corbusier famously announced, is no longer sustainable. The implication of this perspective is summarized in that 1970 paper as the following: building type categorisation is often inadequate. For example, modern ofce buildings are . . . the same inside and out. Yet what goes on in them . . . are quite diferent. The head ofce of an organisation making a specialised product has quite a diferent hierarchy from, say, an insurance company. A cellular block of ofces might be right for one but not for the other. (page 301) I like to think that this perspective fed into the debates emerging in architecture in the early 1970s, about approaches to building design, and the need to recognize the meanings that buildings carried. This was articulated by the architect and architectural commentator Charles Jenks (1977). He expropriated a concept from discussions in English literature of ‘post-modernism’. Most importantly, he argued that post-modern architecture has a ‘language’. What Jenks did not do was to take the next step and recognize that languages have listeners who may or may not understand what is being said. My discovery of

Origins of the theory of place

3

what people brought to their interactions with buildings was opening up as much a perspective that saw them as agents, not passive users. The rather neat title of the paper (reproduced here) that I gave at the legendary 1969 Dalandhui conference (Canter, 1970), ‘Should we treat building users as subjects of objects?’, encapsulated a debate which resonates still today. One intriguing study in my paper, which never found a more conventional journal outlet, was the consideration of where students at the Architectural Association, London, School of Architecture chose to sit, in a seminar series I gave, when I arranged the layout of the room in diferent ways. Quite remarkably, they sat at the front when it was a semi-circular layout but at the back when laid out in straight rows. In those late 1960s days, the AA (as it was known) was somewhat radical. I still remember that the babies taken to my seminar series were not as great a distraction as the dogs brought along. In this context, the reading of the nature of the particular seminar was clearly derived from the layout of the room with the consequent decision of whether to attend the event and, if so, where to sit. It was from this consideration of how people read a room that it occurred to me that the context in which a person is seen could carry meanings about that person. The study for which I supervised an architect, Roger Wools (reprinted here), did indeed demonstrate the power of the context on judgements of people. This idea was elaborated many years later by Sam Gosling (2008) in his popular book. This idea has subsequently been reflected in various television game shows in which people have to guess who the owner is of homes presented to them. Sam’s attempt to participate in such guesses were often not very successful, demonstrating the complexity of these relationships. A follow-up study I did with videos instead of still photographs gave such complex results that it was yet another interesting study that never got published. By the mid-1970s, I was seeking to broaden the reach of my work. The previous focus on buildings and rooms seemed to me to be an unnecessary limitation. A remarkable one-year Leverhulme fellowship to Japan had opened my eyes to the relevance of psychological considerations to larger environmental contexts. The ways in which we build up our understanding of cities had intrigued me as I started to learn how to navigate around Tokyo. I was, therefore, curious about the accuracy with which people generate ideas of where things are, especially how far away they are. From my background in laboratory studies of perception, it occurred to me that in a city where locations cannot be immediately seen, there must be some internal representation. How did this relate to the representation of maps? A very straightforward study suggested itself, examining the estimates people made of distances (reproduced here). By this stage, I had started to use a form of statistical analysis that has proven powerful ever since. This is the representation of relationships as distances in a notional space. However, in the case of distance estimation, it was possible to represent average distant estimates as distances in a notional space. That representation could be compared with other representations. In the complex city layout of Tokyo, there is a public transport system that is represented as a circle, but on an actual map, it is much more elliptic. It was, therefore, rather interesting to see that the spatial representation

4

Origins of the theory of place

of distance estimates was much closer to a circle that an ellipse. This really paved the way for recognizing that places are as much mental representations as geographical ones. Writing my widely cited book, The Psychology of Place, was a journey for me, exploring the psychological implications of places. The introductory summary of that book is reproduced here. The widely quoted and elaborated framework developed in that book from a variety of research sources is that places are aspects of experience that combine physical, conceptual and action facets. These three are always present, helping to indicate what a place is. The implications of this triumvirate have been developed through theories and practice. Note 1 Available in the University of Liverpool Library.

References Canter, D.V. (1970) (Ed) Architectural Psychology London: RIBA publications Canter, D. (1977) “Why architecture is necessary.” RIBA Journal, 84, June 1977: 261–262. Gosling, S. (2008) Snoop: what your stuf says about you. London: Profile Books Jenks, C. (1977). The Language of Post-Modern Architecture. London: Academy Editions.

1

Architectural psychology David Canter

Architectural psychology (known, unsurprisingly, in German as Architekturpsychologie) is an area of social research and practice that considers the implications for human experience, thoughts and actions of interactions with the built and natural environment. Its initial heyday was in the late 1960s and early 1970s with a seminal conference in Scotland in 1969 (Canter, 1970) and another in Australia in 1974 (Canter, 1974) and a book of original research the same year (Canter and Lee, 1974). The term fell into disuse but is making something of a comeback in the 2020s, with a number of design and research groups using the term to describe their activities. There was a psychology and architecture conference at the University of Texas in 20161 and a conference using the term was planned in Germany for 2021. The area of study emerged as the Second World War came to an end. At that time, there was a growing democratization of politics combined with awareness of changes in society. This meant people in authority could no longer decide for the population at large. The growth of the social sciences also reflected a move towards a more systematic understanding of individual and social behaviour of relevance to many areas of decision making. Those in power could no longer be relied on to understand how people lived. In architecture, this was mirrored by a move towards a more scientific basis for design decision making, reflected in the U.K. in the 1951 Festival of Britain which foregrounded the contribution of science and engineering to modern society, including many aspects of building design. Across Europe, there were government-funded projects, starting in the 1950s, to provide guidelines for the massive building programme after the destruction of the Second World War. In the U.K., there were initially studies for housing (Morris, 1961), then ofces, shops and railway premises, then schools and hospitals. These studies, typically conducted by civil servants and sociologists, tended to focus on appropriate levels of spatial provision for diferent activities. Other early studies were carried out mainly by physicists and engineers. They examined aspects of thermal comfort, noise annoyance (e.g., Grifths and Langdon, 1968, who were actually experimental psychologists) and the implications of lighting levels, especially daylight penetration into homes and ofces. Much of this work related directly to the impact of the physical environment on productivity (Oborne and Gruneberg, 1983). DOI: 10.4324/9781003313052-2

6

David Canter

The scientific trend in architecture gave rise to research units, such as that funded by the plate glass manufacturer, Pilkington, who set up the Pilkington Research Unit, initially studying daylight penetration, then ofce design (Manning, 1965) and, subsequently, schools. This then developed into the multidisciplinary Building Performance Research Unit (Markus et al., 1972). A Europe-wide research association grew out of these studies, the International Association for People Environment Studies (IAPS), with the rather more elegant Australian acronym PAPER (People and the Physical Environment Research). The origins in the U.S. in the late 1950s and early 1960s were requests from architects to psychiatrists and psychologists for guidance on the design of buildings which housed people that the designers considered to be very diferent from themselves, notably psychiatric patients and children. An early account by the psychiatrist Osmond (1957) introduced the then rather novel idea that the function of a psychiatric ward should be the basis for design. This was the start of a stream of research that laid the foundations for architectural psychology in the U.S. This gave rise to the Environmental Design Research Association (EDRA) which is still active today. These developments were encouraged by three early publications from the U.S. which had paved the way for the consideration of the spatial aspects of human activities and experience with architectural implications, Lynch’s (1960) The Image of the City, Sommer’s (1969) Personal Space: The Behavioral Basis of Design and Rapoport’s (1969) House Form and Culture. The area of study broadened out to cover aspects of landscape experience and the meaning of building forms. Consequently, although the term ‘architectural psychology’ was widely used, many of those involved were not psychologists and the areas of study went far beyond those aspects of the environment over which architects have control. This has included somewhat limited studies of the psychology of architects and architectural decision making. Not long after the designation of the existence of architectural psychology through conferences in the U.S. and the U.K., the label environmental psychology was introduced in the U.S. with the establishment in New York of an environmental psychology doctoral programme. Although much of the early research under this umbrella dealt directly with aspects of buildings, the promotional power of U.S. researchers led to the term nudging the label architectural psychology out of common use. With the advent of the ‘environmental movement’ and the growing awareness of a climate crisis, many social scientists became involved in what might be called ‘green environmental psychology’. In many cases, this is an aspect of social psychology dealing with attitudes to climate change and related actions such as recycling and being more ‘environmentally friendly’. With that broadening of the meaning of environmental psychology, architectural psychology became identified as a particular area of the wider field. This does add some confusion because there are important epistemological diferences between green environmental psychology and architectural psychology. From its earliest days, there were theoretical explorations of the ways in which people interact with their surroundings. This is often confusingly couched in questions about the impact of buildings on human actions and experiences. These

Architectural psychology

7

questions are based on the assumption of what is known as ‘architectural determinism’. This is the belief that buildings directly shape what people think, feel and do. A commonly cited quotation from Winston Churchill in his discussion of the rebuilding of the Houses of Parliament after being damaged during the Second World War captures the complexity of even an elementary view of the influence of architecture: ‘We shape our buildings, and after that, they shape us’. The interactive nature illustrated by Churchill point to the limits of buildings’ determining human actions. Many studies show that, except at the extremes of environmental conditions, people are remarkably flexible in their response to the physical environment. Therefore, more complex theories have emerged that explore what people bring to their use and experience of buildings as much as what influence buildings have on them. A central theory to emerge as part of the move away from a strongly determinist perspective draws on the concept of ‘place’ (Canter, 1977). This is proposed as an aspect of experience in which physical, social, emotional and behavioural facets are all combined. The central premise is that all actions and experiences are spatial and all spaces carry implications for actions and experiences. In its most extreme form, this is embedded in a phenomenological perspective that accepts ‘being-in-place’ as a wholistic aspect of living (recently reviewed by one of the leading thinkers of this approach, Seamon, 2018). This has its origins in the musings of the French ‘philosopher’ Bachelard (1958). Although, intriguingly, many geographers seem to have gotten bored with map-making, and this phenomenological viewpoint now typically dominates areas of social/human geography (e.g., Malpas, 1999). One particular strand of theory building has been the exploration of the meaning of places. This has taken on the form of elaborating how and what meanings buildings stimulate. This connected with an emerging interest in semiotics and was given impetus by Jencks’s declaration of the existence of a ‘post-modern architecture’ (Jencks, 1977). As long as the dominant style in architecture followed the (rather ambiguous) slogan ascribed to Le Corbusier of ‘form follows function’, the idea that buildings carried meaning was something of an anathema. Once that idea was undermined, there was room for consideration of what the significance of physical forms might be. Early work brought together by architects Broadbent (1980) and Rapoport (1982) provided a variety of perspectives in how buildings generate meanings. The meaning of places is still an active area of scholarship (Castello, 2010). Research into architectural psychology takes on many diferent forms (Groat and Wang, 2002). In general, it eschews laboratory studies because they provide such a distinct context that all that can be studied there is the laboratory experience itself. The research design, regarded as the golden standard in experimental research, of the double-blind controlled trial makes little sense when studying people’s experience of buildings. They always have a related reason for being there, a role in the building, which cannot be readily simulated for the purpose of research. The practical and ethical challenges of randomly assigning people to diferent physical environment also limits the use of this research design. However, the ease of showing people pictures of places and asking them to react to those images has nonetheless provided a trend in publications which has been

8

David Canter

enhanced with the facility provided by digital media, notably videos and virtual reality representations. Surveys and various forms of interviews, which relate directly to architectural experiences, have tended to be the favoured form of research. Within the phenomenological tradition, these interviews have usually been open-ended explorations of direct experience, often bordering on autobiography or descriptive journalism. The influence of architectural psychology is difcult to pinpoint, in part because of the amorphous nature of architecture and the many diferent stakeholders, such as owners, planning authorities, managers and (rarely) potential users, who influence architectural decision making. However, the ways in which architectural education and the fashion for grand architecture has changed over the past half-century must, to some extent, have been influenced by the debates in architectural psychology. These have included the recognition that people bring their own perspective and aspirations to the use of any buildings and that the idea of ‘function’ in architecture carries a complex mixture of psychological implications, an issue discussed directly in the early days of architectural psychology (Canter, 1970). The involvement of potential users in the design of buildings, with all the benefits that it brings, can also be seen as influenced by concepts and research drawn from architectural psychology. Note 1 www.youtube.com/watch?v=eCYApmWq__k

References Bachelard, G. (1958) The Poetics of Space London: Penguin Books Broadbent, G. (with Richard Bunt and Charles Jencks), (1980) Signs, Symbols and Architecture London: John Wiley & Sons Canter, D. (1970) The Need doe a Theory of Function in Architecture Architects’ Journal, 151(5) pp. 299–302 Canter, D. (1970) (Ed.) Architectural Psychology: proceedings of the Dalandhui conference 1969 London: RIBA Publications. Canter, D. (1974) (Ed.) A Short Course in Architectural Psychology Sydney: Architectural Psychology Research Unit, University of Sydney Canter, D. and Lee, T. (1974) (eds.) Psychology and the Built Environment. London: The Architectural Press Canter, D. (1977) The Psychology of Place London: Architectural Press Castello, L. (2010) Rethinking the Meaning of Place London: Routledge Grifths, I.D. and Langdon F.J. (1968), Subjective response to road trafc noise, Journal of Sound and Vibration, 8 (91), pp. 16–32, Groat, L. and Wang, D. (2002) Architectural Research Methods Chichester: John Wiley & Sons Jencks, J. (1977) The Language of Post-Modern Architecture (1977) London: Rizzoli Morris, P. (1961) ‘Homes for Today and Tomorrow’ report by Sir Parker Morris for the Ministry of Housing and Local Government, 1961. BADDA2543. Museum of Domestic Design & Architecture, Middlesex University. Accessed November 03, 2020. https://moda. mdx.ac.uk/object/badda2543/

Architectural psychology

9

Lynch, K. (1960) The Image of the City Cambridge: MIT Press Malpas, J. (1999). Place and Experience: A Philosophical Topography Cambridge: Cambridge University Press. Markus, T.A., Whyman, P., Morgan. J., Whitton, D., Maver, T., Canter, D. and Fleming, J. (1972) Building Performance, Building Performance Research Unit, London: Applied Science Publishers Manning, P. (1965) Ofce Design: A Study of Environment. Liverpool: Department of Building Science, University of Liverpool Oborne, D.J. and Gruneberg, M.M. (1983) The Physical Environment at Work Chichester: John Wiley & Sons Osmond, H. (1957) Function as the basis of psychiatric ward design, Psychiatric Services, 8 (4), pp. 23–27 ‘Homes for Today and Tomorrow’ report by Sir Parker Morris for the Ministry of Housing and Local Government, 1961. BADDA2543. Museum of Domestic Design & Architecture, Middlesex University. Accessed November 03, 2020. https://moda.mdx.ac.uk/ object/badda2543/ Rapoport, A. (1969) House Form and Culture Englewood Clifs: Prentice Hall Rapoport, A. (1982) The Meaning if the Built Environment: a nonverbal communication approach Beverley Hills: Sage Sommer’ (1969) Personal Space: the behavioral basis of design Englewood Clifs: Prentice Hall J. M. WALDRAM, Studies in interior lighting. Trans. ilium. Engng Soc. Lond. 19, 95 (1954). 3. R.G. HOPKiNSON and J. LONGMORE, The permanent supplementary artificial lighting of interiors. Trans. ilium. Engng. Soc. Lond. 24, 121 (1959). Seamon, D. (2018) Life Takes Place: Phenomenology, Lifeworlds and Place Making New York: Routledge

2

Ofce size An example of psychological research in architecture David Canter

Aim of this article Architects are dealing with qualities and quantities which are too subtle to be immediately influenced by one set of research findings; thus, the scientific study of the psychological consequences of architectural decisions will not produce results with an immediate impact on the quality of architectural design. The psychological implications of any design decision will not be revealed overnight by any one experiment, however crucial, and there are still so few psychologists working in this field that a large volume of results is unlikely to be produced at present. If psychology ever afects the design of buildings, it will do so imperceptibly over a long time as the body of information relevant to architectural problems grows and is interpreted. But it will do so only if architects become more aware of the nature of psychological research, for only then will the relevance of research findings become apparent. There are two ways of showing this relevance. One way is to build theoretical models which explain, in detail, the relationships between the discipline of architecture and that of psychology. I think this is essential in the long term. The other way is to make architects aware of the research process and of the problems of carrying out research in architecture. The best way of doing this is to describe a study of a psychological problem that is of concern to many architects. Another need to be met before psychological research will start having an impact on design decisions is development of research strategies which will produce meaningful results for the designer as well as the psychologist. The aim of this article is thus threefold: to show architects the way in which a relevant psychological research project has developed, to present a research strategy which is thought fruitful in the psychological study of architectural problems and to present results which are felt to be of some importance to architects faced with the problem of designing large ofces. However, it cannot be overemphasized at this stage that the actual results are only tentative; much more research is needed before they are firmly substantiated. Problem of ofce size What should be the size of an ofce for clerical workers? This is the problem I have taken, a problem central to the design of any modern ofce block. Nobody, of DOI: 10.4324/9781003313052-3

Ofce size

11

course, expects one ofce size to be suitable for all organizations and all sites, but it would be of value to an architect to know the optimum size in certain situations and within what range it is most advantageous to keep. If the architect could get advice either from his client or directly from research findings on the problem of ofce size, it could influence, in some cases, the whole shape of the building and in others, at least the amount of flexibility for which the architect would provide. Information concerning an optimum ofce size has even greater relevance for the designer if acceptance is given to the suggestions that windows are necessary for view rather than for daylight purposes, provided the levels of artificial illumination are reasonably high (see Wells, 1965a; Larson, 1965; Demos, 1966 [1997]; and Markus, 1967). For if it is not necessary to provide daylight at the back of the room, the depth of an ofce may be anything compatible with the objectives of the organization. However, if an external view from every desk is necessary, the height of internal partitions will be limited so that if the organization requires large ofces for efciency and flexibility, it must have open ofces for psychological reasons. One of the crucial psychological questions then is, ‘What is the optimum size for an open ofce?’ Psychological problems of ofce size The method chosen for measuring room size was based on the assumption that some of the most important objects in the human environment are other humans. The measurement chosen was the number of people in a room. The number who normally work in any given room is directly related to the physical size of that room, especially if density is kept constant. In this study, density was kept constant by selecting similar grades of workers doing similar jobs in the same nationalized industry. Stages of the research project Having decided how to measure the architectural variable, the psychological exploration of its implications proceeded through six stages, the final one being a more sophisticated example of the first one. The rest of this article will be devoted to a detailed account of each of these stages, but they will first be summarized. These stages are all essential for psychological research into architectural problems, although not necessarily in the order presented here: without any one of them, the results achieved will be either unrelated to the likely verbal reactions of the client with whom the architect must communicate, unsubstantiated in the field where the results will be applied, or not related to the definitive findings of the laboratory. The process is also cyclical, for research (like design in many cases) develops through elimination of unfeasible solutions and study in ever-increasing detail of remaining possible areas. Stage 1: Is there a problem? Study of the reactions of the relevant users to diferent degrees of the variable in question. Stage 2: Feasibility of studying the problem. Study of the relationship of the architectural variable to other relevant attitudes.

12

David Canter

Stage 3: Does the variable really influence what people do in real life? Field experiments to establish relationship between architectural variable and behaviour. Stage 4: Does the variable really influence what people say (or the way they feel) in real life? Controlled surveys to establish relationship between the architectural variable and what people say or feel apart from their reactions to the architectural variable. Stage 5: Can the variable be shown to influence behaviour under the rigourous conditions of the laboratory? Laboratory experiments concentrating on the crucial problems thrown up in the other four stages. Stage 6: What is the problem seen to be now? If the earlier stages are carried through with proper scientific rigour and a modicum of imagination, the original problem will now be seen to be quite naive and possibly misguided. A whole host of new problems will have erupted, each demanding a series of similar studies for their exploration. In particular, it is likely that a new look at user reactions from a more sophisticated viewpoint would be of value. Stage 1: User reactions When dealing with variables which the architect will be manipulating, it is important for the psychologist to know the current attitudes towards these variables. First, because these attitudes will indicate what the problem is and the way in which it might be studied, the key problems are more likely to be clarified during this process and conflicting attitudes revealed. Second, because in research intended for eventual application, it is important to know the climate of opinion within which the problem is being studied; this afects the perceived relevance of any findings and the ability of the designer to use them. Finally, knowing user responses to the architectural variable is of great value in its own right, for it is in this way that knowledge of the causal mechanisms behind what people think and say about buildings will be gathered. However, it is unfortunate that so much psychological- and sociologicaloriented research in architecture stops at this first stage. In the research project being presented here, the first stage consisted of investigating people’s reactions to ofce size. This had already been well established by Wells (1965b) (see Manning, 1965). He had shown clear diferences in attitudes towards large ofces between lower grade clerical workers, supervisors and managers. The clerical workers had the most negative attitudes, the supervisors the most positive and the managers were in between. All groups were able to back up their attitudes with reasons, but Wells concentrated on the explanations given by management for and against large open ofces. They felt, in general, that large open ofces were more efcient to clean and organize and easier to supervise but that they provided a less satisfactory working environment for clerical workers than smaller ofces. The other side of the question of attitudes is the one of size of room in which people say they would prefer to work. To explore this a question was put to some 1,200 ofce workers of all grades as part of a wider survey. The question was, ‘Other things being equal, in what size of room would you prefer to work?’ The respondents were given a list of seven diferent room sizes to choose from, ranging from ‘one

Ofce size

13

to four people’ up to ‘more than 80 people’. The percentage of people choosing each room size is shown in Figure 2.1. It can be seen there that the larger the room, the fewer the number of people who would prefer it. It will also be noticed that the relationship is a fairly smooth, curvilinear one. This indicates that although there is a definite relationship, it is a complex one. As this relationship was taken as a starting point for the project, it was worthwhile looking at its complexity a little more closely. Work by Waller and Thomas (1967), Fouilhé (1960), Cohen (1956), Stevens (1961) and others has suggested that people do not deal with physical dimensions in a linear way, be it those of cost, time or the more usual physical dimensions, such as loudness and brightness. Generally speaking, a logarithmic plot produces a more accurate description of estimates of these dimensions than does a simple linear plot. It seemed likely, therefore, that part of the reason for the curve in Figure 2.1 is the lack of validity of the linear room size scale. The other axis of Figure 2.1 is also open to a closer look. This axis describes the percentage of the sample preferring the diferent room sizes. However, it is not uncommon for preferences amongst a homogeneous population to approximate to the Gaussian, or ‘normal’, distribution. It would, therefore, be predicted that if the

Percentage of population

50

40

30

20

10

1-4

5-10

10-20 20-30 30-50 Number in room

60-80

80

Figure 2.1 Percentage preferring each room size shown by answers to the question: Other things being equal, in which size of ofce/room would you be happiest working? Total sample 1,180.

14

David Canter

percentage preferences were indeed normally distributed, in this case, when drawn on a ‘normal’ scale, they would produce a straight line. The substitution of the two scales of Figure 2.1 is shown with the new plot in Figure 2.2, where it can be seen that there is a much neater straight-line relationship than in Figure 2.1. Having shown this straight-line relationship, it is now possible to predict the room size that the average person is likely to prefer and to state that percentage preferring decreases as a logarithm of room size. Taking this relationship and the results of Wells, these are the following implications for the proposed research project:

Percentage prefering/probability scale

1 Room size is a real problem for the architect (a) because there are opposing attitudes within the organization and (b) because these attitudes are measurable, can be expressed as a simple relationship and thus are clearly defined. 2 The first indications are that the average clerical worker prefers smaller ofces, and so if future research findings are to produce constructive applications, this viewpoint must be taken into consideration. 3 In future research, it would be of more value to have room size varying on a logarithmic scale (i.e., 10, 100 and 1,000 people per room rather than 5, 10 and 15), as this is more likely to show significant relationships.

50 40 30 20 15 10 5 2 1

2

3

4 5 678 91

2

3

4 5 678 91

Logarithm of number in room

Figure 2.2 As Figure 2.1 but percentage of preferences is drawn on a probability scale, and room size is drawn on a logarithmic scale.

Ofce size

15

Table 2.1 Showing relationship between room size and other attitudes Question

Is there a relationship with room size?

Taking average conditions, do you find that the available daylight with the supplementary electric lighting is comfortable? In what way have your personal relationships with members of staf outside your department been afected by moving into the new building? Compared with the building you have recently left, do you feel more or less isolated from other people in the organization? Generally speaking, do you think that the physical conditions (i.e., heating, lighting, ventilating, colour and size of the room) in which you work are important for your happiness at work? On balance, do you enjoy your work more or less in the new surroundings? †† Other things being equal, in which size of ofce (room) would you be happiest working? Do you feel that the lift service is adequate? Comparing buildings, do you prefer working in this building or the one you have recently left?

No

*

Yes* Yes† Yes† Yes† Yes† Yes* Yes†

χ2 significant at 1% level (i.e., the relationship should occur 99 times out of 100).

† χ2

significant at 0.1% level (i.e., the relationship should occur 999 times out of 1,000).

††

this relationship does not account for all of the preferences discussed in stage 1: User reactions and Figure 2.1.

Example of analysis of relationships test 16 This test consists of a series of questions involving analysis of relationships of various sorts. For each question, there are five possible answers, but only one is correct. So in each question, you are to pick out the one answer which you believe to be correct and write the letter corresponding to it in the space to the left. Next are some examples already worked out. C

D

A

Father is to son as bull is to: A mare D ewe B colt E cow C calf What number comes next in the following series? 1 2 4 7 A8 D 11 B9 E 12 C 10 The opposite of optimistic is: A pessimistic D silly B neurotic E whimsical C trivial

16

David Canter

Example of clerical aptitude test 15 Average balance

  $

Adams, D. B. Adamson, B. D. Allen, J. F. Anderson, H. G. Andrews, G. L. Armstrong, C. P. Atwell, C. R. Barnes, M. V. Baxter, A. N. Bloomfield, H. T. Bosworthy, A. P. Botts, K. L. Bronson, D. H. Brown, J. R. Browne, J. R. Carey, R. E. Clark, C. U. Clark, C. V. Cooper, B. L. Corbin, G. P. Corwin, G. P. Cosgrove, J. J. Coughlin, M. H. Cowan, F. L. Crane, N. D. Dale, A. M. Daly, L. E. Daniels, R. T. Dearborn, F. F. Decker, M. H.

2375,81 1942,89 267,45 1712,50 2130,26 1543,37 724,16 5004,18 384,62 637,42 1822,77 187,16 4768,33 967,26 1542,87 6144,14 318,89 1731,43 1135,58 395,23 1074,77 2597,16 1374,04 3418,21 954,16 8914,67 2354,36 1317,85 895,22 527,59

Credit rating

 

Balance ($)

Code

999,99 or less 1000,00 to 1999,99 2000,00 or more Name missing

N O P X

Take names in order Begin here 1 Barnes, M. V. 2 Bosworthy, A. G. 3 Clark, C. U. 4 Botts, K. L. 5 Andrews, L. G. 6 Daly, L. E. 7 Cosgrove, J. J. 8 Allen, J. F.

P _____ X _____ N _____ N _____ X _____ P _____ P _____ N _____

9 Decken, M. H. _____ 10 Cooper, B. L. _____ 11 Baxter, A. N. _____ 12 Dale, A. M.

_____ The subject has to fill in code letters for each of the names in the list on the right on the basis of the average balance next to those names in the alphabetic list on the left, according to the codes in the box in the top right-hand corner.

Stage 2: Relation of room size to other relevant attitudes Measurement of attitudes towards ofces of diferent sizes does not, of course, give much information about the relationship of ofce size to other behaviour besides preferences. All it does is suggest that these relationships are there to be found. This is true when studying many other psychological problems in architecture. To establish a preference for a particular level of any variable does not necessarily indicate that that is its optimum level when all other psychological variables are considered, not to say physical, social or economic variables as well. But although there may be many other variables militating against the importance, or indeed the relevance, of user preferences, it is still sometimes possible that clearly defined

Ofce size

17

preferences are the only real manifestation of the architectural variable that the psychologist can show. An example of this is Wells’s (1965a) work on daylighting and the work of Larson (1965) and Demos (1966 [1997]) on windowless schools, for in their studies, attitudes, although present among the users, do not in fact relate closely to other psychological efects of the variables under study. The tentativeness of the implications often reached at stage 1 usually makes it worthwhile to explore other concomitants of the architectural variable under consideration in a simple way in order to see if the problem is rich enough to warrant deeper investigation. In this case, the second stage consisted of giving out a short questionnaire to a large number of people in a new ofce building, working in ofces varying in size from one to 100 people. Answers to each item on the questionnaire were then analyzed to establish to what degree they related to the size of the room in which the respondent normally worked. The questions covered a range of topics to do with the building. The questions and whether they were found to relate to room size or not are shown in Table 2.1. The main conclusion to be drawn from this table is that the size of the room in which a person works does indeed relate to many attitudes which that person has. This study does not enable the nature of that relationship to be explored; in particular, it does not reveal whether these attitudes were directly influenced by the room size or were simply concomitant with it. It is essentially a feasibility study. The results indicate that room size is amenable to investigation at least in so far as that seven out of eight questions did show a relationship, and thus, there is hope that a more detailed study may prove fruitful. The intelligent person who has read this far may wonder at the data collected in these first two stages for no other apparent reason than to see if it is possible to collect such data. There are three comments which may reassure him. First, in a new area of research which rests so heavily on the willingness and ability of people to respond in their workday settings, which demands so much organization beforehand, which in many cases can be carried out only on one special occasion and about which there is little information to help in predicting probable results, it is indeed necessary to see if it is possible to collect meaningful data; otherwise, much efort may be wasted. Second, since it is not the aim of this article to provide research results in great detail, many of the suggestions and ideas which were generated by the data, but were not relevant to the central theme, have not been presented. Third, during these first two stages, much unsystematic information is collected both in developing questions and eliciting answers which is of great value in helping the researcher to formulate the problems in which he has specific interest and in giving him a ‘feel’ for the problem area. Stage 3: Field experiments Stages 1 and 2 have shown that there are likely to be interesting problems associated with room size that were producing efects gross enough to be measured with the crude tools at the disposal of the psychologist. This next stage involves the specific definition of the key problems and the exploration of these problems as rigourously as possible in situations as near as possible

18

David Canter

to those of real life. It is also essential at such a key point in the development of a research project that the psychological variables measured should be activities rather than statements or opinions. The reason for an emphasis on behaviour (as it is usually understood), rather than attitudes, is that at this stage, the interactions of the architectural and psychological variables must be reduced to their simplest level in order to facilitate examination. The study of attitudes at this stage would not facilitate this, for by the time an individual has commented on the efects of a particular variable, so many unknowns would have entered between the comment and the actual efect that interpreting the real causes and their links to the efects would be most difcult. There is another important reason for concentrating on behavioural variables. This is the fact that many clients are mainly interested in what the users of their buildings do, not what they say and certainly not what they say about the building. A final point about this stage before referring to the example is that this is also the point at which theories and hypotheses about the real processes underlying the possible relationships are formally stated. These cannot be stated until the problems have been clarified which, as has been shown, requires stages 1 and 2. Which of the key problems of room size is most amenable to experimental psychological study outside the laboratory? The problem described must contain variables amenable to objective measurement, and its solution must be relevant to psychology, management and architecture; otherwise, it cannot be studied, and the results of studying it would be impossible to validate. The problem chosen was that of the relationship between performance of the job and the number of other people who are normally in the same room. This was chosen because it is central to all discussion of ofce size. For if the size of room causes people to work less efectively, then the reason for having large rooms, namely organizational efciency, would need reviewing. The ideal way of studying this problem would be to measure what, and how much, people actually did at their work in diferent-sized rooms. But in order to do this meaningfully, the work going on in these diferent rooms would have to be identical, and situations like this could not be found. There was always some small diference in the way the job was done, the workflow throughout the ofce, the nature of supervision and the like which made ofces of diferent sizes unsuitable for comparison on the basis of actual work. There was also the difculty that measuring people at their work suggests ‘organization and method’ assessment to which many ofce workers are antagonistic (perhaps with every reason). Performance was, therefore, measured using standardized tests. The tests chosen had been developed in America as methods of selecting clerical staf and, consequently, were closely related to performance of actual clerical jobs. Examples of the two tests used are shown in Table 2.1. There, it can be seen that one of them (clerical aptitudes test), which is timed, is essentially a measure of the speed and accuracy with which a person can find a name from an alphabetic list and code the figure next to it according to a simple rule. The number of names correctly scored within the time limit is taken as a measure of performance. The other test (analysis of relations test) is much more like an ordinary untimed intelligence test, measuring higher mental abilities.

Ofce size

19

What relationship might be predicted between scores on these tests and room size? Many theories would lead to predictions of these relationships, but for simplicity, three will be discussed. The first theory is the frequently stated one that the bigger the ofce, the more distracting stimuli there will be in it, and as a consequence, the poorer will performance be. As developed by Broadbent (1958), the theory states that people can deal with only a given amount of information at any one time, and as a consequence, when doing something that requires searching for information or storing information, other distracting stimuli will be detrimental to performance. One of the consequences of this theory is the prediction that room size will relate to a decrease in the scores on the more complex intellectual task. There are two qualifications that must be imposed on these predictions. One is that disturbance of performance has been consistently shown in the laboratory only when the disturbing noise was above 90 dB (see Rodda, 1967, for summary of this research). In the laboratory, this was usually abstract or white noise, and one might expect disturbance at lower levels with information-carrying noise. The second qualification is that although the frequency of occurrence of noises might increase in larger ofces, the relevance – and, hence, informational value of those noises for each individual – is not likely to increase nearly so quickly. This would suggest that the rate of decrease in performance with increase in room size would slow down after a certain size was reached. In other words, the graph of performance against size would tend to level of if not actually show an increase in performance at very large sizes. The second theory is not nearly so well rooted in the psychological literature. One basic assumption of this theory is that each room has its own peculiar atmosphere which is similar for rooms of similar sizes in which similar work is being carried out. The other is that the social groupings produced in larger ofces are likely to be less cohesive (see Wells, 1965b), and this, taken with the negative attitudes current about large ofces, leads to an inefcient working environment. These two assumptions together would lead to the same prediction as the first theory for lower performance in bigger rooms and also for lower performance on tasks which are untimed and demand more motivation to tackle them, as is the case with the analysis of relations test. The diference between the two theories can be explained simply as follows. The first theory states that negative attitudes towards large ofces are due to the negative environment within them which in its turn produces lower performance. The second theory states that negative attitudes towards large ofces produce negative orientation to work which in turn produces lower performance. Whether the first or second theory best fits the facts can be experimentally shown by testing people in their own room and then in the same room, similar people who do not normally work there. As the first theory depends upon the properties of the physical environment provided in rooms of diferent sizes, this would be expected to influence newcomers as much if not more than people used to it. On the other hand, as the second theory depends more on the ‘atmosphere’ in the rooms and this would not be immediately apparent to newcomers,1 then efects of room size should disappear when people are performing in rooms to which they are not accustomed.

20

David Canter

A third theory states that a certain sort of person will take and keep a job in a large ofce or that diferent people will perform well in ofces of diferent sizes. The personality dimension which seems to fit this theory is that of extroversion/ introversion which is a dimension easily measured by psychologists (see, for instance, Eysenck, 1965). In order to support this latter theory, all that is necessary is to show either a relationship between room size and the extroversion of the staf or a diference between the performance of extroverts and that of introverts in rooms of diferent sizes. In order to test these theories, the two tests shown were given to ofce workers and a measure of personality was taken in six ofces of diferent sizes (varying from seven to a hundred people in the room) in two buildings, both local head ofces of the same nationalized industry. In some cases, the workers performed in their own ofces and in others, in strange ofces in the same building, to which they were taken for the testing. First, the results of the personality measurement were considered. These showed that there was no relation between degree of extroversion and room size or performance. Two possible explanations of this are that extroversion is not an adequate measure of personality referred to in the third theory, or the third theory is not an adequate description of the facts. If the former, some new dimension of personality must be hypothesized and measured. Second, the results of performing in the room in which the clerk normally worked were assessed. These are shown in Figure 2.3. This shows that there is a

Performance percentage/probability scale

84%

Clearical

50% An

aly

sis

of r

ela

aptitude

tion

20%

10

100 Logarithm of room size

Figure 2.3 Relationship between room size and performance.

shi

p

Ofce size

21

significant relationship between room size and performance on both tasks,2 performance decreasing as room size increases. Figure 2.3 also shows that the rate of decrease is greater for the reasoning task than for the clerical task. This supports both the first and second theory. However, the results for the clerks being tested in rooms diferent from their own show no relation to room size. This supports the second theory. There are, of course, other explanations for the relationship described earlier, but all that need be noted at this stage is that performance decreases with increased room size and that there is some evidence that this is not owing to the hypothesized greater distractions within the larger ofces. Stage 4: Controlled survey All organizations are interested in morale, or worker satisfaction, and in most architectural research projects, the influence of the building on user morale is of some relevance. The diference between user morale and the user reactions of stage 1 must be emphasized. In stage 1, it is specifically satisfaction with the building and the built environment that are the dependent variables. In stage 4, attention is turned to satisfaction with the job and the working situation in general. An interest in worker satisfaction is not purely altruistic, as morale, as measured by questionnaires, has been shown to relate to turnover rates, absenteeism and sickness absence in many industries. It is often stated that this or that architectural variable influences morale, but these statements are rarely supported with any real evidence. Thus, when exploring the psychological implications of any architectural variable, it is of value to establish the nature of the relationships of that variable with user morale. Morale, however, is not a simple phenomenon, and it is necessary to identify its component parts in any study. In the study of room size, an area head office was found for the industry referred to earlier. Within it, there were three groups of rooms. One group had on average of ten people per room; another, 20; and the third, 35. The work of the people who occupied these rooms did not difer significantly from one room to the other. Questionnaires were given to all the people in these rooms, asking them about various aspects of their work and how satisfied they were with it. Thirty replies were returned from each group of rooms. The replies were analyzed to find out what the components of morale were in this case and then each of these components was related to room size. Three of the components showed a linear relationship with room size, each one showing an increase in satisfaction with an increase in room size. The three components showing this unexpected relationship were (a) satisfaction with the amount of information received from management and the amount which reaches management; (b) satisfaction with the job itself and the degree to which it is the ‘right job for me’; and (c) lack of dissatisfaction with the number of distracting stimuli and disturbance in the room. Finding 1 suggests that, as managements suggest (see Manning, 1965), there is felt to be an increase in efficiency of communication in a large ofce so that the worker feels more satisfied in that respect. Finding 2 suggests that the larger ofce gives an individual a greater

22

David Canter

opportunity of doing what he wants to do and of finding his niche within the organization. Finding 3 fits in with the conclusion reached at the end of stage 3 that there is no appreciable increase in distractions in the bigger ofces. At first, this support of a tentative hypothesis seems most surprising, especially as it contradicts the quite highly regarded theory of distraction referred to earlier. However, one of the qualifications presented with that first theory takes on a new importance in the light of finding 3. This is the qualification of relevance of information. It was suggested earlier that as an ofce increases in size, the frequency of occurrence of noises is likely to increase, but the probability of the noise being relevant to any given individual is likely to decrease. Therefore, if actual awareness of disturbance – and, hence, dissatisfaction – is a function of both frequency of occurrence and relevance of noise, it seems possible that dissatisfaction will decrease with room size owing to the greater decrease in relevance. In summary for this stage, it can be said that over the small range of room sizes studied, there was an increase in satisfaction as room size increased. Stage 5: Laboratory experiment So far, the results have been culled from the field. The selection of subjects, measuring instruments and as many of the other relevant variables as possible have been carefully controlled, but results were still susceptible to the whims of coincidence. The results might have been produced by uncontrolled variables instead of those hypothesized. Consequently, it is always essential in this type of study to isolate the kernel of the problem and redefine it in terms which can be examined under controlled conditions, even though realism may be sacrificed for scientific validity. In the present case, one of the central problems is the cause of lower scores in the performance tests. The findings of both stage 3 and stage 4 suggest that deterioration in performance is not due to the distracting stimuli in the situation. If, therefore, it can be shown in the laboratory that distracting stimuli similar to those found in offices do indeed afect performance, then doubt will be cast on these findings. Another surprising finding from stage 3 was the lack of relationship with personality variables of performance under diferent conditions. This was re-examined in the laboratory. For this experiment, 48 architectural students were given personality inventories which enabled them to be divided into four groups, approximating to extrovertedneurotic, extroverted-normal, introverted-neurotic and introverted-normal. Groups of eight individuals were then tested under three conditions, the order of conditions being random for each person. In the first condition, they had to complete the clerical aptitude test without any distraction at all. In the second condition, a tape recording of a discussion of the new school of architecture in which these students worked was played to them at 60 dB while they performed the task. In the third condition, the same recording was scrambled by rerecording the discussion out of phase on top of itself and played to the students at 60 dB while they performed the task. The tasks in each of the conditions were similar, but not identical, forms of the clerical aptitude test.

Ofce size

23

The results from this experiment are all negative; it was not possible to show any relationships that were statistically significant. This then supports the argument so far presented. To summarize the results of the five stages, it seems that performance decreases as room size increases, but this is not due to the distractions in the ofce. Perhaps it is due to attitudes towards large ofces and the nature of the social groupings within them. Stage 6: Back to square one Research tends to be a spiral process. The same problems are looked at again and again, in ever-increasing depth and in the light of new findings. This will certainly be true for architectural research for some time. The basic problems will take on diferent perspectives as our knowledge increases. In the problem of room size, it appears that attitude towards the room per se has some relevance for behaviour within that room. Consequently, what is now necessary is deeper study of reactions to room size. No results can be reported on this, but research relevant to it is now being undertaken by the Building Performance Research Unit at Strathclyde University. The problem now is to find the ‘molecules’ of users’ reactions to buildings – in other words, the key dimensions along which they think of buildings – and then to relate these dimensions to behaviour. The results of the initial study along these lines carried out in conjunction with the Birmingham School of Architecture may serve to illustrate the approach. Thirty-two students were asked to describe 22 sketches of house designs by filling in 45 adjective scales. The results from this were then analyzed by computer to see what the main components of these descriptions were, in much the same way that the components of morale were elicited as described earlier. The three main components which were produced were life (boring, dead, commonplace), familiarity (hard, unfriendly, unwelcoming) and stability (stable, coherent, simple). Each of these components is hypothesized to be a basic dimension common to many people (similar studies recently carried out with schoolchildren and ofce workers support this). It is a basic unit that will enable reactions to buildings to be related to behaviour within them. These underlying dimensions will facilitate this because they are more subtle and more deeply based in user reactions than the more conventional approach presented in stage 1. Future problems The methods and results described earlier give rise to problems and warnings for future research. Whether it is the methodology or the result that produces the following comments is an academic question that may be left to academics to discuss. 1 The relationship between verbal behaviour and other forms of behaviour in the building context is not well understood. The main implication of this is that

24

2 3 4

5

David Canter psychological research in architecture should always include examples of both forms of behaviour. The straight-line relationship shown in Figure 2.2 might be a basis for developing a series of laws of preferences for diferent architectural phenomena. By exploring the problems in many diferent ways, it seems possible that some of the difculties of research in this field may disappear. Manning (1967) has suggested that there should now be research into dissemination of research findings. This is necessary, but surely, there is a need also for a reorientation of much research so that it will relate to real-life problems. Research described in this paper has all the formal properties and components of classical scientific exploration, but its general orientation is diferent on two counts: the problem throughout was developed with an eye to the place of the results in architectural design, and it was studied with consideration for the attitudes, reactions and behaviour of relevant building users. It is hoped that this approach has been shown to lead, at least, to research which is more involved with normal people in their normal setting than has much psychology in the past.

Acknowledgment The author wishes to acknowledge the Pilkington Research Unit under whose aegis the research herein described took place; the Building Performance Research Unit for its queries, comments and ideas during the preparation of this paper, and all individuals and organizations who provided indispensable help, time, space and completed questionnaires. The example of an analysis of relationships test shown on p. 883 is from Analysis of relationships (1960). The example of a clerical aptitude test on the same page is from Clerical aptitude (1951). Notes 1 This, of course, is only an assumption at this stage, but it is open to scientific investigation. 2 r for clerical aptitude = −0.47; P = < 0.005; N = 49 r for analysis of relationship = −0.28; P = < 0.05; N = 49

References Bennett, G. K., and Gelink, M. (1951). Clerical Aptitude. New York: The Psychological Corporation. Broadbent, D. E. (1958). Perception and Communication. Oxford: Pergamon Press [Aa3]. Cohen, J. (1956). Le temps psychologique. Journal Psychologique Normale et Pathologique, 53, pp. 285–306. Demos, G. D. (1966 [1997]). Controlled Physical Classroom Environments and Their Efects Upon Elementary Schoolchildren (Windowless Classroom Study). Ofce of Riverside County Superintendent of Schools, Palm Springs, CA. Eysenck, H. J. (1965). Fact and Fiction in Psychology. Harmondsworth: Penguin Books [Aa3].

Ofce size

25

Fouilhé, P. (1960). Evaluation subjective des prix. Revue française sociologique, 1, pp. 163–172. Ghiselli, E. E. (1960). Analysis of Relationships. Palo Alto, CA: Consulting Psychologist’s Press. Larson, C. T. (1965). Efect of Windowless Classrooms on Elementary Schoolchildren. Ann Arbor, MI: University of Michigan. Manning, P. (ed) (1965). Ofce design: A study of environment. University of Liverpool, Department of Building Science, 1 (4), pp. 317–319. [(92)] Ba4]. Manning, P. (1967). Windows, environment and people. Interbuild/Arena, October [(32): Ba4]. Markus, T. A. (1967). The function of windows – a reappraisal. Building Science, 2, pp. 97–121 [(32)]. Rodda, M. (1967). Noise and Society. Edinburgh: Oliver & Boyd [Ab1]. Stevens, S. S. (1961). To honour Fechner and repeal his law. Science, 133, pp. 80–86. Waller, R., and Thomas, R. (1967). The cash value of the environment. Arena, January [Ba4]. Wells, B. W. P. (1965a). Subjective responses to the lighting installation in a modern ofce building and their design implications. Building Science, 1, pp. 57–68 [(92): Ab7]. Wells, B. W. P. (1965b). The psycho-social influence of building environment. Building Science, 1, pp. 153–165 [Aa3: Ba4].

3

Need for a theory of function in architecture David Canter

Why a theory of function is necessary Most architects would agree that, in designing a building, the probable users should be considered (Canter, 1969). The main difculty is that there is no clear understanding of how buildings and users interact. No coherent framework has been developed which presents both the user and the building in the same picture. This is partly why many architects find information from social scientists working on architectural problems irrelevant. It is also partly why most user surveys relate to specific situations only; such surveys usually consist of cataloguing reactions to specific buildings and so provide information about diferences between people in response to an individual building but do not systematically indicate how diferences between buildings produce diferent responses. Architectural training does not encourage a search for general frameworks; emphasis is laid on the success of particular solutions to specific problems. Further, there is no way of recording an architect’s success or failure over a period of time. Normally, it is assumed that what an architect provides is what his client needs, but this premise is never tested. Often, architects avoid understanding the interaction of users with their buildings by getting information from clients, which is the first step towards the design solution. Typical is the space schedule; clients are required to act like architects when asked how much space users need. Answers are commonly based on space used at present, but this assumes that the designer of the spaces with which the client is familiar estimated the user’s requirements correctly. An analogy might be a doctor asking a patient what medicine he thinks he needs. In giving the client what he wants, it is the architect’s task to convert the client’s thoughts, wishes and feelings into architectural terms. But to do this, he ought to understand scientifically how buildings relate to individuals. He must understand what buildings are and how they function for users. This is central to design, but design decisions are rarely consciously based on these ideas. A systematic account of the buildings’ implications for users is necessary; it would enable architects to design adequately for people and to use information from social scientists. Not until the rudiments of such an account are made explicit will architects and social scientists fully appreciate how design decisions are inhibited by the existing vague building/user framework. DOI: 10.4324/9781003313052-4

Need for a theory of function in architecture

27

Development of a valid and meaningful framework should be a joint enterprise between designers and social scientists. This paper aims to start a discussion on how buildings and people interact by using concepts from psychology. It is hoped, at least, to point out gaps to architects. Another aim is to show that the designers’ often covert assumptions about building/user interaction considerably influence design decisions.

An analogy for behaviour

A description of users must be central to any description of interactions between buildings and users because the nature and behaviour of the people whom an architect believes to be the users of his buildings determines the kind of building produced. Any theory of human behaviour can only be an analogy. One of the most widely used analogies is the mechanical one in which individuals are thought of as machines converting inputs into outputs to survive. Inputs are often referred to as stimuli and outputs as responses. When an architect holds this belief, he aims to produce specific inputs. This is suggested by the many architects who draw furniture arrangements and routes they suppose people will take through buildings. This mechanical view of man is common yet inefcient. It assumes a distinction between people and their environment with stimuli from environment and responses from people. Logically and psychologically, this is an arbitrary distinction. The way an individual becomes aware of himself as separate from his environment is complex and perhaps never complete. People constantly act in order to produce certain stimuli and then respond to them. Hence, it has been found (Myrick and Marx, 1968) that people do not use routes predicted by the architect or arrange furniture as he expected. However, mechanistic analogies can be valuable when considering the complex process of adaptation. This is how an individual responds to stimuli from the environment in order to maintain himself within comfortable limits. Some of the most primitive forms of adaptation (often innate) are well-known physiological mechanisms whereby organisms respond to changes in intensity of sensory stimulation by changes in sensitivity. An organism’s efciency in receiving stimulation from the environment can thus be maintained. Many aspects of psychological behaviour can also be seen as adaptation. The limits within which the individual tries to remain are subjective. This is clear when a person maintains a set attitude, although the facts do not support it. For example, people often perceive a matchbox to be the same size no matter how distant. Such processes enable the whole of adult experience to have a more or less stable form. Indeed, much adult behaviour is probably an attempt to maintain stability of a subjective world. To do this, adaptation must be a continuous and dynamic process. Often, the environment presents novel situations, or common situations from an unusual viewpoint. On the other hand, the individual may change his view of particular aspects of the environment on reconsidering his experience or by maturing. Consequently, people are always readapting and organizing their current interpretation of experience to fit in with previous experiences, or they reorganize their memories of previous experience to take account of new ones.

28

David Canter

Thus, the static stimulus-response analogy must eventually lead architects to make design provisions that will quickly be misused, unused or misunderstood. Buildings as inhibiting and facilitating systems

How do buildings afect adaptation? The simplest answer is that most buildings are attempts to hinder particular stimuli from getting to users and to facilitate the reception of other stimuli. Buildings also aim to help or hinder users in reacting to stimuli. If this description of the adaptive process is accepted, one can go further: buildings play a key role in providing a range of opportunities for response from which users can choose. The range must be such that they can make a viable choice whenever they need to adapt to changes which they wish to accept or cannot control. So that users can maintain their subjective balance in the face of changes in the environment (other than in buildings), they must be able to change their use of buildings. Flexibility, therefore, means providing a set of choices (particularly of stimuli) that will make adaptation to changes easier. Privacy becomes the provision of potential for choosing responses and activities. The more activities in which an individual feels he can indulge in a particular space, the more private it is. This idea is presented by Ittelson et al. (1972) in their interesting study of two hospitals. They showed that a patient’s range of activities increases as the size of bedroom decreases so that the widest range of activities took place in rooms in which one person slept. If an individual’s level of satisfaction shows how well adapted he is to his environment, then his satisfaction might be expected to depend on how far he can control the environment. The recurring finding that people are generally satisfied with what they have is to be expected. If very dissatisfied and possessed of some control or choice, people would try to change their situation to make it more satisfactory. Therefore, the wider the choice of stimuli and response that an architect provides, the more likely are users to be satisfied. Buildings as stimuli

Buildings themselves also present stimuli. It is with this aspect – appearance – that architects have been most concerned, and this is the aspect most discussed in the architectural press. Unfortunately, there is evidence to suggest that it is of relatively little interest to users. Herzberg et al. (1954) undertook a study in which he asked people in many diferent types of job what they found satisfying or dissatisfying about their work situation. Physical surroundings were rarely mentioned, except when unsatisfactory. This is common in such studies. Surroundings are not thought so important as pay, supervision, social contacts and the like. Yet no detailed studies of priorities have been made in systematically varied environments. Thus, though appearance is usually of little importance to users, we cannot assume that it is never important.

Need for a theory of function in architecture

29

The making of architectural stimuli

To say that people generally take little interest in their physical surroundings is not to say that the latter have no significance. They may even be particularly important since the appearance of buildings can provide information about their users and owners (Canter, 1969). For example, a cathedral might be expected to look like the sort of place in which God would be at home, just as a bank might be expected to look as if bankers will care for their clients’ money properly. To incorporate this aspect of architecture (i.e., the ‘meaning’ of a building) into the ideas set out previously, the stimulus/response analogy must be expanded. A way of evaluating and interpreting stimuli is needed so that the most efective action can be taken. Although there will be many general similarities due to common cultural background and physical make-up, outcome of this process will differ from individual to individual because everyone has diferent experiences and capacities. An interesting way of looking at this symbolic aspect of the built environment is put forward by Gofman (1959). He suggests that in many situations, individuals seem to act out a part – not from insincerity but because much human activity takes the form of role playing. Maintaining a role is greatly aided and given credibility by the right setting. A building, then, is often designed to be a suitable setting for the activities carried out in it. An individual’s idea of a suitable setting will relate to his notion of who does what in the setting. This might have nothing to do with what actually happens or with the best conditions for the performance of the real activities. Thus, an architect’s ability to design a building that satisfies his client may depend as much on his ability to interpret the client’s idea of what he does, and to provide an adequate setting for notional activities, as on his ability to elicit what the client would actually like to do in the building. In short, the client’s satisfaction might depend more on the architect’s psychological insight than how adequate the ‘brief ’ is.

Facilitation of reception of stimuli

A stimulus is any aspect of the environment to which the individual could react if he knew of it. The last condition is important because the environment overflows with potential stimulation, but we are aware of only a small portion of it. Noise outside a room is unnoticed unless it increases above a certain level and interferes with activities in the room. Tidying a room might not be apparent until a book, previously left in a prominent position, is required. Selectivity of attention and perception is so common that yet another addition to the stimulus response analogy is needed to deal with it. The addition is what Broadbent (1958) calls a ‘filter’ which can select relevant stimuli. This is not always efcient, often because the individual does not have enough information to ‘tune’ it properly. Thus, it is often necessary to attend to more information than can be absorbed meaningfully. To help in such cases, buildings might be designed to act as another filter of stimuli.

30

David Canter

Ideally, a building would let the individual receive only that noise or light – or any other stimulation – that he could deal with at one time. More practically, a building might be constructed to facilitate the human filter’s ‘tuning’, for example, by making noise sources apparent or visual inconsistencies meaningful. Buildings not only provide constant, direct stimulation but also play a critical role in facilitating or limiting reception of most other stimuli which people are likely to receive. At least on a logical level, this facilitating or filtering function of buildings differs from their stimulus properties in important ways. In most cases, such properties remain constant from the opening of a building, although the psychological implications vary from person to person. On the other hand, filtering is related to what goes on in and outside the building at a particular time. The architect can provide stimulus properties in great detail; he might even give detailed instructions for changing the stimuli. But the filtering function has to be estimated or, more often, guessed. The degree of filtration required relates closely to the individual and what he is trying to achieve, and a great deal of psychological information is needed to assess how well buildings filter stimuli. Because people adapt to their environment in a complex way, it is inappropriate to try to control particular physical stimuli without reference to the general context. For example, in cellular ofce buildings, aural privacy is often provided with sound-reducing partitions, which create quiet conditions within each ofce. In open-plan ofces, however, aural privacy is often achieved by masking other people’s speech with relatively high background noise. There is no evidence to show that, provided privacy is achieved, people find either set of conditions unsatisfactory. No simple relationship between aural privacy and quietness can be inferred; the context determines what level of background noise is appropriate. Responses Having provided cues that will lead the individual to believe that he is in the right place for what he wants to do, and having ensured that the required information can be received, architects must also provide opportunities for suitable responses. Architects are usually more subtle than this, for they help to ensure that only a few responses are selected from all possible responses to particular stimuli. For example, if someone wants to speak to a colleague in the same building, his decision to telephone or go and see him is partly influenced by how difcult it would be to make a visit, which, in its turn, is partly influenced by the design of the building. The limiting of responses by a building or group of buildings only recently became apparent to architects. For it is generally held that human beings possess what might be termed free will and so may do as they wish. It is just conceivable that our wish to perform an act might be influenced by the ambience of the building we occupy, but, this apart, many architects would hold that buildings cannot influence

Need for a theory of function in architecture

31

actions. But making an action seem more natural, desirable, efcient or comfortable within a particular room is tantamount to limiting the actions that are carried out in it. It is also clear that there comes a point beyond which a particular action is impossible and before which, though possible, it is difcult. The distinction between the impossible and what only seems so is vague. In many ordinary situations, we assume things are impossible and so never try to do them. Hence, we never discover whether they really are impossible, and our assumptions are untested. Because the design of a building leads to certain assumptions about what seems to be possible, our behaviour may often be indirectly influenced because we do not try to do things which are really possible, believing them to be impossible. The influence of relationships

Some of the most important stimuli and responses are exchanged between people, on the basis of which personal relationships develop or fade. Therefore, one of the most important efects of architectural design is that of helping or hindering social interactions. An aspect of social interaction that has received detailed study relates to friendship formation. The underlying assumption in theorizing about research findings on friendship formation is that one can only become friendly with a person with whom one interacts. Furthermore, other things being equal, the more one interacts with the same person, the more likely is the relationship to develop positively or negatively. Thus, in a building designed so that people can meet often, one would expect them to be more friendly than where similar people doing similar work cannot interact often (though, of course, many other factors afect friendship formation). It follows that an individual who normally occupies a position in a building through which many people pass will have more acquaintances than someone in a more isolated position. There have been several pieces of research, in housing units (Festinger et al., 1950), in ofces (Wells, 1965; Gullahorn, 1952) and in university halls of residence (Warr, 1964), which bear this out. They showed clear relationships between building layouts and friendship patterns of occupants. It was even found, in one study carried out in the ’30s (Bossard, 1932), that there was a simple relationship between residential propinquity and probability of marriage. All these studies show how design and layout of buildings, by limiting the possible range of responses, influence aspects of behaviour seemingly far from architecture. Generally speaking, while buildings may not influence behaviour itself, they often contribute towards it. Applications of the model to the concept of building type

One of an architect’s most important tasks is to decide with his client on priorities of diferent design requirements. These will, of course, be concerned with the stimuli and responses necessary for performing the major tasks to be carried out in the building.

32

David Canter

Building types might thus be distinguished by categorizing the stimulus and response requirements of activities that each type requires. A more important distinction might be evolved by examining priorities of these requirements. Looked at this way, the old system of building-type categorization is often inadequate. For example, modern ofce buildings are visually pretty much the same inside and out, yet what goes in them and the people and organizations that use them are quite diferent. The head ofce of an organization making a specialized product has quite a diferent hierarchy from, say, the main ofce of an insurance company. A cellular block of ofces might be right for one but not for the other. Classifying buildings into conventional types is, therefore, not usually helpful; categorization according to type of organization would probably be more useful. A hierarchy of needs

So far, human needs and motivation have not been discussed. Some architects seem to work on the basis that human beings live at one of two levels: they require either succour for physiological needs or beauty for aesthetic reasons. This picture is obviously incomplete. If buildings are to be designed to enable people to deal more fully with their world, a more accurate classification of human wants and needs must be obtained. Such a classification has been studied by psychologists for many years but, as yet, inconclusively, and some have even suggested there is no place for the concept of needs with a scientific discipline. However, one theory is worth describing to show that more knowledge of motivation is necessary. This theory assumes that individuals always develop because of, and in spite of, interaction with the world around them. Some psychologists describe this in terms of gradual flowering of the inner essence of the individual; others see it as a refinement of the mechanisms of human behaviour owing to the nature of adapting and learning processes. But the result is the same whether the individual is the instigator of his development or is propelled by external forces. As he interacts with people and things and as his more essential psychological needs are met, he becomes motivated by subtler, less dominating forces. As Blai (1964) says: The chief principle of organization in human motivational life is the arrangement of needs in a hierarchy of lesser and greater priority of potency, and, under conditions of equal deprivation, the prepotent needs are more urgent and insistent than the others. The chief dynamic principle animating this hierarchy organization is the emergence of less prepotent needs upon the gratification of more prepotent ones, and until the prepotent ones are relatively satisfied, the others do not emerge as consistent motivators of behaviour. The individual is dominated and his behaviour arranged only by unsatisfied needs. If hunger is satisfied, it becomes unimportant in the current dynamics of the individual.

Need for a theory of function in architecture

33

Maslow (1962) has described this hierarchy of basic needs, mainly from clinical studies and observations: 1 2 3 4 5 6 7 8

physiological safety belonging respect information understanding beauty self-actualization

This is not to say that architects should not consider satisfying the need for beauty before they have helped to satisfy that for belonging. Rather, they must realize that in any building where many jobs are being done, diferent people at diferent points in the hierarchy will be dominated by needs according to their particular level. Blai’s study shows that the lesser prepotent needs of self-actualization, advancement, interesting duties and leadership were all selected more in the professions than in the trades. Conversely, the more prepotent needs of respect, job security, congeniality and independence were selected in greater amounts by lower socio-economic groups. These are only tentative findings from one questionnaire study of only 480 people. However, if this sort of information were firmly established, it could be used, together with details of the organizational structure, to advise the architect how the building he designs must help reception of important stimuli and delimit appropriate responses so that the individual and the organization are satisfied. Summary and implications

This article was first drafted some four years ago. Since then, many architects have developed attitudes to their profession, like those described previously. Two books (Sommer, 1969; Moller, 1968) have been published which deal with the relationship between buildings and behaviour. However, no attempt has yet been made to develop a theory of man-environment interaction that is of value during the design process. Disciples of systematic design have developed processes which do not consider people’s behaviour in buildings. This paper is only a beginning, but already, it is clear that there are many points in the design process where implicit assumptions about human behaviour lead to inadequate solutions. Probably the worst efect of these assumptions is that they limit the architect’s creative freedom by imposing unnecessary relationships on him. The variety of choices of action a client wishes to be available to users of his building should be precisely elicited, and they should form the core of the brief. The architect should also elicit types of information that the client wishes to be made available to the users of his buildings so that they will have data on which to make

34

David Canter

their choices. Thus, most difculties in turning the brief into a design will be caused by the conflicts and interactions arising from attempts to facilitate all the required stimuli and responses. Finally, stimuli provided by the buildings’ fabric should be taken into account because they enhance the efectiveness of activities carried out in the building by subtle mechanisms, such as expectancy and ‘meaning’. Until an acceptable theory of how architecture functions is available, a fruitful dialogue between social scientists and architects cannot develop. References Blai, B. (1964). An occupational study of job satisfaction and need satisfaction. Journal of Experimental Education, 32 (4) [(E2s)]. Bossard, J. H. S. (1932). Residential propinquity as a factor in marriage selection. American Journal of Sociology, 38, pp. 219–224 [(E2s)]. Broadbent, D. E. (1958). Perception and Communication. London: Pergamon Press [(E2u)]. Canter, D. V. (1969, April). Attitudes and perception in architecture. Architectural Association Quarterly, pp. 24–31 [(E2s)]. Festinger, L., Schacter, S., and Back, K. (1950). Social Pressures Informal Groups. New York: Harper [(E2s)]. Gofman, E. (1959). The Presentation of Self in Everyday Life. New York: Doubleday [(E2u)]. Gullahorn, J. T. (1952). Distance and friendship as factors in the gross interaction matrix. Sociometry, 15, pp. 123–134 [(E2s)]. Herzberg, F., Mausner, B., and Snydermann, B. (1954). The Motivation to Work. New York: Wiley [(E2s)]. Ittelson, W. H., Proshansky, H. M., and Rivlin, L. G. (1972). Bedroom size and social interaction of the psychiatric ward. In Wohlwill, J. F. & Carson, D. H. (eds), Environment and the Social Sciences: Perspectives and Applications  (pp. 95–104). American Psychological Association. Washington https://doi.org/10.1037/10045-009 Maslow, A. H. (1962). Toward a Psychology of Being. London: D. Van Nostrand [(E2u)]. Moller, C. B. (1968). Architectural Environment and Our Mental Health Horizon. New York: Horizon Press [(E2s)]. Myrick, R., and Marx, B. S. (1968). An exploratory study of the relationship between high school building design and students’ learning. Ofce of Education Bureau of Research. US Department of Health, Education and Welfare [713 (E2u)]. Sommer, R. (1969). Personal Space: The Behavioral Basis of Design. Englewood Clifs, NJ: Prentice-Hall [(E2u)]. Warr, P. B. (1964). Proximity as a determinant of positive and negative sociometric choice. British Journal of Social and Clinical Psychology, 4, pp. 104–109 [(E2s)]. Wells, B. W. P. (1965). The psycho-social influence of building environment. Building Science, 1, pp. 153–165 [(E2s)].

4

Should we treat building users as subjects or objects? David Canter

Research motivations One of the attractions of the realm of architecture to the psychologist is possibly the fact that it entails an involvement with real-world decision making. At times, designing buildings must appear to psychologists like the experiments they dream about. By traditional research standards, the investigation runs forever, the number of subjects involved is infinite and the resources for building the apparatus magnificent. The possibility is provided of influencing and observing something that takes place, life-size, outside the constraints of the laboratory. The architectural psychologist is a person whose research motives are diferent from many of his academic colleagues. This very diference has probably led him to take an interest in problems other than those usually considered central to psychology. It should, therefore, be possible to learn much about the problems, hypotheses and results, of which architectural psychology at present exists, by examining the types of motivation which the research workers in this field have. The psychological architect (for want of a more apt term) is often attracted to this field by the distant possibility of producing buildings that influence people or can be used by people in the way the architect really intended. He is looking for categoric information that will enable him to produce the subjective and objective efects which he wants. One way in which architects sometimes express this need is by asking when psychologists will have their volume of ‘specifications’ ready. This invaluable document would contain lists of all the sizes, shapes and relationships possible in buildings, and next to each would be lists of responses that would be products by various sub-groups of the population (classified throughout according to SfB). Human motivations Architectural psychologists and psychological architects are both motivated by the desire to understand the way in which buildings afect people. Underlying this desire are assumptions concerned with what makes people do things. Both groups assume that something determines behaviour. The diferences of opinion that do exist with regard to this assumption cut across both architects and psychologists. These differences may be summarized by saying that some research workers seem to think DOI: 10.4324/9781003313052-5

36

David Canter

of people pulling and others think of them as being pushed. That is to say some theoretical formulations seem to develop from the idea that people are motivated by forces either deep inside them or completely external to them but always pushing them around and difcult to control. Other theories see people trying to get to grips with their environments, trying to understand and organize them, pulling them into an amenable shape. Both the ‘push’ and the ‘pull’ theorists are led into formulations of human behaviour which are directly influenced by their basic assumptions. They also take up radically diferent attitudes towards their experimental material, namely, people. These assumptions consequently have a far-reaching efect on the sort of problems they undertake to solve and the methods they see as appropriate for solving them. The attitudes of diferent environmental scientists (if you like) towards the people who provide experimental data for them is, I think, the most important concomitant of the various underlying assumptions. The ‘push’ theorists look at their subjects from the outside, so much so that it would be more accurate to call their subjects objects. (Chambers Dictionary’s definition: subject – a thing existing independently, the mind regarded as the thinking power [opp. to the object about which it thinks].) The research carried out by these theorists is peopled with unspeaking mechanisms, whose internal processes can only be inferred by physical actions. The objects of their research are not recognizable as people like the ones you and I know. Rather, they are complex primates impelled by basic (sometimes obscure) drives. They interact with one another by means of simple signals or expressions and on the basis of abstract spatial rules derived from innate needs. Needless to say, it is easier for these theorists to have animals as the objects of their studies rather than actual people, and it is certainly easier for them to derive their detailed theoretical formulations from work carried out with animals or from laboratory objects which are required to perform simple, and often meaningless, tasks. The ‘pull’ theorists, on the other hand, use people very much as the subjects of their experiments. These subjects are often remarkably verbose and are assumed to know more or less what they require of their environment. They are aware to some degree of the ways in which they interact with the physical environment and are constantly desiring to modify it and create it to be in their own image. These subjects, then, are active purposive organisms. They know what they want and will welcome any opportunity that helps them to get it. As a consequence, the best way of studying these subjects, and of finding out the proper environment with which to provide them, is to ask them. The diferences between these two approaches are best highlighted by the conflicts which exist between them. When the first group of theorists presents information about the objects of their research, the second group refuses to recognize these objects as people and insists that their humanity has been drained for them. Yet when the second group gives details of the responses of their subjects, the push men insist that these responses are inevitably invalid, as no subject can see himself objectively. Even if he could, he would be trying to infer hidden impulses behind the investigator’s actions and so would modify his own responses in order to push the investigator into reacting to him in a satisfying way.

Should we treat building users as subjects or objects?

37

Professional roles I said at the start that one of the main attractions of architectural psychology to members of both the contributing professions was its potential for afecting behaviour or for making decisions of importance to many people. This attraction may be considered a political awareness of, and an interest in, society at large. Consequently, the two sets of assumptions about people are likely to give rise to professional roles which have diferent political biases. If one thinks of a client or a building user as an object, then it is likely that one knows best and, as a consequence, the most appropriate approach to take is the authoritarian one. This is the approach which most professions take to the recipients of their services. I have not used this terminology to indicate an unfavourable attitude towards this approach. There are many occasions on which it is a very important part of a professional role to persuade or influence the client to behave in certain ways. Considering the user as a subject leads to what might be called a democratic approach to the professional role. The aim of professional contact when this role is taken often is to find out what the user or client requires and to set a process in motion which will enable him to get it. The job of the environmental scientist then becomes one of a communication channel. He has to find out as accurately as possible what it is that the people want, perhaps interpret this and pass it on to the designer. I hope it is obvious by now that the latter set of orientations underlies user requirement studies (which were, interestingly enough, developed by government agencies) and the former set gives rise to many academic outpourings, such as those on territoriality or on stimulus-response theories for design. Research problems The series of dichotomies described previously are, of course, theoretical approximations to what actually exists. Few individuals always sit on one side of the fence and never venture next door. In fact, the fence straggles across such difcult terrain that many people spend as much time, possibly unwittingly, on one side as on the other. I am no exception. Yet I think that an awareness of these distinctions helps us to locate central research problems and also shows us situations in which one approach can in fact contribute towards the solution of problems posed by the other. In order to illustrate the difculties which arise by staying always on one side, I propose to deal with two sets of studies I have carried out and to draw from the results of these an approach which combines both of the previous examples. Seat selection One theory that treats people very much as objects is the one that explains the distance which people keep from one another in terms of territoriality or personal space. This is thought (by Hall, 1966, for instance) to be a possibly innate mechanism and is certainly considered to be analogous to phenomena observed in studies of animals. Fields or bubbles can be postulated as existing around people, and certain

38

David Canter

behaviour can be interpreted as an attempt to reduce the possibility of a bubble being pierced or a field entered. Some fascinating research exploring these ideas has been carried out by Sommer (1969), and he has shown that in many specific situations, behaviour can be observed which is in accordance with the assumptions of a ‘bubble’, or a theory of ‘personal space’ as Sommer calls it. The advantages of this approach, if the details could only be finalized, to the authoritarian architect are, of course, tremendous. If people do indeed behave according to relatively simple rules that push people about below the level of awareness of those people, then it should be possible to design spaces that would take advantage of these rules. At Strathclyde, we were interested in one space in particular – a seminar room. We were interested in one interpersonal distance in particular, that between students and tutor. How could we design seminar rooms that could optimize this distance? Our interests in this developed from the observation that students will frequently sit at the back of a room when there are seats available at the front. We wondered whether this indicated the operation of ‘personal space’ and an attempt being made to optimize some distance from the lecturer or the back walls of the room. We examined this problem in two ways, by questionnaire and by observation. I will only deal with the observation results. The questionnaires corroborate these results very closely. A lecture room in the School of Architecture building was selected, and in the first case, the chairs in the room were arranged in a semi-circle as shown in Figure 4.1. Groups of eight students were asked to go to this lecture theatre to take part in an experiment. When they entered the room, a lecturer called them to him and gave them questionnaires. They were then told to sit down and complete the questionnaires. Note was made by the lecturer of the seats the students occupied. Two conditions were introduced into this situation. In one, the lecturer was 1/2 a metre or so from the students as shown in Figure 4.1. In the other case, the lecturer was some 3 metres from the front row. Three groups of students were used in each condition. Analyses were carried out to see whether or not the rows or angle of seats occupied in the two conditions were any diferent. No statistically significant diferences could be found. However, there was a clear pattern in both conditions. As illustrated in Figure 4.1, there is a tendency for students to sit to either side of the centre of the semi-circle. This suggests that no matter what the initial position of the tutor, people do not wish to sit central to this line of view. One of the difculties of the authoritarian approach to people is apparent in this result. People are not so easily manipulated but tend to interact with the surroundings. The activities in which they are engaged also confuses this interaction. The second case which we tried consisted of arranging the furniture so that it formed a rectangle as illustrated in Figure 4.2. The same two conditions were introduced. The distance of the tutor was varied. Figure 4.2 shows the seats chosen by the students when the tutor was near to the front row, and Figure 4.3 shows the position in which the students sat when he was further away. Analyses of variance show these two sets of seating preferences to be significantly diferent from one another.

Should we treat building users as subjects or objects?

39

1

1

2

3

1 3

12

11

6

8

2

+A

1 1

2

1

2 2

5

+B

A = ‘near’ position of lecturer B = ‘far’ position of lecturer

Figure 4.1 Total frequency of each seat being occupied by a student over four trials with tutor ‘near’ position and four trials with tutor at ‘far’ position.

It is apparent in the previous study that the position in which earlier students sit reduces the number of seat choices available for later students. Consequently, there is interaction between the users as well as between the users and their environment. One way of obviating this complication in this particular study was to give the students questionnaires with the seat pattern on them and ask them on which seat they would prefer to sit. This, in fact, was done, and Figures 2.4 and 2.5 facilitate the comparison of the two sets of observations. It is clear from these figures that there is a great similarity between the pattern of preferences in the two modes of data collection. The use of a questionnaire in this situation is really one stage towards asking the student what he thinks of the room, where he thinks he would like to sit in the room and the way in which he thinks the room, or his position in the room, might afect

40

David Canter

1

2

1

0.5m

1

1

2

3

2

1

1

1

2

2

1

1

1

1

Tutor’s position

Figure 4.2 Frequency of seat selection in rectangular furniture layout in three sessions when tutor was ‘near’.

him. This information about the student’s thoughts and feelings is very important, and without it, a full understanding of the mechanisms being studied could not be achieved. Using a questionnaire is one stage towards getting the user to set up hypotheses about the efect of the physical environment and to explain his interaction with it. In other words, it is getting the subject to take over a considerable portion of the role of the experimenter. All that the experimenter does is observe the subject’s estimates of the results of a theoretical formulation of which the experimenter has no clear description. This is really the essence of the democratic approach. The investigator does not presume to understand or to hypothesize the nature of the mechanisms by which the subject deals with the physical environment but, rather, to get the subject to show how satisfied he is with the functioning of the environment in which he is. With the results of this experiment on seat selection, it is very difcult to give an explanation in purely objective terms. An idea such as that of personal space is difcult to apply unless some modification is produced in the shape of that space by furniture layout. This modification would need to be quite complex if it were to explain the diferential efects of the tutor’s position with the diferent layouts. Any such modification would have to take into account the interpretation of the situation by the subject.

Should we treat building users as subjects or objects?

41

1

1

1

1

2

2

1

2

1

1

1

1

2

1

1

1

2

2

3m

Tutor’s position

Figure 4.3 Frequency of seat selection in rectangular furniture layout in three sessions when tutor was ‘far’.

If we need to take this interpretation into account, why not start with it? If we do, a much less complex hypothesis emerges, as follows. On entering the room, the furniture layout informs the subjects of the nature of the activity in which they are to take part. If they interpret it as a formal activity, then they attempt to optimize their distance from the tutor. If they see it as an informal activity, then distance from the tutor is irrelevant and other factors exert an influence. At present, we only have tangential evidence for the various assumptions within this argument. First, studies reported by Canter (1967), Canter and Wools (1970), and Wools (1970) in the present volume indicate the influence of furniture layout on the interpretation of what takes place within a room. Second, Sommer (1969) has shown that the amount of tutor-student interaction in which a student takes part relates quite closely to the position within which the student sits, in a rectangular layout of furniture. It seems likely, therefore, that the degree to which students decide to partake in a seminar relates to where they position themselves at the beginning of it and that this decision relates to a complex cognitive process based on trying to predict the type of seminar it will be.

42

David Canter

% of total number of subjects

50% 42.2% 40% 35.9% 30%

20%

10% 5% 1

2

3

4

5

6

7

Distance from lecturer (metres) Distance from lecturer Distance from lecturer Figure 4.4 Questionnaire results.

% of total number of subjects

50%

40% 39.2 30%

28.1%

20%

10% 5% 1

2

3

4

5

6

Distance from lecturer (metres) Distance from lecturer Distance from lecturer

Figure 4.5 Results from observations.

7

Should we treat building users as subjects or objects?

43

Individual diferences It follows that the two approaches to research I have described are complementary. Without observations of people as objects and of what they actually do, it is difcult to predict how they will behave in specific situations. On the other hand, without an understanding of how they feel and think about the situation and how they understand the workings of the environment, it is difcult for the research worker to develop a sophisticated-enough theory to deal with the majority of day-to-day contingencies. The investigator can enter this cycle at any point, but it is clear that an understanding of the subjects’ cognitions – and, hence, of their verbal responses to the physical environment – would be of great value. The great problem, however, in trying to study peoples’ reactions to the world about them is deciding which people to study. Does one select people in terms of their background, experience, age, sex, personality or some combination of all of these, or does one select people on the basis of diference in their response to the environment? At Strathclyde, we have been exploring the latter possibility because the little research which has been carried out with the former possibilities has been notoriously unsuccessful in facilitating the understanding of people’s satisfaction with, or interpretation of, the environment in which they work. It will be seen that what we have attempted to do is to select by some means a group of subjects who difer in a known way from others in the population in terms of their response to the physical environment. By studying these people, we hope to understand more fully the way the environment afects people and also to increase the efciency of our research by getting rid of one of the main sources of error variance in our studies, that due to individual diferences. To do this, we must treat our subjects as objects and try to classify the structure of their verbal responses. The aspect of response to the physical environment with which we have been concerned has been the degree to which the subject is able to distinguish between diferent parts of the physical environment and also the degree to which he is able to distinguish between diferent ways of describing the physical environment. We feel that the degree of sophistication which a person has in his concepts, descriptions and interpretations of the physical environment is likely to relate to his satisfaction with that environment, although this relationship will probably be very complex. Furthermore, we hypothesize that the sophistication of this diferentiation will relate to his reliability as a subject. We have attempted to measure this aspect of cognition in a number of ways. I think the idea behind it becomes clearer when one examines the techniques used for measuring it. One of the first techniques I used was taken from Sherif and Sherif (1967). Teachers were presented with 44 photographs of diferent types of classroom arrangement. These photographs were in fact models that we had prepared specially. An example is given in Figure 4.6. The teacher was required to sort the photographs into as many piles as he wanted in terms of the degree to which they provided a satisfactory teaching environment. Having done this, the teacher was then asked which of the piles contained classrooms that were acceptable to him and which contained

44

David Canter

Figure 4.6 One of the photographs used in the ‘own categories procedure’.

those that were unacceptable. This gave a number of cards remaining in a neutral category. The idea was that the less sophisticated a teacher was in discriminating between classrooms, the more likely he was to leave classrooms in the neutral category. Teachers with a high degree of discriminating ability and with a sophisticated way of dealing with classrooms would be able to decide about every classroom as to whether it was acceptable or unacceptable. A week or so after this task was completed by the teachers, I got them to complete repertory grids dealing with classrooms (see Bannister and Mair, 1968). They themselves generated the elements (or places) for the grid in answer to my queries about a range of places in which they could conceive of teaching. The constructs of the grid (or descriptions of places) were elicited by the method of triads and by discussion with the teachers about the similarities and diferences between the teaching places they had listed for me. The teacher was then asked to rank each of the teaching places on each of the descriptions of those places in turn. The particular property of this set of rankings in which I was interested was the similarity between the diferent descriptions which the teachers had used. In other words, did all the classrooms fall into the same rank order on each construct or in a completely diferent rank order? The more similar the rank orders, the greater the similarity between the diferent constructs in the way that they were used by the teacher. I would expect that a teacher who was able to diferentiate between many diferent aspects of the environment would place the rooms in a diferent order on each description. In order to measure this property of the individual, I calculated the correlation between every construct and every other construct. From this matrix of correlations, I added together the highest one for each construct and took the mean of these correlations as my measure of ability to discriminate between constructs. The higher the mean correlation, the less able is a person, or the less prepared, to distinguish between diferent constructs. I would expect there to be a relationship between this mean and the number of items in the neutral category found

Table 4.1 Group repertory grid used with first-year architecture students Next is a list of places in which you might study. You are to put this list in rank order on each of the aspects given. Name

Your kitchen at home

Your worktable in this school

A table in On a park A study the library bench on a at home sunny day

A study in An ideal the school study (like a staf place room)

______________________________________________________________

Should we treat building users as subjects or objects? 45

Privacy (rank 1 for a lot) Heating and ventilation (rank 1 for a lot) Control over room (rank 1 for a lot) Distance from lecturer (rank 1 for a little) Availability of references (rank 1 for a lot) Amount of distractions (rank 1 for a little) Availability of meals (rank 1 for easy to get) Overall preference (rank 1 for prefer most) Friendliness of environment (rank 1 for very friendly) Adequate furniture (rank 1 for very adequate) Lighting (rank 1 for good) Ease of communication with others (rank 1 for easy) Where do you usually study?

Your living Your room at bedroom home at home

46

David Canter

Number of photographs in the neutral category

earlier because both measure the same thing, sophistication of response to the physical environment. The number in the neutral category deals with discrimination of places, the mean correlation with discrimination of descriptions. Figure 4.7 shows the relationship found. It approaches a rank order correlation of unity. Thus, even with the very small number of subjects involved, there is support for the suggestion that sophistication of response to the physical environment is an enduring aspect of the individual which is consistent over time, even when measured by two quite different techniques. A further check of this consistency was provided in another study which we carried out with our first-year students. The students completed the standard repertory grid shown in Table 4.1. This deals with their ranking of the various places which they might use for studying on a series of descriptions of those places. Instead of looking at the average correlation as we had done previously, we completed a principal components analysis for each grid for each student, using Patrick Slater’s programme (1964). It occurred to me that a more sophisticated version of the average correlation is probably provided by the percentage of the variance taken up by the principal component in the analysis. Bannister and Mair (1968) refers to this percentage as an estimate of an individual’s ‘cognitive complexity’. His definition of cognitive complexity relates closely to the previous description of sophistication (see Canter et al., 1968, for a further discussion of this point). Two months prior to the completion of the grids, the students were asked to rate a number of drawings on ten bipolar adjectival scales. The assumption was that the more discriminating the students, the wider range of ratings they would give. A student who could not distinguish well between various drawings would say they were all either good or bad. In order to measure this degree of ‘spread’ of the ratings, we used a technique

20

10

0 00

70 80 90 Mean correlation on repertory grid

Figure 4.7 Relationship between ‘own categories procedure’ and ‘cognitive complexity’.

Should we treat building users as subjects or objects?

47

developed by M’Comisky (see Canter et al., 1968) which is essentially a nonparametric measure of variance. A test of the consistency of the variable of cognitive complexity is provided by the correlation between the percentage of the variance taken up by the principal component and the ‘spread’ score. In our case, with 28 subjects and a gap of over two months, including the Christmas vacation, between the administration of the two quite separate tests, we obtained a product-moment correlation of 0.40. This is significant at the 2.5% level. Whatever this attribute of cognitive complexity is, it would seem to be moderately consistent and relevant to a wide range of ways or recording responses to the physical environment. Having established some consistent variable which relates to a person’s way of thinking about the physical world, in other words, having gone some way towards categorizing the ‘objects’ from which we elicit ‘subjective’ reactions, the next stage is to see in what way this variable relates to a person’s satisfaction with certain aspects of that world. In order to do this, we used the data we had collected from the repertory grids of the students mentioned earlier. What we did was to estimate the degree of satisfaction of the student with his study-bedroom. On the grids we had given the students, one of the elements to be ranked was an ideal study place. To find out how satisfied the student was with the place in which he studied at the moment, we needed to find out the degree to which it related to his ideal. The statistics of doing this are a little intricate, but all that it amounts to is the comparison of the profile of rankings on the study-bedroom with the profile of rankings on the ideal. We took it that the more similar these profiles were, the more satisfied students were with their study-bedroom. Figure 4.8 shows the scattergram for the relationship between the student’s spread scores calculated as described earlier against degree of similarity of the two profiles just mentioned. It will be seen from this figure that as the student’s spread scores increase, that is, as his degree of ability to discriminate increases, so the variance between the students decreases. This idea is represented schematically by means of the triangle drawn over the points in Figure 4.8. The diference in variance between the top half and the bottom half of Figure 4.8 is significant at the 5% level when tested with an F test. The means for the two halves are not significantly diferent. A feasible explanation for Figure 4.8 would seem to me to be along the following lines: the more able an individual is to discriminate between diferent aspects of the environment, the more likely is he to have a number of diferent ways of assessing the environment he inhabits at the moment. Consequently, when he makes a judgement about, say, his study-bedroom, he brings to bear many diferent viewpoints. This leads to his decision about the room being closer to the average for the population. On the other hand, a person with low abilities to discriminate would only use one or two aspects on which to base his judgement and, consequently, would be more likely to have an extreme judgement than his more cognitively complex fellows. If this relationship was validated by future studies, its value would be quite considerable in practical terms because it provides the possibility of selecting individuals whose response is more in accord with that of their peers, is more likely to

48

David Canter 70 60

‘Spread’ scores

50

40

30

20

10

0 40

60

80

100

120

140

160

Closeness of actual study place to ideal

Figure 4.8 Relationship between ‘cognitive complexity’ and ‘satisfaction’.

be typical of the population at large and is probably more consistent or reliable than average. Our eventual hope is that we will be able to ‘objectively’ select individuals that will give us valid subjective responses and further that we will need fewer subjects. This relationship is probably of more value at present for theoretical reasons in that it is the beginning of an insight into the way in which people deal with the physical world. Subsequent studies have helped clarify the relationship between cognitive complexity and satisfaction with the physical environment. They have shown that in some cases, the relationship is a linear one such that more cognitively complex people are more satisfied with a particular aspect. They have also indicated that the building form itself possibly influences the level of complexity of the user with regard to responses to particular aspects of the environment. Subjects or objects? In conclusion, it can be said that an understanding of the interaction of people and buildings and a furthering of the various roles of the environmental scientist cannot be fostered by treating people solely as objects. Information concerning their own

Should we treat building users as subjects or objects?

49

thoughts and feelings about the environment is essential whether the designer finds himself in an authoritarian or a democratic relationship with his client. Further, it can be suggested that architectural psychologists and designers should not be afraid of looking at clients and building users from an external viewpoint. This is the case particularly with regard to the selection of respondents. It is only by building up an understanding of the interaction between the set of assumptions which deal with people as subjects (who experience) and that set which deals with them as objects (who at) that we can hope to turn architecture into a science with firm roots in psychology and to turn psychology into a science which is of relevance to the everyday world. References Bannister, D., and Mair, J. M. M. (1968). The Evaluation of Personal Constructs. London: Academic Press. Canter, D. (1967). The measurement of appropriateness in buildings. Transactions of the Bartlett Society, 6, pp. 43–60. Canter, D., M’Comisky, J., and Johnson, J. (1968). Familiarity with architectural concepts and academic achievement of architecture students. Perceptual and Motor Skills, 28, pp. 871–874. Canter, D., & Wools, R. (1970). A technique for the subjective appraisal of buildings. Building Science, 5, pp. 187–198. Hall, E. T. (1966). The Hidden Dimension. New York: Doubleday. Sherif, M., and Sherif, C. W. (1967). The own categories procedure in attitudes research. In Fishbein, M. (ed) Reading in Attitude Theory and Measurement. New York: Wiley. Slater, P. (1964). The Principal Components of a Repertory Grid: An Account of the Computing Programme Grid Analysis 2 with Some Notes on Grid Analysis 3. Vincent Andrews. Sommer, R. (1969). Personal Space: The Behavioral Basis of Design. New York: Prentice Hall. Wools, R. (1970). The assessment of room friendliness. In Canter, D. (ed) Architectural Psychology: Proceedings of the Conference Held at Dalandhui, University of Strathclyde, 28 February–2 March 1969. London: RIBA Publications.

5

Judgements of people and their rooms *

D. Canter , S. West and R. Wools†

Many studies have indicated that a wide variety of peripheral attributes of people influence judgements made of them. For instance, glasses (Thornton, 1944) or lipstick (McKeachie, 1952) or the type of clothes worn (Gibbins, 1969) all relate to judged attributes of the wearer. Indeed, early studies in person perception (e.g., Thorndike, 1920; Asch, 1946) showed that seemingly independent judgements did in fact relate to one another. More recently, sociological studies (Laumann and House, 1970) have shown that living room furnishings can predict the social attributes of their owners. It might be expected, then, that the physical environment also plays a role in person perception. The surrounding environment forms an enduring and central part of the impression of a person (considerable efort and money is frequently used to create the right ‘atmosphere’ or best ‘image’, whether in private rooms or directors’ ofces). In dramaturgical terms, the architectural setting forms part of a front, the expressive equipment used by the performer during his performance (Gofman, 1959). A number of studies have indicated that rooms act as a context for judgements made within them. Rosenthal (1966) found relationships between judgements of people and their surroundings; the experimenter was perceived as more ‘statusful’ in an untidy laboratory. Kasmar et al. (1968), however, found no efect of diferent room furnishings on judgements of psychiatrists by patients. The Maslow and Mintz (1956) study of the efect of room context on perceptual judgements showed faces were judged diferently in a ‘beautiful’ room from the way they were judged in an ‘ugly’ room. Canter and Wools (1970) showed a similar efect when subjects judged drawings of rooms. The previous considerations gave rise to the general hypothesis that relationships will be obtained between ratings of people and the rooms with which they are associated. The most parsimonious inference rule which might underline these relationships was assumed to be the one of similarity between a person and their room. The hypothesis was investigated using increasingly complex stimuli. Experiment I In this experiment, line-drawings of rooms containing outlined seated figures were used in order to produce a wide range of standardized room variations. As factor analytic studies of environmental stimuli (Vielhauer, 1966; Hershberger, 1969; DOI: 10.4324/9781003313052-6

Judgements of people and their rooms

51

Canter, 1969; Canter and Wools, 1970) had indicated, the four commonly occurring major dimensions of friendliness, harmony, activity and formality were used to rate the rooms in this experiment. Method

1 Sixty-seven students from many disciplines, but none from architecture or design, rated 15 monochromatic line-drawings of rooms. The drawings were selected to give a wide range on the four dimensions earlier. Ratings were made from one to seven on four bipolar adjectival pairs for each dimension (e.g., for Friendliness, friendly – unfriendly, unwelcoming – welcoming; for Harmony, unstable – stable, discordant – harmonious; for Activity, calm – lively, active – passive; and for Formality, statusless – statusful, formal – informal). 2 A further 24 non-architecture students assessed the same 15 line-drawings, rating the person shown in outline from one to seven on four item scales representing Osgood’s dimensions, Evaluation (e.g., good – bad, pleasant – unpleasant), Potency (e.g., strong – weak, brave – cowardly) and Activity (e.g., active – passive, tense – relaxed). For scoring, each set of items was formed into a Guttman scale by dichotomization (cf. Wools, 1971). Summing the ratings over the four bipolar scales gave scores for each dimension. Means were calculated from the scores for each drawing on the seven dimensions for all subjects in each group. It was thus possible to compare judgements of the rooms with judgements of the person in the room by intercorrelating the two sets of mean scores. Results

These correlations lend support to the general hypothesis of a similarity relation between ratings of rooms and people associated with them. The activity dimensions are the only ones which do not give significant correlations, although three of the four bipolar adjectives were the same for the two groups of subjects. The lack of correlation for the activity dimensions implies that the subject groups were actually making diferent ratings and not ignoring instructions by making similar judgements in both cases. If the activity rating of rooms had simply been unreliable or random, it would have been impossible to isolate an activity dimension Table 5.1 Product-moment correlations between mean ‘room’ judgements and mean ‘person’ judgements in Expt. I  

Room: Friendliness

Harmony

Activity

Formality

Person: Evaluation Potency Activity

0.90* 0.61* 0.02

0.84* 0.63 0.13

0.55 0.23 0.05

0.89* 0.62* 0.05

d.f. = 13. *

P < 0.01.

52

D. Canter, S. West and R. Wools

or, from it, produce a Guttman scale with an acceptable level of reproducibility. If activity ratings of the people in the drawings had been unreliable or random, the scores would not have correlated significantly (0.66) with the potency dimension as they did. A further implication is that these rooms give little indication of the ‘activity’ of their occupier and vice versa. These results show that the general similarity hypothesis is not supported for all dimensions. Experiment II To test the previous findings with more realistic architectural stimuli, colour slides of real rooms were used. The slides were of ofces in Edinburgh University, and although there were no people in them, there was considerable evidence of habitual use by one individual (e.g., cofee mugs, briefcases, etc.). In order to examine the possible efects of training and professional orientation on judgements, 41 architecture and 32 psychology students were used as subjects. They were presented with 24 colour slides of rooms unfamiliar to them. All slides were rated for both room and user using a split-half design such that each subject viewed and rated each one only once. The users of the rooms were rated on five bipolar adjectival scales representing the dimensions of Evaluation (e.g., kind – cruel), Potency (e.g., bold – timid) and Activity (e.g., fast – slow). For the rating of the rooms, two sets of ten adjectives were used, one set (e.g., friendly – unfriendly, happy – sad) being a development of the ‘friendliness’ dimension used in Expt. I (Canter and Wools, 1970). The second set was drawn from the main dimension, ‘adequacy’ (e.g., comfortable – uncomfortable, adequate – inadequate), found by Canter (1971) in his study of school buildings. Both sets had been found highly reliable in ratings of buildings and formed Guttman scales with high coefcients of reproducibility. Results

No significant diferences were found between the means for the two groups using t tests, variances of the two groups using F tests or between the correlation coefcients using z̄ the normally distributed diference between standardized correlation coefcients. As with the line-drawings, there is a significant correlation between ratings of evaluation of the users and friendliness of the rooms. However, potency of the user Table 5.2 Product-moment correlations between mean room judgements and mean user judgement for the two subject groups in Expt. II  

Architects users:

Rooms: Friendliness Adequacy

Evaluation 0.70* 0.78*

d.f. = 22. * P < 0.01.

Psychologists users: Potency 0.07 −0.10

Activity 0.47 0.50*

Evaluation 0.66* 0.63*

Potency −0.36 −0.31

Activity 0.39 0.35

Judgements of people and their rooms

53

shows no significant correlation with the room ratings. Activity does show some small but significant correlations. These two studies together indicate that the central evaluation dimension correlates in either mode but that the correlations of other variables are mode specific. Experiment III As Expts. I and II showed that ratings of central dimensions of rooms and people were associated, it was decided to examine the efects of the room background on judgements of people in photographs. Besides enabling us to examine the contribution of an actual photograph of a person to the general room efect, this experiment also has everyday relevance in that the mass media frequently show certain people against particular backgrounds. As these visual representations may help influence public opinion, the environmental context may also be important. A further point is that frequent association of certain people and environments is possibly one way in which attitudes develop towards those environments. Expt. III was a first attempt to study these problems experimentally. Method

Through the assistance of a local newspaper, nine photographs were prepared from three head-and-shoulder photographs of people and three of rooms such that each person appeared in the same position in each room. The three people were all smiling, well-dressed men wearing glasses, two in their 40s, the other looking rather older. Room A was a Victorian window overlooking an industrial scene, room B was the library of the Q.E.2 and room C was a large IBM ofce with filing cabinets and luminous ceiling. Thirty architecture students rated the person in each of the composite photographs on the three five-item evaluation, potency and activity scales used in Expt. II. As the photographic superimposition became clear to the subjects at the latest upon seeing the fourth photograph, the presentation order was varied for each of three groups of ten subjects. This ensured all nine photographs were seen by ten naive subjects and provided a three-variable, factorially designed experiment, each of the variables of order of presentation, person and room existing at three levels. The photographs were converted into black-and-white slides for presentation. Results

To see if any room efect existed, t tests were carried out between the means for each combination of backgrounds for each person. As shown in Table 5.3, nine of a possible 27 t tests were significant. An analysis of variance showed that although the people were rated diferently at the 0.01% level for all three dimensions, the room efect was only present on the potency dimension at the 5% level. The efect of order was present on the activity dimension also at the 5% level. There was also a significant person × order interaction on all three dimensions.

54

D. Canter, S. West and R. Wools

Table 5.3 Semantic diferential means showing significance of diference between means  

Person 1

(a ) Means for rooms A and B A E 18.6 A 21.1 P 22.0

Person 2

Person 3

B 16.7 19.9 20.1

A 19.1 19.9 16.7

B 22.2 16.6* 16.2

A 19.0 21.2 21.5

B 21.8* 20.0 21.5

(b) Means for rooms A and C A C E 18.6 17.4 A 21.1 20.8 P 22.0 23.0

A 19.1 19.9 16.7

C 19.1 16.7* 18.1

A 19.0 21.2 21.5

C 18.1 15.9*** 18.1*

(c) Means for rooms B and C B C E 16.7 17.4 A 19.9 20.8 P 20.1 23.0

B 22.2 16.6 16.2

C 19.1* 16.7 18.1

B 21.8 20.0 21.5

C 18.1* 15.9*** 18.1*

* P < 0.05. *** P < 0.001.

The picture presented by these analyses and by further analyses dealing only with first, second or third ratings for each subject is a very complex one. They indicate that the room efect is masked by interactions between person and order efects. However, it does appear that there is a room efect and that this is most marked for person 3. Furthermore, the library/ofce comparison produced the most consistent results on all three dimensions. General discussion The three experiments taken together strongly support the general hypothesis of relationships existing between judgements of people and their rooms. There is also an indication that the inference rule of similarity between room and user underlies these relationships. Expt. III suggests that this inference rule can lead to what may be regarded as a ‘room efect’, whereby judgements of people are modified by the room in which the person is seen to be. The preparation of this paper was partly assisted by an award from the Architects Registration Council of the United Kingdom. Notes * Now at Department of Psychology, University of Surrey, Guildford, Surrey. † Now with Alex Gordon and Partners, London.

Judgements of people and their rooms

55

References Asch, S. E. (1946). Forming impressions of personality. The Journal of Abnormal and Social Psychology, 41, pp. 258–290. Canter, D. V. (1969). An inter-group comparison of connotative dimensions in architecture. Environment and Behavior, 1, pp. 38–48. Canter, D. V. (1971). Scales for the Evaluation of Buildings. Glasgow: University of Stratchclyde, mimeograph. Canter, D. V., and Wools, R. (1970). A technique for the subjective appraisal of buildings. Building Science, 5, pp. 187–198. Gibbins, K. (1969). Communication aspects of women’s clothes and their relation to fashionability. British Journal of Social and Clinical Psycholog, 9, pp. 301–312. Gofman, E. (1959). The Presentation of Self in Everyday Life. New York: Doubleday Anchor Books. Hershberger, R. (1969). A study of the meaning of architecture. Environmental Design Research Association, 1st Annual Conference, Chapel Hill, NC. Kasmar, J. V., Grifn, W. V., and Mauritzen, J. H. (1968). Efect of environmental surroundings on outpatients’ mood and perception of psychiatrists. Journal of Consulting and Clinical Psychology, 32, pp. 223–226. Laumann, E. O., and House, J. S. (1970). Livingroom styles and social attributes: The patterning of material artifacts in a modern urban community. Sociology and Social Research, 54, pp. 321–342. Maslow, A. H., and Mintz, N. L. (1956). Efects of aesthetic surroundings. I. Initial efects of three aesthetic conditions upon perceiving ‘energy’ and ‘well-being’ in faces. The Journal of Psychology, 41, pp. 247–254. McKeachie, W. J. (1952). Lipstick as a determiner of first impressions of personality: An experiment for the general psychology course. The Journal of Social Psychology, 36, pp. 241–244. Rosenthal, R. (1966). Experimenter Efects in Behavioural Research. New York: AppletonCentury-Crofts. Thorndike, E. L. (1920). A constant error in psychological rating. Journal of Applied Psychology, 4, pp. 25–29. Thornton, G. R. (1944). The efect of wearing glasses upon judgements of personality traits of persons seen briefly. Journal of Applied Psychology, 28, pp. 203–207. Vielhauer, J. A. (1966). The development of a semantic scale for the description of the physical environment. Dessert Abstract, 26, p. 4821. Wools, R. M. (1971). The Subjective Appraisal of Buildings (Unpublished Ph.D. thesis, University of Strathclyde).

6

Distance estimation in cities David Canter and Stephen K. Tagg

AUTHORS’ NOTE: This research is supported by the Social Science Research Council of Great Britain. A property of the physical environment of distinct psychological importance is the fact that the environment completely surrounds us. Thus, it is not possible for us to experience or perceive all of it at any one instant. We can only turn our attention to discrete aspects of the environment at successive points in time. However, in order for our behaviour to be appropriate, efective or adequate in relation to the physical environment, it is necessary for it to proceed in a continuous fashion. To explain the way in which this discrete experience can produce continuous interaction, it is necessary to postulate some representational process on the part of the individual. This ‘representation’ must amalgamate experience into a form which links discontinuities in perception and allows extrapolations to facilitate preparation for future action. In the case of an individual coping with a city, it is further necessary to assume that the postulated amalgamation of experiences is somehow summarized. This summary consists of a pattern or structure, the resultant efect of which is to organize the ‘representations’ of the experiences of the city and their implications. Note that the term ‘representation’ is in quotation marks. This is because there is nothing within our description of it to say that it must have any direct relationship to the usual methods of representing cities. Indeed, because our assumption of a representational process derives from the relationships between present experience and continuing behaviour, it seems likely that representations other than visual ones (akin to those used by geographers – ‘maps’) are involved. These concepts derive from the work of Bartlett (1932) and Lynch (1960). More recently, Stea (1969) and Ittelson (1973) have argued in some detail for crucial diferences between the perception of environments and the perception of objects. Stea, in particular, has asked, ‘Can urban imagery be quantified?’ For the present paper, we have moved some way to answering this question. We have examined some aspects of the metric properties of urban ‘images’. Once it is accepted that internal representation is central to our abilities to cope with cities, then the question of the veracity of this internal process, the properties it handles accurately and those it distorts becomes paramount. One of the most crucial properties is that of scale, or distance. This is crucial because by its very nature, DOI: 10.4324/9781003313052-7

Distance estimation in cities

57

any representation of the city environment cannot be of the same physical size. Yet many of the qualities of any representation relate to its scale (i.e., the ratio between distances on the representation and actual distances). Indeed, for geographic surveys, we may consider an inch-to-the-mile map appropriate for dealing with cities but not for countries. Are there any analogous psychologically appropriate metrics for cities? Research in environmental distance estimation has already revealed a number of consistencies. There is thus the likelihood that such inviting but formerly speculative concepts as a city’s image (Lynch, 1960; Appleyard, 1970) now may be examined precisely and numerically by the use of distance estimation. For example, Ekman has carried out a number of studies (summarized by Lundberg et al., 1972). The central aim of these was to show that a square root relationship held between subjective distance and emotional involvement for capitals around the world – a relationship such that intensity of emotional involvement was inversely related to the square root of subjective distance. Such a relationship has been demonstrated in a number of studies (Dornic, 1967; Bratfisch, 1969), but unlike the studies from the same laboratories on subjective temporal distance (Lundberg, forthcoming), no evidence seems to have been published to show that this relationship provides the best or most parsimonious account of the data. The reasons for the use of ‘subjective’ as opposed to ‘objective’ distance estimations, Lundberg ascribes to Ekman’s (1961) suggestion that psychophysical relationships should be explored on the ‘purely psychological level’. However, in the case in which comparisons are being made with emotional involvement, it would seem that (in English anyway) both judgements may be really synonymous for the same concept. How subjectively far away a place feels can only mean the degree to which one feels linked with it or concerned with activities which take place within it, in other words, how emotionally involved one is with it. Thus, finding a relationship between these two statements only helps to show they do indeed have a similar meaning. It does not tell us much about the relationship between psychological and physical variables. The linear relationship between subjective distance and geographical distance which Lundberg et al. (1972) reports does help to provide some information in this direction. However, in this case, attitudes towards the various countries and knowledge of them clearly play a key role. Lundberg’s findings possibly contribute more to geopolitics than psychophysics. Evidence in favour of the weaknesses of the Ekman and Lundberg studies is given by Stanley (1968, 1971), who, when using judgements of estimated objective distance, found no evidence for the inverse square root relationship with emotional involvement. Cadwallader (1973) has tackled the problem more directly by asking people to estimate the distance from their home to each of 30 cities in the Los Angeles basin. To examine the importance of the type of distance estimation, he had his subjects use two methods. Using absolute mileage estimations, he obtained the regression Y = 1.09X + 5 (r = 0.96), whereas using ‘magnitude estimation’, the regression was Y = 0.59X + 0.46 (r = 0.94), where Y was the mean of the estimated and X the actual distances. Golledge and Zannaras (1973) also showed that high correlations could be obtained between ‘perceived’ and actual distances. Judgements were made of distances

58

David Canter and Stephen K. Tagg

between the university campus and a series of locations along a ‘major north-south artery’. However, they found newcomers less accurate in their distance estimates than established residents. They also report that there was ‘an underestimation of distance moving away from the down town area’. In other words, distances into the town were perceived as relatively longer than distances out of it. Briggs (1972) followed up this finding and found some support for it. This contrasts markedly with the study by Lee (1970) in which students estimated the walking distance between pairs of points in Dundee. The pairs were so selected that some of the distances were inward and some outward. Lee found that there was an overestimation of outward journeys. It is intriguing that this supports Lee’s essentially cognitive orientation. He sees distortions in estimation as brought about by factors such as desirability of the journey destination. Lee argues that the centre of Dundee was more attractive to his students than its periphery. The Golledge and Briggs studies, however, led support to the learning theory stance which they take. They see experience of the journey, with increasing trafc towards the downtown area, as the key determinants of distortions in perception. Yet Lee and Golledge are both dealing with quite diferent city forms: the grid pattern of northern Columbus contrasting with the contour following plan of Dundee. Indeed, in the latter case, there will be a slight tendency for the outward distances to be uphill away from the Firth. The final environmental study we must refer to is that of Lowrey (1970). He examined in detail the scalar properties of distance estimations to known urban facilities. He came to the conclusion that people were indeed able to estimate distances in a manner akin to ratio scaling. These studies have covered a wide range of distances, from hundreds of yards to hundreds of miles, and they have also included a range of types of distance estimate. But in particular, there seems to be some confusion in the literature as to whether actual routes are estimated or direct (‘crow flight’) distances. At the large country or city scale of Lundberg or Cadwallader, crow flight distances are seemingly the only meaningful ones to ask subjects to estimate. Route distances would be confused by the range (or lack) of routes available. But at the city scale, it would seem that actual route distances would weigh the estimate in terms of experiences directly related to following the route itself, details such as time and direction having a considerable impact. Crow flight estimates may be drawn more directly from some abstracted representational process which the person has developed. Admittedly, this process may be built upon many experiences, such as travelling about the city, but crow flight estimates should enable us to deal with the general residue of this experience, rather than its particular manifestations. For our studies, then, we have concentrated upon crow flight distance. However, in dealing with direct distances, it is easy to confuse distance estimation with the psychophysical studies of length estimation. Both Ittelson (1973) and Stea (1969) have emphasized the diference between the latter in which a target object is present for perception and the former in which the object cannot immediately be perceived. Furthermore, even the longest lengths judged in psychophysical studies, around 700 metres (Kuroda, 1971), were from a seated position along a straight

Distance estimation in cities

59

road. More typical are the studies by Kunnapas (1968) with lengths up to four metres being judged under highly controlled and often excessively distorted conditions. In these situations, many of the depth cues, such as size of the image on the retina or familiarity of the object perceived, have been shown to have a marked efect on judgements of their distance from the observer (Kunnapas, 1968; Gogel, 1969). In the same vein, Stevens has developed the Weber-Fechner functions to show a power relationship between psychological judgements and physical measurements. This power function seems to vary according to the sensory attribute being judged (Stevens, 1957), but there is some discussion as to what its value is for distance. In his 1957 paper, Stevens indicates that it might take on a variety of values for estimates of distance under diferent conditions, being close to one for physical length (Stevens and Galanter, 1957). Others have argued (Künnapas, 1960; Teghtsoonian, 1971) that the exponent depends upon the range of the stimuli being judged, and when this is taken into account, an exponent of 0.03 copes with most data. But here again, the lengths judged are always immediately available to the senses, often in the region of one to three metres, and thus are qualitatively diferent from distance estimation in cities. The question is thus raised whether the same relationships will hold for ‘cognitive’ estimates as for ‘perceptual’ ones. One further value of studying distance estimations is that they are a significant component of our environmental vocabulary. The study is thus especially valuable if the units of estimation are drawn from the commonly available language of yards or miles (and possibly minutes and hours). Whether it be as a basis of signposting or word-of-mouth comment, distances frequently occur as environmental descriptors with implications for behaviour. The fact that there is a publicly available criterion against which to validate these descriptors does not necessarily imply that they are used exactly in accord with that criterion. The study by Cadwallader suggests that they are not. In summary, then, the studies to date have shown that consistent (if consistently inaccurate) estimates of distances between environmental locations can be made. These estimates might well be of a diferent kind from those normally examined by psychophysicists, and they are also likely to be sensitive to aspects of environmental experience. They thus have great potential for contributing to the understanding of the representations of cities which we all develop. Of the questions requiring future study which emerge from this survey of research, the one we have identified as central is this: What general statements can we make about the relationship between estimated and actual distance? In other words, is this relationship constant across cities and cultures? A corollary of this question will also be examined: What distortions in estimations can be related to the form of any given city? Method In all, 11 distinct studies were carried out in seven cities in five diferent countries. A summary of these studies is presented in Table 6.1. In all cases, the points were selected by a resident to give a reasonable distribution over the city and to relate,

60

David Canter and Stephen K. Tagg

in general, to reasonably well-known, identifiable locations. The type of location depended on the city, but they tended to be landmarks and stations. However, some attempt was made to introduce one or two less well-known points for later comparisons. Subjects were presented with the names of points in the form of a matrix: they had one set of names across the top and the other set down the side. For studies 1 to 6, the two sets of names were the same; for studies 7 to 11, they were diferent.1 Thus, for the first 6 studies, all interpoint distances were obtained but not for the remainder. For the reasons discussed earlier, it was decided to obtain estimates in a form common in daily speech. Each subject estimated the direct (as the crow flies) distance between every pair of places to the nearest half mile (or half kilometre). He was told to guess if he was not sure. He thus completed every cell in the matrix for the latter five studies and one-half of the matrix for the former six. After their responses had been collected, it was apparent from subjects’ comments that their estimates did not usually come from any clear ‘mental map’. They did, however, frequently express a reasonable confidence in them and did not appear to find the task too onerous. For the purpose of analysis, arithmetic means for each group were calculated for each of the interplace distance estimates. The relationships between estimates and distance Reference to Table 6.1 shows that all the mean estimates are greater than the actual mean distances; all further comments must be taken against this background of overestimation. In order to answer our central question, linear regressions were calculated between all the mean estimates and the actual distances. For instance, with 5 points, there are ten estimates; with 7, there are 21. Correlations for these regressions were also calculated to see if the assumption of linearity was acceptable. To further test the possibility derived from the work of Stevens (1957) that a nonlinear power function is involved, regression of the log mean against the log actual distance and their correlations were calculated. A summary of these results is presented in Table 6.2. The results in that table have been arranged in order of regression slope so that the studies can be seen in the sequence from those which give a decrease in overestimation against length (i.e., the A-coefcient is less than 1.00) to those which give a general overestimation of distances. However, before referring to the implications of this sequence, two general points emerge from Table 6.2. First, in general, these subjects, when their responses are averaged, do give answers which correlate highly with the actual distance. Yet because few of the slopes are 1.00, there is a tendency for their estimates to have a consistent bias (or misestimation) in proportion to the actual distance. (The practical implications of this in such areas of decision making as trafc engineering could be considerable.) The one group which stands out in respect of correlations is that of study 1. This is the only group of ‘real’ people (i.e., nonstudents accosted in the street and asked to answer some questions). They would be expected to come from a background of experience and training very diferent from the students. Such a discrepancy between their

Table 6.1 Summary of 11 distance estimation studies City

Study number

Number of subjects

Number of points

Number of distance estimates

Mean actual distance

Standard deviation actual

Mean estimated distance

Standard deviation estimate

Subjects

Pedestrians in Central Glasgow Architecture undergraduates

1

50

5

10

2.67

0.86

3.97

0.76

Glasgow (inner distance) Glasgow (outer distance) Edinburgh

2

38

7

21

1.36

2.33

4.18

2.89

3

38

7

21

6.22

2.58

7.53

2.07

Architecture undergraduates

4

35

7

21

1.28

0.60

1.92

0.88

Heidelberg

5

24

7

21a

2.50

1.21

3.15

1.02

London

6

25

5

10

3.96

2.24

4.77

1.74

Sydneyb

7

23

12

36

13.61

6.53

17.19

7.70

Tokyo

8

64

11

30a

6.89

3.39

10.59

5.00

Tokyoc

9

41

13

30a

6.69

2.78

11.77

4.63

Nagoyac

10

18

10

40a

5.33

2.36

7.51

3.57

Nagoyac

11

20

12

35a

5.16

2.35

6.46

2.80

Art school undergraduates Psychology undergraduates Architecture undergraduates Architecture undergraduates Architecture undergraduates Technical college undergraduates Architecture undergraduates Technical college undergraduates

a. b.

Estimates in kilometres (all others in miles). We are grateful to Ross Thorne for conducting this study. We are grateful to Dr. Inui and Dr. Miyata for conducting these studies.

61

c.

Distance estimation in cities

Glasgow

62

David Canter and Stephen K. Tagg

Table 6.2 Regression of mean estimates against actual distance (X = actual distance) City

Study Linear regression Correlation for LogLog regression (Log Correlation for Log number (Y = AX + B) Y = AX + B Y = ALogX + B) Y = ALogX + B

Glasgow 1 Heidelberg 5 London 6 Glasgow 3 Glasgow 2 Nagoya 11 Sydney 7 Nagoya 10 Edinburgh 4 Tokyo 8 Tokyo 9

0.45 + 2.72 0.70X + 1.41 0.74X + 1.84 0.76X + 2.81 1.01X + 1.28 1.07X + 0.95 1.12X + 1.19 1.27X + 0.77 1.29X + 0.27 1.39X + 1.03 1.54X + 1.46

0.45 0.83 0.95 0.95 0.93 0.90 0.95 0.84 0.88 0.94 0.93

0.35X + 0.45 0.53X + 0.30 0.73X + 0.36 0.59X + 0.42 0.79X + 0.26 0.92X + 0.16 0.89X + 0.03 1.04X + 0.13 1.03X + 0.17 0.93X + 0.24 0.82X + 0.39

0.52 0.87 0.97 0.92 0.94 0.91 0.97 0.95 0.93 0.95 0.92

averaged ability to estimate distances and that of the undergraduates poses a very important and potentially fruitful question for future investigation. The second point emerges from a comparison of the two regression equations and their correlations. From this comparison, although there is a tendency for the LogLog regression to account for a little more of the common variance, this is certainly not consistent. In the light of this, our future discussion will deal only with the linear relationships. There is one practical value of the high correlations. They do support the assumption that the ratio of estimate to actual distance is a reasonably constant one for any set of distances. Thus, we may consider, in general, average misestimation without the fear that these average ratios (say, of actual divided by the estimate or the slope of the regression line) will relate to the actual magnitude, or scale, of the distances involved. Accepting this, we may then question the sequence of the studies in Table 6.2 to see what hypotheses they suggest for such notable diferences between cities. Regressions and Pregnanz The studies fall into three groups in relation to their linear regressions. The first group covers the Glasgow, Heidelberg and London studies. In four cases, the Acoefcients are less than one and the B-values are greater than one. The general trend for estimating distance, which emerges from these relationships, would seem to be to add a constant distance and then to underestimate the actual distances. This produces overestimation for shorter distances. However, once the actual distance is above 6 or 7 miles, underestimation is the tendency. In the main, these cities appear to have some property which leads the respondents to think of longer distances being, on the average, about three-quarters of what they actually are. The second group contrasts with this. In this group, drawing studies from Nagoya, Edinburgh and Tokyo, there is a tendency to consider places about a third further away again

Distance estimation in cities

63

than is actually the case, as well as adding a small constant. One study from Glasgow and one from Nagoya can be treated with the Sydney study as forming a third group which falls in between the previous two. An examination of the general form of the seven cities involved was carried out, on discovering these groupings, to see if there were any obviously distinct properties of the cities which might account for the groupings. Lynch’s (1960) concept of ‘imageability’ was kept in mind during our search. It is plausible that the more a city possesses what Lynch describes as those ‘physical qualities which relate to the attributes of identity and structure in the mental image’, the more likely are people to be able to relate their judgements to specific, identifiable locations and thus produce accurate estimates. However, we are dealing with estimates taken across the whole city and so only the most general and obvious indicators of imageability are likely to be relevant. One answer seemed to lie in the fact that the cities are either essentially coastal, with a large bay or sea acting as the most distinct boundary of the city, or they are placed upon a large river which efectively divides the city in half. Glasgow, Heidelberg and London are divided by rivers; Nagoya, Tokyo and Edinburgh are bounded by bays or oceans. Sydney forms an intributing combination of the two, being placed on either side of an estuary with the Pacific Ocean as a boundary. The only other distance regression we could locate (Cadwallader, 1973) has been quoted. It will be noted that it falls between our Sydney and Nagoya regressions. As this study was in the Los Angeles basin, it does not go against the general pattern described here. Before considering the implications of this, it must be emphasized that there are two studies which rather muddy the water, one from Glasgow, the other from Nagoya. Thus, any general hypotheses that are derived from considering them will require considerable modification in relation to specific groups of respondents in specific cities. The sources adding to the variations in responses should be noted. We are dealing with responses generalized across locations and across respondents. As Cadwallader (1973) has illustrated, there are large diferences between people for the correlations between estimated and actual distances. We have comparable data which will be the subject of later reports. Lowrey (1970) has emphasized the difering role of diferent types of location in the estimation process. We can find less evidence for this in our studies. This is nonetheless a potential cause for variability in our results; it is so difcult to equate location types across cities as varied as Heidelberg, Sydney and Nagoya. However, from the generalized view adopted in this paper, the type and variety of locations and individual’s perceptions of them are crucial components of the ‘mental image’. We, therefore, argue that there is considerable validity in making cross-city comparisons on the basis of distance estimations. In general, for the concept of imageability to find some empirical validity from our data, it is necessary for it to be modified. We would like to relate it more closely to the Gestalt concept of ‘Pregnanz’. If we accept that people place some structure upon their ideas of a city in order to cope with that city, then it is plausible that a ‘simple structure’ is involved. This would mean that people would place on their concepts some simplifying scheme which would enable them to remember and act on their image the more readily.

64

David Canter and Stephen K. Tagg

For example, in the case of a city with a river running through it, it is possible that the river would act as a reference line. As it takes up geographical space, we may expect people to add the psychological attributes of that space as a constant to their judgements. Support for this is found in the comparison of study 2 with study 3. In the former, in which distances are across and around the Clyde at its narrow point, the B-value is 1.28. In the latter case, where the wider part of the river towards the estuary is included, the B-value is 2.81. Can the B-values be used as a test of the psychological dominance of major geographical features? The role of the ‘reference line’ in longer distance estimates, it is hypothesized, may well be to facilitate estimation due to the facilitation of accurate location in the conceptual structure of the estimated points. If we further assume that actual size of this reference line is discounted in the estimation of the distances, we would obtain the situation illustrated here: an overall reduction in the error of estimated distances as the actual distance increases. These hypotheses derive, in the main, from the structure of our linear regression equations, a structure which contains two distinct components linked to the A- and B-coefcients. That certain geographical features, such as rivers, may play difering roles in each of these components is one of the major hypotheses requiring future testing to be derived from our studies. In a city lacking an obvious formal component, like Tokyo, we would expect people to place a simplifying form upon it. Attempting to find a circle, or rectangle, in the jumble of streets or railway lines would be one method of doing this. People need to encompass the city in order to cope with all of it. The lack of a formal component would tend, psychologically, to act in the opposite way to a central river (drawing places away from the centre), thus leading to overestimations. At best, this is a very loose formulation! Very few locations were used in each city and Tokyo and Edinburgh, for instance, both have centrally located castles, which might be expected to aid in structuring. Therefore, for further clarification, we shall next examine particular studies from each of the three groups. Even if later evidence discounts this formulation, it is clear that the precision of distance estimation and the regression model does contribute considerably to our definition and understanding of city images. For example, a number of detailed hypotheses may now be put forward which would test our briefly stated formulation. (We would welcome collaboration with anyone wishing to cooperate in these tests.) We would predict, as an illustration, that cities with a clear structure (e.g., Paris, San Francisco, New York, Stockholm?), especially when placed on either side of some dividing topographical feature, would produce slopes of actual to estimate of less than one and B-values greater than one. Whereas cities with a less readily definable simple form would produce overestimates (e.g., St. Louis and Los Angeles in the United States, Birmingham or Milan in Europe?). Comparisons of configurations In order to explore further the patterns revealed by distance estimations, one study from each of our three groups has been selected, a Tokyo study (number 8), the Sydney study and the Glasgow study (number 3). For each of these sets of data,

Distance estimation in cities

65

a configuration of the locations in two-dimensional space was required. This was in order to see what types of distortions from the actual geographical configurations were present in the estimated distances. The idea behind the use of distance estimations is, of course, that they must in part be drawn from, or relate to, some conceptual configuration. Thus, by working backward, so to speak, and producing a configuration based upon the estimates, it should be possible to highlight important psychological properties of the ‘conceptual configuration’. The process used to generate the configuration was the nonmetric, multidimensional scaling procedure developed by Guttman and Lingoes, ‘SSA-1’ (Lingoes, 1973). In essence, this is an algorithm which takes as its starting point the rank order of, in our case, the mean distance estimates. It then, by a series of iterations, produced a two-dimensional arrangement of the locations such that the rank order of the distances between them is the same (within specified limits) as the original rank order of the means. However, both because the number of locations is limited and because of the nonmetric nature of the programme, even if the actual distances were used instead of estimates, a geographically accurate arrangement would not necessarily be produced. It is thus essential to compare arrangements based upon estimates with one based upon actual distances, both produced by the same programme.2 Figure 6.1 shows the ‘actual’ and ‘estimate’ configurations for each of the three studies. In interpreting these configurations, it must be remembered that only the distance between locations is used as data. Thus, any simple transformation which maintains the relationships between these distances is possible. Two configurations which were, for instance, mirror images of one another or upside down in relation to each other would be identical in terms of the statistics on which the programme is based. The crucial diferences between actual and estimated configurations are as a consequence not those relating to orientation of the points in space (which is an artifact of the programme) but the relative distances between the points. Furthermore, because the programme has only limited information about the points, it is possible for quite large distortions from geographical accuracy to occur. However, because an identical algorithm is used for each set of data, comparisons are still meaningful even when these distortions are gross (as with the Sydney study). To facilitate the interpretation of the configurations, a key geographical feature has been indicated in Figure 6.1. Two things may be noted in the results for Glasgow shown in Figure 6.1. First, the estimated locations tend to be distorted in relation to the actual locations by, in general, being closer to the Clyde. This is shown clearly by reference to GuttmanLingoes’s derived distance coefcients. For the cross-Clyde distances between points 4 and 5, and 3 and 6, they are 1.15 and 0.89 for the actual configuration, respectively. But they are 0.78 and 0.76 for the estimated configuration. However, for the distances parallel to the Clyde between points 4 and 3, and 5 and 6, they are 0.79 and 0.63 for actual and the scarcely diferent values of 0.79 and 0.65 for the estimated distances. The second point to emerge from the Glasgow study is that the points towards the estuary (points 1 and 2 at the right of Figure 6.1) are much further away from each other in the estimated than the actual (1.27 estimated, 0.83 actual). This goes

66

David Canter and Stephen K. Tagg

Estimate

Actual

2

2

Glasgow

3

4

3

7

7

1 4 5

6

6

1

5

River Clyde Indicated

11

8

3

9 4

Sydney

2

8 7 10

6

9

5

2

7

12

12

3 1 6

4 1 11

10

Coastline and Bay Indicated 4

9 4

9

1 6

6

10

Tokyo 8

2

10

7 3

2

11 3 5

7

11 1

8 5

Main Circle Railway Line Indicated

Figure 6.1 SSA-1 configurations for Glasgow, Sydney and Tokyo based on estimated and actual distances.

Distance estimation in cities

67

against the general trend for Glasgow but seems to relate to the bridges available; there are many in the city centre at the left of our configuration. This is a point to which we shall return. Jumping to the Tokyo study, we can see how distances are overestimated there. The railway stations, which form the named locations most readily available in Tokyo, form a configuration approximating a circle when based on estimates, whereas they are actually in a form nearer to an ellipse. This implies that north-south estimates (along the line) are reasonably accurate but that east-west judgements are overestimated. This relates well to our finding that when ten respondents were asked to draw a map of Tokyo, all immediately produced a circle to represent the main railway and elaborated from there. This railway system is represented as a circle in generally available schematic diagrams. A consideration of Tokyo will show that it is such an intricate city that there is no other overall structure to which reference can be made when representing it schematically. The implications of the simplifying circle are thus clearly revealed in the distance estimates. Sydney’s complex form together with the sparse information available in the data leads SSA-1 into some difculties. It was found necessary to curve the line of points along the coast (12, 10, 1, 6 and 11) in order to fit them into its two-dimensional space. The estimated distances lead to the opposite trend, folding the geographically straight row of points in to bring the extremes closer. The corollary of this seems to be to push the inland locations, which are far from the city centre, much further away from the coast in the estimated situation. Thus 12 and 9, and 11 and 8, whilst very close in the actual configuration, are a long way from each other in the estimated. The resultant of these two processes appears to average out overall and give us a regression slope close to unity. It may be noted again that the distances across Sydney harbour (e.g., 7 to 4), where there is a major road/rail crossing, were little distorted from one configuration to the other. But (as in Glasgow) those distances which did not share any main transportation artery were most susceptible to distortion (e.g., 12 to 9 and 11 to 8). Summary and conclusions The seminal work of Lynch has produced studies based, in the main, upon data extracted from the maps which people have drawn. These studies have proven time consuming to conduct, and their analysis has often erred on the side of illustrative quantity rather than statistical quality. Yet many of the ideas relating to city images do have quantitative implications, especially for perceived or estimated distances. An examination of these implications should lead to a clarification of the basic assumptions. Thus, at the large scale, the imageability or ‘legibility’ of a city should relate to the general pattern of distance misestimation. Yet we have seen that whilst a city with a confusing image (like Tokyo) may well lead to general overestimation, a more ‘legible’ city (like Glasgow or Heidelberg) may still produce distortions in the direction of underestimations. The commonsense quality of these relationships only serves to hide a profound psychological question as yet unanswered. If you have a clear idea of where places are in relation to each other, do they seem closer together?

68

David Canter and Stephen K. Tagg

Why? And is this a part of the conflicting efects of direction of distance in relation to the city centre reported by Lee, Golledge and others? Our study indicates the critical roles which may be played by road and rail systems, general topography and major geographical features. This leads to the argument that the contrast between the findings of Golledge et al. and Lee possibly relates to the fact that the former were dealing with estimates along a major road in, essentially, a featureless countryside whilst Lee was not. Indeed, the role of rivers and roads may help to bridge the theoretical diferences between these two groups. Lee’s explanation in terms of the valence of the city centre may be replaced from our results with the valence properties of a whole series of places linked in some conceptual structure. This structure, in its turn, may be based upon the types of experiences and learning processes which Golledge and Zannaras emphasize. Furthermore, when dealing with concepts at the scale of cities, it would appear that roads, railways and rivers, rather than acting as ‘barriers’, may act to give the city its coherent structure. Indeed, they may all be more fruitfully construed as diferent types of conceptual ‘links’. Finally, then, we can point to the readily available crow flight distance estimations as a valuable part of the arsenal of techniques available for examining the internal representations of cities. It is hoped that more people will make use of them in the future. Notes 1 The slight diferences in the number of points in each study can only be explained as a function of time and distances at which the studies were carried out. The most efective procedure, embodied in studies 1 to 6, emerged during the course of the other studies. 2 Like all other nonmetric, multidimensional scaling algorithms, SSA-1 involves the iterative refinement of a starting configuration in terms of the given constraints. The nature of this configuration can be important. In this example, the default option of the eigen vector starting configurations was used. When a geographical, grid-based starting configuration was substituted, the resulting configuration proved more difcult to interpret.

References Appleyard, D. (1970). Styles and methods of structuring a city. Environment and Behavior, 2 (1), pp. 100–116. Bartlett, F. C. (1932). Remembering. New York: Cambridge University Press. Bratfisch, O. (1969). A further study of the relation between subjective distance and emotional involvement. Acta Psychologica, 29, pp. 244–255. Briggs, R. (1972). Cognitive Distance in Urban Space (Ph.D. dissertation, Ohio State University). Cadwallader, M. T. (1973). A methodological analysis of cognitive distance. In Preiser, W. F. E. (eds) Environmental Design Research, vol. 2. Stroudsburg, PA: Dowden, Hutchinson & Ross. Dornic, S. (1967). Subjective distance and emotional involvement, a verification of the exponent invariance. Psychological Laboratory Report 237. University of Stockholm. Ekman, G. (1961). Some aspects of psychophysical research. In Rosenblith, W. A. (ed) Sensory Communication. New York: John Wiley, pp. 35–47.

Distance estimation in cities

69

Gogel, W. C. (1969). The efect of object familiarity on the perception of size and distance. Quarterly Journal of Experimental Psychology, 21, pp. 239–247. Golledge, R., and Zannaras, G. (1973). Cognitive approaches to the analysis of human spatial behaviour. In Ittelson, W. H. (ed) Environment and Cognition. New York: Seminar Press, pp. 59–94. Ittelson, W. H. (ed) (1973). Environment perception and contemporary perceptual theory. In Ittelson, W. H. (ed) Environment and Cognition. New York: Seminar Press, pp. 1–19. Künnapas, T. (1960). Scales for subjective distance. Scandinavian Journal of Psychology, 1, pp. 187–192. Kunnapas, T. (1968). Distance perception as a function of available distance cue. Journal of Experimental Psychology, 77, pp. 523–529. Kuroda, T. (1971). Distance constancy: Functional relationships between apparent distance and physical distance. Psychologische Forschung, 34, pp. 199–219. Lee, T. (1970). Perceived distance as a function of direction in a city. Environment and Behavior, 2 (1), pp. 40–51. Lingoes, J. (1973). The Guttman-Lingoes Nonmetric Program Series. Ann Arbor: Mathesis. Lowrey, R. A. (1970). Distance concepts of urban residents. Environment and Behavior, 2 (1), pp. 52–73. Lundberg, U., Bratfisch, O., and Ekman, G. (1972). Emotional involvement and subjective distance: A summary of investigations. The Journal of Social Psychology, 87 (2), pp. 169–177. Lundberg, V. (forthcoming). Individual functions of subjective time distance and emotional reaction. Scandanavian Journal of Psychology. Lundberg, V., Bratfisch, O., and Ekman, G. (1972). Emotional involvement and subjective distance: A summary of investigations. Journal of Social Psychology, 87. Lynch, K. (1960). The Image of the City. Cambridge, MA: MIT Press. Stanley, G. (1968). Emotional involvement and geographic distance. The Journal of Social Psychology, 75 (2), pp. 165–167. Stanley, G. (1971). Emotional involvement and subjective distance. Journal of Social Psychology, 84, pp. 304–310. Stea, D. (1969). The measurement of mental maps: An experimental model for studying conceptual spaces. In Studies in Geography No. 17. Evanston, IL: Northwestern University, pp. 228–252. Stevens, S. S. (1957). On the psychophysical law. Psychological Review, 64 (3), pp. 153–181. Stevens, S. S., and Galanter, E. H. (1957). Ratio scales and category scales for a dozen perceptual continua. Journal of Experimental Psychology, 54, pp. 377–411. Teghtsoonian, R. (1971). On the exponents in Steven’s law and the constant in Ekman’s law. Psychological Review, 78 (1): 71–80.

7

The psychology of place David Canter

Introduction My motivation in writing this book grew out of an increasing awareness that there exist a number of diferent studies, in a wide range of discrete areas (geography, planning, psychology, architecture, urban sociology) which all have a common concern: how people make sense of and cope with their surroundings, whether it be nature trails or nursery schools, crowded kitchens or city centers. Furthermore, all these studies appear to have a special interest in what are variously called ‘pictures in the head’, ‘mental maps’, ‘urban images’ and so on – the notions which we ourselves have of the places we experience. Yet no overview of this material appeared to exist beyond the introductions to books of readings. No attempt had been made to draw this material together and provide even the rudiments of a framework for it. It struck me as a worthwhile task to undertake because its potential ramifications are so great. As I proceeded with this task, it dawned on me that what all the diverse profession and disciplines shared was an interest in ‘places’. The idea of ‘place’ seemed to be one which, if explored, could provide a concept which would act as a bridge between the various fields of inquiry. However, further consideration suggested that central to the interest in ‘places’ was the desire to understand the ways in which we represent them ‘in our heads’. That is how the term ‘psychology’ has crept into the title of this book – not as a belief that psychologists are the only ones for whom its contents are relevant but as a way of indicating that the focus of the book is on individuals and their comprehension of places. Although I took psychology into the title for this book, in the text I have eschewed, in the main, any reference to studies performed within the major psychological rubric, the laboratory experiment. This was not as much a deliberate policy as the desire to deal with research clearly and directly related to the experience of the environment. By their very nature, laboratories lead to an abstraction of particular stimuli from their natural habitat. Even the so-called ‘simulation’ of environments by the use of photographs, slides and the like leaves much to be desired in the way of real motivations or experiences on the part of the respondents. I, therefore, found myself avoiding studies which did not draw immediately and materially upon actual surroundings, unless their point was very obviously pertinent DOI: 10.4324/9781003313052-8

The psychology of place

71

to the issues at hand. It surprised me that, once the earlier guidelines for the book had emerged, the research of quality available was limited, but it was even more intriguing to discover how powerful was the message of those authors who had originally turned people’s attention to the psychology of places (yet how few of the implications of their work had been explored). I even found that most of the research problems still being actively pursued were clearly nascent in work carried out in the early part of this century. The objective of the book thus emerged to provide, first, a sketch of the attempts which have been made to explain how we conceptualize places and, second, an outline of the methods which have been used to study those conceptualizations. The third objective, of pointing the direction in which to look for the applications to decision making of the products of the explanations and methods, followed inevitably from the first two objectives. My hope was that if the book were at a sufciently introductory level, it would enable members of the various disciplines to begin to benefit more fully from talking and working together. Like any village marriagemaker, I shall judge the success of my introductions by the health and number of the ofspring produced. The book ahead This book is about the psychological processes which enable us to understand places, to use them and to create them. It is not about any of those topics so favoured by experimental psychologists: space perception, object perception, colours or shapes. Rather, it is concerned with those situations in which people live and work, converse with others, are alone, rest, learn, are active or still. This does not mean it is concerned with activities alone or only with the buildings which house them. It is about those units of experience within which activities and physical form are amalgamated: places. However, my use of the word ‘place’, in this book, is slightly unusual, and so in Chapter 1, I attempt to put it in perspective, by showing how places have their impact, how our understanding of them influences our actions and how places are qualitatively diferent from objects. If we are to understand people’s responses to places and their actions within them, it is necessary to understand what (and how) they think; and thus, this book will concentrate more upon conceptual systems than behavioural systems. This is not to dismiss the great importance of the actions of others in influencing the course of our lives. It is rather to look within the individual for the causes of his actions, at his interpretation of the context within which he finds himself. Thus, if we are led to believe by the facade, say, that a particular building is a youth club, then we are likely to enter it in a diferent frame of mind and behave, at least initially, somewhat diferently within it than if it looked like a Methodist chapel. Our understanding of the situation may be thought of as producing, or at least influencing, our behaviour. The essence of this argument (which has its origins in the work of Mead, 1934; Kelly, 1955; and more recently, Harre and Secord, 1972) is that any act is made in relation to the context within which the individual thinks himself to be. In other words, the organism recognizes, acknowledges or in some way, takes account of its

72

David Canter

context, failing which, action is difcult and appropriate action is impossible. Even an animal, which is thought of as reacting instinctively to a stimulus, must first be aware of that stimulus before it can act. Dealing with the infinitely more complex human being (in which instinctual responses, if any, are very rare indeed), it is clear that for sane survival, the stimulus must be recognized, interpreted and reacted to in an appropriate way, bringing to bear all the internal mechanisms available for structuring experience. As a consequence, although in this book I shall turn to observation of behaviour from time to time, the central interest will be in those cognitive systems (rather than purely perceptual processes) which enable us to act appropriately in places. So this book is about places and the conceptual systems we employ in order to facilitate our operation within places, and perception will play a minor role in my argument. It is particularly important to examine the relationship of perception to cognition because geographers and architects, as well as psychologists, have taken these technical terms from the laboratory (in which they blossomed) and assumed that they can be used in an identical way with reference to the environment (Goodey, 1971; Koroscil, 1971). One further distinction may help to focus the content area of this book. That is the distinction between places, which are units of the environment, and objects, which exist within an environment. It is necessary to clarify this distinction, like that between perception and cognition, because of the widespread belief, among academics and laymen, that places may be regarded just as big objects. The consequence of this is that attempts are made to apply theories and methods developed from dealing with objects to the study and even the design of places. It is a fallacy to which geographers (who are nowadays increasingly involved in the study of such issues) as well as architects and planners are liable. Themes and issues There is one particular advantage in examining the impact of our surroundings by studying conceptual systems of places and that is that we are not tied to any given environmental scale. In the book ahead, we will range from beds on hospital wards to regions of Britain. Underlying the exploration of this range is the assumption that the psychological processes involved are similar. As the book progresses, readers will be able to judge for themselves the validity of this assumption, but it is an assumption which helps to bring the formulations of the psychology of place, which at times are rather abstract, closer to daily experience. For example, it helps to show why we ‘use similar words to describe’ our feelings both about buildings and about landscape, and it helps to explain the layman’s frequent difculty in making the distinction between the architect and the planner. Indeed, the bizarre political battles which sometimes occur in local authorities between diferent departments, such as the parks department fighting the housing department for land, or the planning department insisting that playgrounds are not the responsibility of the architecture department, are battles frequently caused by ofcials who forget that they are dealing in places. The number of square metres involved or the size of the area which is a particular colour

The psychology of place

73

on a map are the considerations which lead to confused conflict. When the experience of places is taken as a focus, then scale does not have qualitative significance and many conflicts can be resolved. In studying the cognitive systems pertinent to the environment, we focus upon two recurrent themes. One is the development of a theory which will enable us to describe and understand the structure of these systems. The other is an explanation of the procedures available for bringing these internal processes into public view. One of the criticisms of many psychological theories is that they present a picture of people somehow less complex than we know ourselves to be. Fickleness, confusion, agitation, inconsistency and many other supremely human states are all part of the daily load, yet if a psychologist presents an account of them, they frequently seem to change into something less than fully human. As a consequence, in Chapter 2, when we explore theories of cognitive systems of place, we will draw upon the ideas of a physiologist and an economist as well as those of an architect. By combining their diferent viewpoints with those of cognitive psychologists, we can begin to sketch the properties of these inner processes in a way which captures something of our daily experience. But we need to do more than just capture experiences of places. In order to develop a scientific understanding, we must record and measure them (our second recurrent theme). Everyday activities remind us of the variety of methods available for describing and representing places, such as photographs, maps, books, caricatures and the like. In Chapter 3, the techniques which have been used by psychologists, geographers, architects and planners for revealing people’s inner processes are compared and contrasted. From these comparisons, we can begin to see why the residents of a rambling city like Tokyo rely on maps for directions, while the gridiron plan of a North American city generates a dependent on the numbering of blocks. We are then in a position to consider the general question of whether the reliance on particular modes for representing places does itself leave its mark on how we make use of our conceptual systems. For example, are land values more related to city centre development if the maps commonly available emphasize the crucial location of that centre? Maps, of course, are a much-favoured device for locating and describing places. Yet as discussed in Chapter 4, they have their problems. One of these is the extent to which they strengthen particular facets of our conceptual system, and this may have a particular impact in those situations in which we rely upon a schematic map, such as an underground railway system. But they may generally lead to an emphasis on the geometric structure of an area, in two dimensions, rather than the essentially three-dimensional experience of moving through it. A procedure which we commonly use for representing what we know of places, yet one quite diferent from mapping, is the estimation of distance. Given the wide use of distances in signposting, it is surprising that there has not been more study of the implications of distance estimations. In Chapter 5, the few studies which do exist are discussed. We may pre-empt that discussion by mentioning the overall accuracy and consistency of distance judgements. Even though this accuracy occurs against a background of very great individual variation, it does demonstrate the sophistication of the cognitive process we develop for coping with our surroundings. Interestingly, it also lends

74

David Canter

support to the notion that our internal representations do contain information akin to that contained in cartographers’ maps. One simple, practical implication of these studies is that we may find it more appropriate than it is commonly believed to use distance information when, for example, assisting people with route finding. Although mapping and distance estimation do give much useful information about people’s conceptions of place, they still leave great areas of our experiences untouched. Most notably, they do not give us insight into the feelings which people have about places or how they evaluate them. As a consequence, Chapter 6 is devoted to illustrating what can be learnt about place conceptions from the ways in which people describe places. What emerges is that there are two important aspects to descriptions. One is the evaluations assigned to a place by people, and the other is the range and type of activity associated with it. From this is apparent the extent to which even the most abstract or romantic accounts of places contain within them evaluations and expectations directly linked to the behaviour considered appropriate for them. However, because of the richness and complexity of verbal descriptions, they reveal more clearly the diference between people than do the other procedures to be discussed in this book. It is, therefore, appropriate that in Chapter 7, we turn our attention to accounting for these diferences. In the three previous chapters, the range of ways in which conceptual systems may be explored are presented so that by Chapter 7, we are in a position to look at the origin and variation of these systems. In many ways, the focal issue for utilizing information derived from place psychology is the possibility of systematizing individual variation. It is frequently the starting point for popular interest in the topic. In Chapter 7, it is argued that the major cause of diferences between people in their environmental conceptions is the diference in their roles vis-à-vis the environment. In other words, their difering experiences give rise to difering perspectives. This has one very great implication for the processes used for creating environments. By virtue of being an environmental decision maker, a designer’s conceptual system is likely to be diferent from that of people not in that role. There is thus great possibility for a mismatch between creator and user. This mismatch may only be resolved by utilizing the procedures which will be referred to in earlier chapters for making public the conceptual systems of all involved. Chapter 8 draws the book to a conclusion by presenting the outlines of procedures which may be used for incorporating the place conceptions of many groups into design decision making. In drawing up proposals for participation in this way, we come to rely heavily upon the possibilities of identifying and describing the conceptual systems of the various interest groups. This puts considerable emphasis on our abilities to analyze complex systems and seems to imply that the use of computers may well become important for this purpose. The book proceeds, then, from an examination of the nature of those conceptual systems which enable us as individuals to create and make use of places. This leads to an examination of the ways in which we may elucidate and describe our internal representations of places. This finally puts us in a position to discuss variations between people and the consequent possibilities of generating new procedures for designing and producing places.

The psychology of place

75

It is worth noting that by the end of the book, we are able to comment on the construction of places, whether it be the design of a living room or the redevelopment of a city centre. This emerges as a goal because one of the implications deriving from our explorations in place psychology is that the identification and use of places in a clear and consistent way is of direct value to people. It also seems to be the case that, with the centralization of decision making and rapid communications, fewer conceptual systems are exerting increasing influence. The net efect of this is the increasing international homogeneity, the same architectural and urban forms being reproduced throughout the world. It is hoped, therefore, that the psychology of place will be one of the forces bringing back the possibility of a wider range of environmental experiences through the creation of places more appropriate to their inhabitants.

Part 2

Elaborating the theory of place

Once The Psychology of Place had been published, various opportunities arose to develop many of the ideas only briefly sketched out in that initial volume. Place evaluation was a particularly interesting context for this development. In the paper reproduced here, the agency inherent in human activity, the purposes which people strive to achieve, at the heart of the theory of place, is taken as the basis for developing evaluations of places. This purposive focus crucially enhances the usual satisfaction survey by giving people objectives when evaluating a place. In the results of large-scale surveys using this approach, there was an intriguing finding. This related to the way the diferent scales of place were considered related to each other. This was whether it was the house, location or neighbourhood – when evaluating housing or close, local or distant interactions in hospital wards. The general assumption is that these diferent scales nest within each other. In other words, the feelings about a house influence the satisfaction with the location, and they in turn influence views about the neighbourhood. However, what emerges is that these diferent scales operate independently of each other. Our reactions to places immediately around us have the same psychological structure as the larger scale, independently of each other. Once this idea is expressed, it becomes obvious. I can hate the house I live in but love my neighbourhood. A nurse may regard access to reach patients in distress on the ward as very easy but dislike the overall layout of the hospital. Here is a crucial feature of the psychology of places. Because places are essentially conceptual structures, our reactions to them can operate at any scale. But those reactions always have the same natural facets of meaning, space and its social implications and aspects of the physical structure. The purposive perspective implies some form of action in a place, even if that is quiet contemplation. This idea is taken a step further by drawing on social psychological, or even sociological, understanding of how actions are structured by human interactions. This draws attention to the remarkable absence in social psychology of consideration of the physical context of social processes. As Barker and his associates in their presentation of an ‘ecological psychology’ pointed out (Barker), the best predictor of the actions a person will be engaged in is not their traits or other aspects of their personality but the setting in which they are. Unfortunately, the learning DOI: 10.4324/9781003313052-9

78

Elaborating the theory of place

theory tradition from which Barker emerged blinded him to any consideration of the physicality of those settings. Putting situations in their place (reproduced here) takes further the significance of place-related roles and rules. By connecting directly with issues in social psychology, it is possible to elucidate many aspects of the psychology of place. The pattern of place use is predictable from the rules associated with the diferent roles in a place. How students sat around Plato in the groves of Academe was bound to have the same broad similarity to students and the teacher in a modern classroom. Archaeologists have even formed views of gender diferences in neolithic houses from where the remains of activities have been found. The further consideration of the interactions at the heart of place psychology is the recognition that the nature of places is inevitably a dynamic, unfolding process. This idea has been developed in interesting ways at the architectural scale by Sophia Psarra (2009) and at the urban scale by Kim Dovey (2010). The paper reproduced here deals with the essence of these dynamic processes, the interplay between actions and other aspects of places. I gave this the perhaps arcane title an ‘existential dialectic’. In addition, I wanted to break away from the often rather rigid way of organizing academic papers by arranging the chapter around the headings a suite of music by J.S. Bach. I was pleased to realize how readily the themes I was exploring fitted the diferent types of dance forms that Bach drew upon. Even more surprising was when a student, knowledgeable in musicology, wrote a detailed analysis of my essay with cross-refences between the psychological and musical ideas. Sadly, this analysis has gone missing. References Dovey, K. (2010) Becoming Places London: Routledge Goodey, B. (1971) Perception of the Environment Birmingham: University of Birmingham Harre, R., & Secord, P.F. (1972) The Explanation of Social Behaviour Rowman & Littlefield Kelly, G. (1955) Personal Construct Psychology New York: Norton Koroscil, P.M. (1971) The behavioural environmental approach. Area, 3(2), 96–99. Mead, G.H. (1934) Mind, Self, and Society, Vol. 111 Chicago: University of Chicago Press Psarra, S. (2009) Architecture and Narrative London: Routledge

8

The purposive evaluation of places A facet approach David Canter

AUTHOR’S NOTE: I am grateful to Dan Stokols for his detailed comments on earlier drafts of this paper. The need for a theory of evaluation In reviewing environmental psychology as a whole, Stokols (1978) singled out evaluation research as being especially deficient in theory. Four years later, when Russell and Ward (1982) reviewed conceptual and theoretical developments in environmental psychology, they found it worthwhile to refer to only three researchers involved in ‘the assessment of places’, and even in these cases, it is the technology that they emphasized rather than the theoretical models. In commenting on this gap in the theory of evaluation, when considering conceptual aspects of environmental quality assessment, Canter and Kenny (1981) have labeled environmental evaluation as ‘the pragmatic endeavor par excellence’. Across a decade of major publications in environmental evaluation (BPRU, 1972; Craik and Zube, 1976; Friedman et al., 1978; Marans and Spreckelmeyer, 1981), there have been few cumulative research findings. Each study has emerged almost as if no others had ever been conducted. In responding to the need for a theory of environmental evaluation, this article proposes a preliminary model that draws upon a phenomenological perspective of the nature of environmental evaluation. It harnesses the facet approach to theory development and converts the perspective into a form that is amenable to empirical examination. Published studies are drawn upon to help clarify the model and to demonstrate its validity. Objectives of the model The central proposition of the proposed model is that the fundamental processes underlying human experiences of places are consistent across people and places. It is the content of these processes and their emphases that will vary in important and identifiable ways from one person to another and from one place to another. Furthermore, it is proposed that a model of environmental evaluation should achieve the following objectives: 1 The derivation of hypotheses about consistencies in environmental evaluation across cultures and across settings. DOI: 10.4324/9781003313052-10

80

David Canter

2 The generation of explanations and predictions about diferences between individuals in their evaluations of the same and diferent settings. 3 The specification of hypotheses about the consequences for user performance and well-being of any given place, developed from an examination of the placespecific content of the model. 4 The provision of a general template for evaluation instruments, the contents of which can be specified from knowledge of the place under study. This maintains the possibility of comparability across settings while avoiding the use of bland, general-purpose instruments that are not sensitive to the nuances of particular settings. 5 Identification and classification of some of the consistencies and confusions in environmental evaluation literature and indication of the relative importance of the various components of evaluation. These objectives will be assessed in the final discussion when examining the implications of the proposed model. For the present, they serve to emphasize that any model proposed should have theoretical, methodological and practical implications. The goal of achieving this combination of objectives underlies the proposed model. A consensus on place experience One reason it ought to be possible to achieve such ambitious objectives for a model of environmental evaluation is because of recent developments in understanding within environmental psychology and related areas of social psychology. These developments, in combination with the methodological opportunities generated by the facet approach, make a powerful tool for theory generation available. The methodological consequences of the facet perspective will be turned to later, but first, it is important to recognize the emerging theoretical consensus on which the proposed model draws. Possibly the most fundamental shift in environmental psychology, which now makes general model building viable, is the widespread acceptance that the experienced environment must be considered as a set of related multivariate phenomena existing over time. This seemingly obvious point was first clearly articulated by Ittelson (1973) and is in marked contrast to the perspective of environmental engineers and those psychologists who follow them, who focus on specific physical variables, such as temperature or noise level, in isolation. Ittelson (1973) briefly mentioned the further important point since elaborated by many people (Relph, 1976; Canter, 1977b; Lef, 1978), namely, that the environment is used and experienced rather than simply looked at. This implies that people have a purpose for being in any given location and that their plans and intentions must consequently play a role in any understanding of the psychological implications of that environment. One further consensual trend has been the general acceptance of the cognitive component of experience, which emphasizes the primacy of perception over sensation. This cognitive emphasis, which parallels and probably has roots within similar

The purposive evaluation of places

81

developments in experimental psychology, has been shown to have implications for the understanding of such phenomena as crowding (Stokols, 1978), stress (Lazarus, 1966) and environmental learning (Downs and Stea, 1977). Thus, most theoretical views of people’s interaction with their surroundings now emphasize the understanding that people have of their environment and how they model that understanding. As a way of enshrining in a new technical term the more cognitive and wholistic properties of environmental experience, a number of authors have used the term ‘place’ (e.g., Relph, 1976; Canter, 1977c; Seamon, 1979; Tuan, 1977; Stokols, 1981). Thus, places, at a very minimum, are recognized (a) as being multivariate, (b) as existing in some form of cognitive representation within individuals and (c) as being purposively used by people as a way of completing plans or achieving objectives. Indeed, the term ‘place’ is used throughout the recent review of environmental psychology by Russell and Ward (1982). At the same time that environmental psychologists have embraced the richness of places as a context for human actions, a number of social psychologists have recognized the value of studying ‘situations’ in which behaviour occurs (reviewed by Argyle et al., 1981). As Canter (1983b) has discussed in detail, their descriptions of situations bear a close resemblance to the important aspects of places already mentioned: places are construed as multivariate entities, people are assumed to be intentional in how they behave within places and an understanding of a person’s perception of the situation is emphasized. From a number of diferent directions, then, the constituents of a general model of place experience are emerging. However, these constituents have not been explicitly directed to the question of place evaluation. Evaluations have tended to be conducted in a strictly pragmatic spirit, without a conscious resource to any specific theoretical framework for the nature of place evaluation. Purposive evaluation The most fruitful starting point for a theory of place evaluation is the notion of intention or purpose. The attraction of evaluations as a possible contribution to personal or professional decision making implies that evaluations are directed towards some goal. It is thus no major step within the cognitive framework to accept the consequences that the Building Performance Research Unit (1972) postulated from their goal-oriented model: It is central to our goal oriented model of people that if they had no concept of their state when their goals were achieved then there would be little directed impetus towards achieving them. It must also therefore be possible for people to estimate at what position they are in relation to their goals. Statement of this position is a statement of satisfaction. (p. 68) This viewpoint on place evaluation is directly in accord with the more focused examination of the environmental consequences for stress considered by Stokols

82

David Canter

(1979) in his discussion of ‘people’s subjective appraisals of their environment’ providing ‘a reasonably straightforward index of the quality of their experiences in those settings’ (p. 36). He points out that stress may occur where the ‘straightforwardness’ of this relationship does not pertain, but this aspect of the extremes of experience will be discussed after the development of the proposed model. The referent of experience

There is one set of empirical questions, deriving from the purposive framework, which has not received wide attention. This is the classification of the objectives people have in their use of places. The multi-modal qualities of place experience attest to the fact that a person’s interactions within a place always have a number of related objectives, or referents. The essence of the argument here, which has been presented at length elsewhere (Canter, 1977a), is that although the experience of a place can be broken down into distinguishable components, the experience itself is unitary. Thus, the components must all be present in any examination of place evaluation. They can operate as foci for attention (in other words, as referents for an evaluation), but a place always exists in relation to a number of referents. Putting this viewpoint more concretely, a person, for example, always has a physical and a social existence. Questions can direct attention to how a place contributes to the physical or social well-being of a person without their implying that the physical and the social are totally distinct systems, especially without assuming that they are orthogonal dimensions. It is the recognition that the physical and the social aspects of experience can be treated as foci of attention, rather than independent domains, which frees a purposive model from the conundra that beset environmental psychologists who follow, even if only implicitly, a stimulus (environment)-response (social) approach. It is at the heart of the questions raised by phenomenologically oriented geographers, such as Relph (1976) and Seamon (1979), and central to the discussions of crowding by Stokols (1978). It is also the basis of the debate about the locus of the environment in which Wohlwill (1973) places the environment at one side of the stimulus-response divide and Russell and Ward (1982) look for it somewhere else. By accepting that both physical and social components are present in place experience, it is possible to make an important distinction. This is the distinction between, on the one hand, the relationship between the social and physical aspects of experience and, on the other hand, the quite separate issue of how the physical and social components of experience relate to ‘objective’ measurements, such as descriptions of layouts and measurements of distances. These measurements are representations of physical phenomena the individual does not experience directly. Thus, the central question for environmental psychology is no longer the understanding of the nature of the relationships between people and their physical surroundings but an understanding of the ways in which the experience of places maps onto accounts of places drawn from sources other than direct experience, sources such as architect’s plans, cost estimates or decibel recorders. This is not simply a change in

The purposive evaluation of places

83

terminology, as important as that may be, but has implications for the direction and methods of research and the application of that research. The identifiable referents of environmental interaction, then, are the social and the physical. These are not considered distinct orthogonal dimensions but ways of classifying the components of place experience. Since these fundamental aspects of place experience are not seen as orthogonal dimensions, it is not surprising that the research literature based on factor analytic and principal component procedures does not make this distinction. Researchers who have eschewed the factor analytic approach reveal some interesting nonorthogonal distinctions. In his review of housing satisfaction, Rapoport found it of value to classify the components of evaluation under the headings of ‘social’ and ‘physical’. In their overview of approaches to environmental evaluation, Friedmann et al. (1978) place as central the study of what they label (a) ‘the users’ and (b) ‘the proximate environmental context’. In her study of the quality of life of young men, Levy-Leboyer (1978), using correspondence analysis, identified what can be best translated as ‘needs’, which she put under the headings of ‘full social life’ and ‘environmental comfort’. Her remit did not deal directly with the experience of places, but with the quality of life in general, supporting the possibility that what is being explored in place evaluation is one aspect of the more general processes underlying people’s life experiences. What also emerges from this literature is that the physical component of human experience of places usually has two identifiable aspects. One relates directly to the spatial component, with all its connotations of demarcation and privacy, while the other relates to the services and comfort which a place provides. Levy-Leboyer distinguishes between ‘environmental comfort’ and ‘secure personal space’. The fact that these areas of human physical experience have been dealt with independently by researchers (they are handled in introductory textbooks under quite separate headings) supports the contention that they are important and distinct aspects of place experience. They are always co-present in any place. In summary, this consideration of the referents of place experience provides a threefold classification: (a) social aspects, (b) spatial aspects and (c) those aspects relating to environmental services, especially those services relating to comfort and convenience. The degree of focus of the referent

Beyond the referents themselves, additional aspects of place experience can be identified. For example, in their recent exploration of the contribution of the concept of place to the study of settings, Stokols and Shumaker (1981) highlight the degree to which a person’s association with a place is ‘specific’ or ‘nonspecific’, whether there is a regular association between people and particular places or not. In presenting this conceptualization, Stokols and Shumaker (1981) are concerned with the description of settings. Yet if the perspective is changed to that of the experience of places and the purposive nature of that experience is taken into account, then the experience itself can be seen to be related to the objectives a person might have for being in that place.

84

David Canter

In efect, the specificity of experience of a place is a reflection of the degree to which the place contributes to objectives a person might have. Specificity includes the facilitation of those objectives and the hindering of them, as when a person is in a place against his or her will. A number of objectives exist within any given place experience, and aspects of any particular experience can be examined to see the extent to which they are central to the objectives of that setting or the degree to which they are focused on the major referents of that setting. The term focus is especially apt here, when considered in relation to Craik and Appleyard’s (1981) defense of the value of Brunswik’s lens model for environmental assessment. I suggest that some aspects of place experience bring important distal and proximal variables more readily into focus than do others. Some physical cues, being more central to a place’s experience, have higher ecological validity and so are more ‘in focus’. The concept of the degree of focus is valuable and can be readily illustrated. For example, a nurse in a hospital ward can consider the design of that ward in relation to how well it contributes to the provision of care to patients in bed (high focus). On the other hand, a nurse can be asked about the various facilities on which she needs to draw in order to provide that care, such as the design and location of the utility room (low, or indirect, focus). The distinction between levels of focus should not be confused with the technical issue of whether a question is focused or not, in the sense of multiple choice as opposed to open ended. Indeed, some open-ended questions may deal more directly with the essence of a place and may be more highly focused than those that are very precise questions. For example, the question ‘How do you like living here?’, although open ended, may well deal with the essence of the experience of that place, but the question ‘State on a seven-point scale how convenient the location of the shops is in your neighbourhood’ is less focused on the essence of the experience of the neighbourhood. As can be appreciated from these two very diferent examples of hospital wards and neighbourhood, the form the variation in degree of focus can take is likely to be a function of the type of place being studied. In an informal setting, this is likely to be a particular/general distinction. In a more highly serviced setting, such as a hospital, it is likely to relate to the ancillary/principal activities of that setting. A taxonomy of place experience (seen in degrees of focus) does not lend itself to distinct categories in the same way the referents do. Instead, a continuum from high to low degree of focus can be seen, which will divide on the basis of the practicalities of any situation. Another important property of this aspect is that it does not have a logical existence on its own in the same way that the referent does. The ‘degree of focus’ can only vary in relation to some particular object or referent. It modulates the referent. This has an important advantage, which will be seen when empirical examples are discussed, namely, that the focus can be used as a way of identifying what is central to the purposes of an individual in any given place. Level of interaction

Beyond the focus and the referent of experience of places, the most commonly discussed property of places is their hierarchical relationship to one another. Stokols

The purposive evaluation of places

85

and Shumaker (1981) show how the idea of hierarchy is inherent in the early writings of Barker (1963), and Russell and Ward (1982) take this quality of places as an aspect for special mention. Yet beside the discussions of hierarchy of place in Rapoport’s (1977) and Canter’s (1977a) work, there has been little theoretical consideration of the implications of environmental hierarchy for the experience of places. The pervasive existence of a place hierarchy does pose some fundamental questions about the human experience of place. For example, are the goals or objectives a person might bring to the evaluation of a place diferent at diferent levels in the hierarchy? If this is so, how do the diferent goals interact with each other if any given place simultaneously has diferent requirements of it, depending on what part of the hierarchy is considered? Difering hypotheses about the possible psychological consequences of the hierarchy of place have been put forward in some detail by Canter and Rees (1982) in the context of housing satisfaction, but they can be stated more generally. They argue that the nature of the hierarchy proposed by diferent researchers can be very diferent. They cite a comparison of Rapoport’s (1977) views on what place hierarchy means with those of Canter’s (1977a). In essence, Rapoport proposes that levels in the hierarchy are nested within each other; those places closest to the individual, such as a room (the innermost), and larger places, such as a city, are conceptualized as surrounding the inner places. This is, of course, a direct geographical metaphor for place experience; the experience is seen as a direct reflection of the geographical relationships between places. Canter (1977a) argues that the geographical relationships are only one component of place experience. He suggests that the critical aspect of the hierarchy of a place is the level of interaction it implies for an individual. In other words, levels in the hierarchy imply diferent ranges over which an individual interacts with a place. From this perspective, the levels of the hierarchy have a linear order to each other. No one is seen as central to all the others. The comparison of these two approaches is important for a number of reasons. First, it shows that the notion of hierarchy can have diferent substantive meanings in diferent formulations. This is likely to have theoretical as well as practical consequences. Second, the contrasting hypotheses derived from the diferent formulations are open to direct empirical test. The testable hypotheses become apparent when the diferences of each formulation are examined more closely. Rapoport sees each of three hierarchical levels as an outgrowth of an experientially prior one. The level closest to the individual he calls the ‘behavioral environment’, and the next level away from the person he calls the ‘perceptual environment’. His model implies that the experience of the latter is mediated, and presumably influenced, by experience of the former. The third level away from the individual Rapoport calls the ‘operational environment’. The model implies that the experience of this is a function of the experience of the ‘behavioral’ and ‘perceptual environments’. In other words, for Rapoport, satisfaction with the ‘behavioral environment’ is likely to be most closely correlated with satisfaction with the perceptual environment and that, in turn, will relate to satisfaction with the operational environment.

86

David Canter

Because the behavioural environment is primary, Rapoport’s formulation would predict that satisfaction with this environmental level would be the best single predictor of satisfaction with other levels. This suggests that items referring to satisfaction with the behavioural environment would have a higher average correlation with all other items than would items dealing with other levels. The correlations for the operational environment items would be expected to have the lowest average correlations, and perceptual environment questions would be between these two in their average relationship to all the other items. A nested set of correlations is predicted to parallel the nested relationships between the environments. If questions about satisfaction are thought of as points in a space and the correlations between questions are thought of as inverse distances between the points, Rapoport’s geographical analogy predicts a distribution of points with the behavioural environment questions at the center of the space, the perceptual environment in a region around them and the operational environment questions around them. In opposition to this, Canter presents a view of environmental experience in which each level has a separate focus. In this model, the mode of environmental interaction at the large scale is similar yet distinct from the mode at the small scale. Thus, although satisfaction with each level of the place hierarchy is seen as being most closely related to satisfaction with those levels directly adjacent in scale, no primacy is accorded to any one level. In terms of the spatial parallel, the diferent hierarchical levels are predicted to move across the space. No single set of items is central to all the others. Canter’s model of place hierarchy, then, predicts groupings of correlations at each hierarchical level, but these groups would be arranged along some axis, not nested within one another. The facet approach Discussion of the diferent formulations of place hierarchy serves not only to illustrate distinctions between the purposive model of evaluation and other formulations but to also show that detailed, empirically testable hypotheses can be derived from the purposive model. However, because of the complexity of the purposive model, it is necessary to harness an approach to theory development that will facilitate the conversion of the formulations into directly testable hypotheses. In order to convert the multiple classification scheme of referent, focus and level into a specific set of related hypotheses, the facet approach to theory definition is of great value. It is important to emphasize, however, that the facet approach only provides a structural framework. It is not an explanatory theory itself in the normal use of the term. Facet theory, as such, is a metatheory, a theory about how to formulate and test theories. In the present context, facet theory is a vehicle for developing and clarifying the purposive model of place evaluation. This model is the basis for an explanatory theory of evaluation. In order to avoid confusion between the purposive evaluation theory and facet theory, the former will be referred to as the purposive model and the latter as the facet approach. The facet approach was developed by Guttman (1965), described in detail by Borg (1977) and Shye (1978) and reviewed in relation to applied psychology by

The purposive evaluation of places

87

Canter (1982, 1983a). Within this approach, a theory is taken to consist of three parts. The first is a formal, detailed definition of the domain of concern. The second is empirical evidence that observations within the domain of concern have a structure in accord with the definition. The third is a rationale for the correspondence between the definition and the observations. The facet approach to theory shares much with many other approaches. Any diference resides in the emphasis given to the definitional system. In common with the work of Toulmin (1958), elaborated by Bromley (1977) in his approach to case studies, facet theory requires a precise definition of the research problem. The properties of this definition have direct implications which can be examined by nonmetric, multidimensional scaling procedures, such as smallest space analysis (SSA), thus allowing a direct empirical test of certain aspects of the proposed model. One fundamental principle used in applying a facet design to a research problem is the proximity principal: the more similar observations are in terms of how they are defined, the more closely they will be related empirically. These examples illustrate a practical application of facet design, that of generating specific observations. In this instance, the observations are questionnaire items, but the approach can be used efectively for content analyzing open-ended data or as a fruitful post hoc examination of an existing data set. The specification of facet elements that are being explored in any particular set of observations is also the first stage in providing a precise definition of the overall problem. As illustrated later, the empirical test of the proposed facet structure is whether the relationships between questions are predicted by the facet elements to which they belong. To take the previous example again, question (1) (House – General) shares a common element with both question (2) (House – Specific) and question (3) (Neighborhood – General). However, it has no common element with question (4) (Neighborhood – Specific). Therefore, the prediction would be that question (1) is more highly correlated with question (2) and question (3) than with question (4). Of course, this prediction would only hold for the population in general. Any particular subsample, especially one chosen because of its specific housing experiences, may well generate idiosyncratic patterns of correlations. Such patterns might, as a consequence, be of particular interest in revealing important aspects of the experience of that subsample. The multidimensional analysis (SSA) represents the observations as points in a space, with the degree of relationship between the observations being represented by the inverse of the distance between the points (the closer the points, the more highly correlated). The prediction of where a point will lie in the space is derived from determining to which elements of each of the facets the observation, represented by this point, belongs. Interpretation of the spatial analysis is, therefore, most efectively achieved through identifying facet elements. The empirical test of the existence of an element is whether or not items containing it form a discreet region in the space with the questions sharing that element being found within that region. The proposed facets can be further clarified by specifying the relationship between the elements of a facet. The major distinction is whether the elements form

88

David Canter

a logical order amongst themselves or are unordered. In general, elements that are quantitatively diferent form ordered facets. Facets whose elements are qualitatively diferent are unordered. For instance, in the example earlier, both facets would be expected to be ordered because one is increasing in specificity and the other in level of interaction. However, the social, spatial and service elements of the referent facet have no obvious, logical order to them. The relationships between the facet elements lead to predictions of the locations of the regions in the space. Using the principle of proximity, items that have the same elements will be found in the same region of the space, and elements which are most similar will have regions that are adjacent. These relationships will be described more fully later. The total definitional system being applied to the research problem is obtained through specification of all the facets and the relationships between the facets. Using the facet approach in this manner, a model of place evaluation can be proposed, the validity of which can be directly tested empirically. The previous discussion of purposive evaluation can now be expressed as the definition of three facets: the referent facet (R), the focus facet (F) and the level facet (L). Together, these three facets can be used to define, or prescribe, any set of observations used for the evaluation of places. The common range of satisfaction

One further step in the argument is necessary before turning to empirical examples. In order to define the relationships between facets, it is necessary to establish their correspondence to a further facet, common to them all, which has a specified ‘common range’ to it. Given that facets are frequently used to generate questions, the common range is often provided by the answers to those questions. A common range is taken to mean that there is a common direction to the answers of all questions that make up the domain being studied and that this common direction has the same meaning for each of the questions. In place evaluations, this common range may be regarded as the degree to which a person feels his or her purposes are facilitated by being in a given place. If the answers to a set of questions are all expressed in terms of the extent to which the place being evaluated contributes towards or hinders a person in achieving goals or purposes (i.e., how satisfactory or unsatisfactory it is), then, provided this range of satisfaction has a similar meaning for all questions, it can provide the common range for mapping the relationships between the facets that constitute the questions. The general mapping sentence (GMS)

Specifying the three purpose facets and the common range provides the basis for a more formal statement of the nature of evaluation. In efect, it is being posited that a place evaluation involves an assessment by people of the extent to which their use of places enables them to achieve their purposes. Using the notation of facet theory, each of the purpose facets can be related to each other. This set of verbally connected facets, mapped into a common range, is called a mapping sentence (Borg, 1977). It provides the principal mechanism for specifying

The purposive evaluation of places

89

the formal definitional system for which a correspondence with empirical observations will be sought. The general mapping sentence shown in Table 8.1 is sufciently detailed to use as a framework for formulating specific questions. For example, a person can be asked of the place in which he lives: ‘How satisfied are you with the general friendliness of the neighbourhood?’ (F2 r1 l3) ‘How satisfied are you with the heating system in your house?’ (f3 r3 l3) A nurse on a hospital ward could be asked: ‘How well does the layout of the bedspace help nurses treat the patients?’ (f1 r1 12) or ‘How convenient is the access to the pantry?’ (f2 r2 l3) The sample questions here derive from a more detailed elaboration of the elements within any facet than given in GMS and upon the specification of a particular context. They serve to illustrate the principle involved. The cylindrex hypothesis

There is one further property of the model provided by the three purpose facets that needs to be mentioned. The multivariate structure specified by the three facets of the form described is three dimensional and hypothesized to be cylindrical, as shown in Figure 8.1. The reasons for this cylindrical hypothesis derive from the relationship that exists between the facets. The degree of focus is hypothesized to modify the referent facet. Thus, these two facets will be found in the same plane of a multidimensional space. Furthermore, the referent facet is regarded as having qualitatively, not quantitatively, difering elements. The most direct structural form to correspond with this is a radial, or circular, configuration (Degerman, 1972). If the referent of the interaction is represented as the wedges of a circle, then their modification by a simply ordered facet would correspond to concentric zones of that circle. Together, these two facets form an adix in two dimensions (Guttman, 1957; Lingoes and Borg, 1977). The third facet, level of interaction, being independent of the other two, would correspond most directly to a further dimension. Two hypotheses can be generated for the structural relationship of such a facet to the other two. As discussed earlier, the hierarchy of place could be seen to modify the referents. In this case, it would be found as an elaboration, in the same plane as the other two facets. However, following Canter’s (1977a) hypothesis of the hierarchy of place experience, such a facet linked to a radex forms the axis of a cylinder, and each stratum of the cylinder is a radex covering a diferent range of interaction. The structural model for the content domain is thus hypothesized to be a cylindrex. Such a structure has a number of interesting properties. For clarity, these will be discussed once examples from actual data have been presented.

90 David Canter

PERSON (x) evaluates the extent to which being in PLACE (p) facilitates

FOCUS - F

[f1 [f2 [f3

the overall essence ] the general qualities ] specific aspects ]

by stating that

Table 8.1 General mapping sentence

REFERENT - R

of his/her

[r1 [r2 [r3

social ] spatial ] service ]

LEVEL - L

objectives at the

[l1 [l2 [l3

Local ] Intermediate ] Greater ]

[ It greatly facilitates ] [ ] [ ] his/her objectives [ ] [ ] [ ] It interferes with to Where (p) is a place of which person (x) has direct experience

Levels of interaction

The purposive evaluation of places

R1

F1

R3

R2

91

L1

F2

R1

F1

R3

L2

R2

F2

R1

F1

R3

L3

R2

F2

R1

F1

R3

L4

R2

F2

Figure 8.1 Schematic representation of the general model of evaluation (see Table 8.1). NOTE: F = the degree of focus facet; R = the referent facet; L = the level of interaction facet.

Context-specific methodology To carry out studies to test and elaborate the cylindrical theory of place evaluation, as summarized in the general mapping sentence, it is necessary to contextualize the mapping sentence and to derive specifications of particular observations that can be made in the given context. When the observations take the form of questionnaire

92

David Canter

items, as often is the case, then the general mapping sentence acts as a framework for the conduct of pilot interviews and as a template for the generation of questions. These questions must, of course, emerge from the language and conceptualizations of those who will be called upon to answer them so that, as in all survey research, careful piloting is essential. The piloting must also find which of the many possible questions that can be formulated in accordance with the facet specifications do make sense to the people concerned. The first goal of this exploratory exercise is to identify what the particular facets and their elements mean within the context being studied. The second is to generate a number of questions for each of the diferent combinations of elements and to select those that are most appropriate to the people and place being studied. Fruitful analysis deals with a number of questions for each combination of elements, helping to specify the range of particular issues covered by each combination. Of course, further elaboration of the mapping sentence is also possible at this state, to postulate further facets to identify more specific questions and distinguish between questions having the same element combinations in the general mapping sentence. Thus, through a specification of the general mapping sentence for the particular places being studied and the carrying out of pilot research, an instrument is produced appropriate to the context. This instrument is based on general, context-free, yet testable hypotheses. This process can also work in reverse. An existing set of observations, say, a set of questionnaire responses, can be examined to see if they reveal the structure predicted by the general mapping sentence. Unfortunately, many sets of observations, when subjected to this scrutiny, are found to be based on such ambiguous questions or to have such large gaps in any identifiable facet structure that interpretation of the results is extremely problematic. Two studies to date have been published that provide tests of the GMS in two very diferent contexts, yet a comparison of their results has not been presented before. One is the study by Canter and Rees (1982) of housing satisfaction; the other is the study by Canter and Kenny (1981) of the evaluation of hospital wards by nurses. Further unpublished reports of studies within the GMS framework of diferent contexts, such as ofces (Donald, 1983), prisons (Canter and Ambrose, 1981) and living rooms, do exist, but the general utility of the approach suggests that a comparison of existing publications is of value at this stage. Housing satisfaction The housing satisfaction mapping sentence presented by Canter and Rees (1982) is given in Table 8.2. Here, it can be seen that the three facets from the general mapping sentence have each been directly translated to the context of housing. The questionnaire details are given in Canter and Rees (1982), including the facet element combinations for each of the 44 questionnaire items used. The questionnaire had been developed after pilot, open-ended interviews with 140 householders. The questionnaire developed from these pilot interviews was then piloted itself with 100 households. From the pilot responses, a modified 44-item questionnaire with

The purposive evaluation of places

93

ROLE The extent to which

[1. [2.

Husband ] Wife ]

(x) states that

LEVEL - L [1. The house itself ] [2. The location ] [3. The neighbourhood ] FOCUS - F [1. Overall ] [2. In general ] [3. In particular ]

REFERENT - R is satisfactory for

[1. [2. [3. [4. [5. [6. [7.

[1. Social contact ] [2. Space ] [3. Services ]

Not really at all ] Not very ] Slightly ] Fairly ] Quite a lot ] Very much ] Very much indeed ]

Where (x) is an owner/occupier evaluating the place where he/she lives Table 8.2 Housing satisfaction mapping sentence

seven-point rating scales was prepared, and two copies were sent to a random sample of 2,000 households together with instructions specifying that one questionnaire was to be completed by the husband and one by the wife. All households in which both the husbands and the wives answered all the questions were used for analysis (total N = 902 couples). A role facet

The reference to husbands and wives introduces a further facet into the place evaluation model. This is a diferent facet in that it is a classification of the respondents rather than of their responses. In the context of housing, there is an eminent logic to this facet, as discussed by Canter and Rees (1982), but it also serves to illustrate a further general facet, evidence for the existence of which is discussed by Canter (1980). It is anticipated that diferences between role groups in their evaluations could be found in any study that collected responses from diferent role groups. For the present, it serves to illustrate how additional facets, specific to a particular study, can be explored without losing the possibility of examining the evidence for general facets, such as those in the GMS. Smallest space analysis Each household was treated as one respondent who had answered 88 questions (44 from the perspective of the husband and 44 from that of the wife). These 88

94

David Canter

responses were intercorrelated across all 902 households, and the resulting productmoment correlation matrix was input to a smallest space analysis (the programme SSAI was used as the most appropriate for the data matrix) as described in Lingoes (1973) and Shye (1978). Essentially, SSAI creates a configuration representing the correlations between the items as the inverse of the distances between points in multidimensional space. It does this by relating the rank order of the correlations to the rank of the distances through a series of iterations. Thus, in the final configuration, the closer together any two points are, the more highly correlated are the items which those points represent. As mentioned earlier, the proximity principle postulates that the more similar questions are in their facet elements, the more highly intercorrelated they will be. Thus, it is a central hypothesis that contiguous regions of the SSAI space will correspond with the elements of each facet. If a set of regions (i.e., a partitioning of the space) cannot be found for any facet, then there is no empirical evidence for that facet. Furthermore, the structure of the regions (how they are arranged in relation to each other) provides a more specific test of the hypothesized relationships between the facet elements. The approach, then, converts the mapping sentence into a number of related regional hypotheses which are tested by examination of the SSAI configuration. Results In order to establish whether there is any correspondence between the empirical structure of the data as reflected in the SSA and the facets of the mapping sentence, each projection of a four-dimensional1 SSA was examined to see if any partitioning of the projection revealed regions relating to the proposed facet elements. The results of this examination will be presented separately for each facet. Facet L – Level of Interaction. Figure 8.2 shows the projection of the SSA space which most clearly reveals three regions corresponding to the three elements of this facet: the house, the location of the house and the neighbourhood. The order of these facets, increasing in scale, is also as predicted. There is strong evidence, then, for this three-way classification of housing satisfaction questions. The structure of the regions in Figure 8.2 also demonstrates the difering conceptions of the relations between these levels of interaction. As discussed earlier, Rapoport (1977) proposes a ‘concentric’ relationship, whereas Canter (1977b) suggests a simply ordered structure. The structure in Figure 8.2 strongly supports Canter’s (1977) hypothesis. The three levels of interaction are on a continuum; none is paramount in influencing the others. The Role Facet. Because each question was answered separately by husbands and wives, the SSA space can also be examined to see if there are distinct regions for each role group. Figure 8.3 shows the partitioning of a projection which clearly shows two distinct role-related regions. Only one of the questions answered by the wives is found in the region of the husband questions. None of the questions answered by the husbands are in the wife region. There are, thus, clear diferences in the conceptualizations of the husbands and wives.

The purposive evaluation of places

95

54 64 20 10

HOUSE

63 61 62

18

66

6 22 45

58 46

14 8

81 12 70

76

52

32

25

24

7 28

56

60

35 42

21 79 39

63 63

27

17

26

2

72

50 71

86

83

41

5

49

23

LOCATION

15 44

95

31

67

13 86

33

19 9

81 75

1

37

69

43 40

59

34

11

36

57 63 77

87

4 84

78

53

29 38

80

55

3

16 30 48

47

62

73 60

NEIGHBOURHOOD 74

Figure 8.2 Projection of SSAI space showing partition for level of interaction facet for housing satisfaction.

Facet F – The Degree of Focus of the Interaction. Questions were categorized as reflecting three degrees of focus, being three elements of facet F. In terms of regional hypotheses, the most general questions are predicted to be at the center of the configuration and the least at its periphery. Those questions of moderate generality are expected to form a band between these two extremes. The projection showing the regional structure relating to facet F is presented in Figure 8.4. There, the overall (f1) questions can be seen to form an elliptical region at the center of the configuration. The reason for the shape of this region can be discerned from the fact that the f1 items for the husbands are some distance from those for the wives, generating two epicenters to the ellipse. The region around this central area contains all the f2 items, as predicted, supporting the three-element structure of facet F. However, a number of items from element f 3 are also found in

96

David Canter

7 51

33

5

34

27

HUSBANDS

38

29

30

6

28

21

25

73

3 82 36 49

66

71

72

47

77

14

20 43 10 24

74

44

19

15

39 8

1

52

31 12 11 4 2

35 65

16

6

80 87

68

63

61 45 58 62

75

57

66

67

13 32

18 42 17

64 54

78 84 70

40 26

61

56

83 50 53 70

41 46

76 48 85 68

60

WIVES

60

55

50

Figure 8.3 Projection of SSAI space showing partitions for the role facet for housing satisfaction.

the region close to the center. This implies that these questions are more general than was assumed when assigning them to facet elements. Consequently, these are items that should be examined more closely and possibly modified in future uses of this questionnaire. Facet R – The Referent of the Interaction. The concentric structure of the degree of focus elements reflects the fact that they are modulating some other facet. The mapping sentence shows that they are modifying the referents of the interaction, facet R. It follows, therefore, that the projection which reveals the focus facet should also capture the referent facet. Furthermore, the modulating structure of facet F implies that any facet it modifies will have a radial structure in relation to it (i.e., a set of wedges with their points meeting in the central region of element f1). Figure 8.4 also shows the region found for facet R. The three regions (one for each

The purposive evaluation of places

97

Figure 8.4 Projection of SSAI space showing partitions for focus and referent facets for housing satisfaction.

element) can be clearly seen in Figure 8.4. The set of items relating to social contact are distinct from those for spatial provision, which are distinct from those for the services. The radial structure also shows that these distinctions are qualitatively different rather than quantitatively so. No continuum or simple order can be proposed for these elements. It is worth noting that such a structure makes sense for the referents of purposive activity if it is accepted that people have many diferent goals at any one time. Any simply ordered structure would imply one underlying goal. However, conventional forms of dimensional analysis, especially forms based on the assumptions of linearity of principal component analysis, would not reveal such a qualitative structure so readily. The superimposition of facet F on facet R generates a complex structure referred to by a number of authors as a ‘radex’ (cf. Shye, 1978: 334); a radex being any

98

David Canter

set of variables whose correlations reveal both a circular order and a related simple order. In simple terms, an adix is a set of correlations which reveals both concentric circles of variables and a qualitative order around those circles. Conclusions of housing study

The SSAI, then, lends strong support to the cylindrex hypothesis and the housing mapping sentence. The related yet distinct perspectives of husbands and wives are also shown, pointing to the value of exploring role diferences in evaluation in future research. Next, the rather diferent context of hospital wards will be examined as a further exploration of the general mapping sentence. Ward evaluation In order to examine nurses’ evaluations of hospital ward designs, Kenny and Canter (1981) developed a questionnaire on the basis of mapping sentences derived from the preliminary analysis of open-ended interviews and pilot questionnaires. In all,

Nurse (p) evaluates the extent to which ward (w) facilitates

LEVEL - L [ l1. Observation to ] [ l2. Movement to ] [ l3. Contact with ] [ l4. Reducing ] [ disturbance ] [ from ]

by stating that it

REFERENT - R [ r1.other people [ r2. locations [ r3. engineering [ services

FOCUS - F ] ] [ f1.direct ] care and ] in order to provide [ f2. indirect ] comfort ] to patients

[1.helps a great deal ] [2.helps ] [3.helps a little ] [4. neither helps nor hinders ] to provide nursing care [5.makes it slightly ] and comfort [ difficult ] [6.makes it difficult ] [7.makes it very difficult ]

Where p is a nurse who is evaluating the ward (w) in which she/he works: these being adult aoute wards in hospitals built since 1960. Table 8.3 A mapping sentence for nurses’ evaluation of wards

The purposive evaluation of places

99

they used three pilot stages leading to the final 93-item questionnaire. Each item was answered with a seven-point scale which was completed by 1,921 nurses. The results presented here derive only from the final survey. The mapping sentence for nurses’ evaluations of wards is given in Table 8.4. The classification of items according to this is given in Canter and Kenny (1981) and Kenny and Canter (1981). Full copies of the questionnaire are available from the author. It is apparent that in the context of wards, the general mapping sentence takes on a very diferent form from that associated with the housing satisfaction realm. The level of interaction facet generates four distinct elements, whereas only two elements can be identified for the degree of focus facet. The spatial element of the referent facet can also be seen, within the physically limited setting of wards, to relate mainly to the location of facilities. Results of smallest space analysis A projection of the three-dimensional SSAI solution is presented by Kenny and Canter (1981). As it happens in this case, the cylindrical structure can be directly perceived in one projection. Every item is contiguous with items that have the same composition of elements. Furthermore, the order through the regions of the level facet is as specified in the mapping sentence moving from 11 at the left of the space to 14 at the right. Conclusions of ward evaluation

Two clear implications for the understanding of nurses’ reactions to ward design emerge from the cyclindrex established – one from its structure, the other from its content. The structure demonstrates that nurses’ reactions have a focus to them, epitomized by the core of the cylinder. Their reactions are not the set of isolated dimensions which might be indicated by a factor analysis nor are they a set of separate categories as revealed by naive content analyses. They form an organized system of responses, an ordered set of reactions which, while operating at a variety of levels, still maintains a common core. Discussion A framework for purposive place evaluation has been drawn from the existing evaluation literature and translated into a set of empirically testable hypotheses. Two distinct formulations have been brought together in developing and testing this framework. One is the goal-directed model of place experience; the other is the facet approach to theory construction. Thus, while each of these formulations have value independently of each other, in combination, they provide a powerful basis for a theory of place evaluation. This theory has been summarized as a general mapping sentence which predicts a cylindrical structure to place evaluations. An additional facet beyond those in the cylinder (role) has also been posited. Two published accounts of evaluations in difering contexts have been considered, and both provide strong

100

David Canter

evidence for the cylindrex hypothesis derived from a general mapping sentence of place evaluation. It is, therefore, now appropriate to consider the consequences and implications for the proposed model in reference to the objectives outlined earlier. 1 Consistencies Across Places. The rationale for the cylindrex hypothesis is, in its present form, context free. Similar patterns of relationships between items which can be classified according to the general mapping sentence would, therefore, be hypothesized for other contexts than housing and hospital wards. Research is in progress that looks at evaluations of ofce buildings; but all other places, natural and man-made, are open to exploration within this framework. It would be remarkable if the framework withstood such an onslaught of investigation. It has already been seen that elaborations of additional elements within a facet and the addition of other facets are feasible and logical for any particular context. However, the question of the cylindrical hypothesis and its stability across diferent places must remain open to empirical challenge. Indeed, it is the establishing of the conditions under which it does not pertain which will be especially important for theory development. It might be anticipated, for example, that the cylindrical hypothesis would be culture specific. However, support for cylindrical hypotheses have been found in other facet studies of attitudes outside of the environmental framework (Levy, 1981; Shapira and Zevulun, 1978). Thus, the generality of the cylindrical hypothesis is worthy of exploration. 2 Diference Between Individuals. The diferences between husbands and wives found in the study of housing satisfaction suggests one hypothesis for future study (i.e., that the pattern of correlations for diferent groups may well be the same but that the focus of their evaluations may difer). However, the theoretical formulations so far are not specific enough to predict exactly what diferences between groups might be hypothesized. Given that the theoretical basis for the formulations so far are the purposes people have for being in a place, then it would be hypothesized that the major diferences between people in their evaluation of places would be related to diferences in their reasons for being in a place. As discussed elsewhere (Canter, 1979), the set of variables most directly available for study which clearly link to diferences in place-related purposes are roles, especially ‘environmental roles’ (Canter and Walker, 1980). In order to specify these hypotheses more fully, it is necessary to understand what aspect of any general model is likely to be modified by role diferences. Is it the facet structure itself or the diferences in the levels of satisfaction any individual achieves within an overall structure? 3 Consequences for Performance. The examples discussed have drawn upon questionnaire surveys of stated satisfaction. There are a number of weaknesses when this data base is used alone. Other modes of expression of experiences of places are not covered (e.g., what people do in a place, how well they perform specific tasks in a place and other expressions of well-being). Furthermore, as mentioned, Stokols and Shumaker (1981) point to the possibility that stress may be typical of places in which a simple match is not present between these diferent modes of expression. There are thus many further questions to be answered. However, the purposive model allows precise formulation of these questions.

The purposive evaluation of places

101

If performance, for example, is taken as a reflection of interactions with a place, then it would be expected to have a similar structure to satisfaction. Questions could then be specified about whether it was variations in performance at specific scales of interaction which were to be investigated or performance relating to specific referents at diferent scales. Having established this, the relationship between the performance structure and the satisfaction structure can be investigated. In efect, the mode of expression of place experience becomes a further facet. In studies which have included behavioral observations and questionnaire responses in the same analysis, it has been found that the modes appear to play a modulating facet, observations being more functionally diferentiated than verbal statements (Canter, 1982). This would mean, if established generally, that verbal response could be more readily predicted from observations of behaviour than vice versa. In order to predict actions from verbal accounts, some detailed specification of the context of the action would be necessary. 4 A Methodological Template. The way in which the general mapping sentence can be used to generate new evaluation instruments and guide pilot research has been illustrated a number of times in this article. The idea that one standard evaluation instrument is necessary for cumulative research has been thrown into doubt. The proposal that a standard evaluation instrument could be used for all studies of places has been an aspiration for many researchers since the early semantic diferential studies of environments. Yet any detailed consideration of places and their use reveals that their qualities lie in the unique aspects of the experience of them. The search after relevance which has characterized evaluation studies has implicitly recognized this by the way evaluators have felt impelled to develop instruments specific to the settings they are studying at the expense of generality. Taking the approach illustrated here, the best of both these aspirations are possible, combining comparability with context relevance. 5 Salience of Facets. Having established the basis of a multivariate, purposive model for place evaluation, the question arises as to what the contribution is of the diferent components of the model, how they might be weighted or what their ‘importance’ is. This is a far more complex set of questions than might at first appear. In understanding the way to unravel these questions, it is worth noting that the model is derived from a number of consistencies in the literature. Thus, because these earlier studies are based on a variety of diferent perspectives, there are many diferent ways of approaching the problem of salience. However, within the proposed model, there are a number of clarifications possible. The purposive framework of the model does indicate that salience or importance will be a function of the diferent goals that a person may have in any given place. By directing attention, then, to one referent or another, diferent sets of priorities may emerge. The cylindrical structure itself ofers some indications as to how salience may operate in considering the evaluation of places. For example, the axis of the cylinder, relating as it does to levels of interaction, indicates that no simple formula will reveal one level to be more important than another. However, because these are organized in a linear fashion, it is also indicated that no place can be experienced in the same way at the same time.

102

David Canter

At its simplest, this suggests that a house might either have the advantages of spaciousness or a very convenient location but is unlikely to have both. On the other hand, the radex framework does give an indication of diferential importance because it does give the cylinder its continuous ‘core’. While this ‘core’ should not be thought of as any general measure of satisfaction with a place, it can be regarded as crucial to any general evaluation. As long ago as 1941, Lewin was using a form of topology not greatly diferent from the regional analyses carried out here, producing a stratification diagram akin to a radex. He wrote, ‘The state of the whole depends more on the state of the central cells’. He also pointed out that those ‘central cells’ are ‘more “sensitive” to the state of the whole’ (Lewin, 1951: 123). He thus served to emphasize the significance of the central region of any structure of psychological relationships. But at a more general level, the model raises the question as to whether all places have an identifiable psychological core to them. Another important question is the relationship of the facets and their cylindrical structure to design variables. To date, the establishment of links between evaluations and actual design parameters has been remarkably limited. The model proposed here does, nonetheless, indicate a number of fruitful directions. Again, it is the level of interaction facet that may well bear the most direct relationship to variations in physical design, for it is here that the compromises central to design may well leave their greatest mark. Trying to optimize the benefits of general accessibility (a distant level of interaction) with space for particular activities (a close level of interaction) is central to many designs. These included designs where being able to see patients from all points must be balanced against space around the bed for nursing activity, or in housing where detached spacious plots inevitably lead to the urban sprawl that makes distances between facilities so great. The focus facet also ofers some intriguing possibilities for links to design. The metaphor of ‘central/peripheral’ is readily applied to diferences in degree of focus so that there may well be physical correlates of a similar form worth exploring. One final and particularly interesting possibility for the development of this work is the provision of criteria for classifying places according to either the facet structure or the facet content that emerges from users’ evaluations. Work currently in progress suggests that places can be distinguished on the basis of their facet structure. Indeed, in a recent publication, Canter (1983a) has suggested that a superordinate classification scheme may be possible. This would be a classification of facet structures. It is clear from this discussion that the proposed model opens up many important lines of inquiry. Its potential specification of hypotheses derived from a purposive model of place evaluation would seem to be considerable. Whether or not that potential is achieved remains in the hands of the research community. Note 1 There is no strict theoretical basis for determining the number of dimensions appropriate for any SSA. Because the interpretation of the analysis is not based on the labeling of dimensions per se, this arbitrariness is of little practical consequence. However, two broad guides to the most appropriate number of dimensions for an SSA solution are available. One is the model being explored. In the present case, the cylinder might be expected to require three dimensions and the role facet a further dimension (four dimensions in all). The second guide is a ‘stress’ measure of the fit between the rank of the correlation matrix and the rank of the distances between items in the space. Two measures of ‘stress’ are

The purposive evaluation of places

103

generally available: the Guttman-Lingoes coefcient of alienation (GL) and Kruskal’s stress (K). See Lingoes (1973) for further details. In the present case, GL was 0.14 and K was 0.13; both of these values are well within the recommended limits.

References Argyle, M., Furnham, A., and Graham, J. A. (1981). Social Situations. London: Cambridge University Press. Barker, R. G. (1963). On the nature of the environment. Journal of Social Issues, 19, pp. 17–34. Borg, I. (1977). Some basic concepts in facet theory. In Lingoes, J. C., Roskam, E. E., and Borg, I. (eds) Geometric Representation of Relational Data. Ann Arbor, MI: Mathesis. B.P.R.U. (1972). Building Performance. London: Applied Science. Bromley, D. B. (1977). Personality Description in Ordinary Language. London: Wiley. Canter, D. (1977a). Children in hospital: A facet theory approach to person/place synomorphy. Journal of Architectural Research, 6 (2), pp. 20–32. Canter, D. (1977b). The Psychology of Place. London: Architectural Press. Canter, D. (1977c). The Psychology of Place. London: St Martin’s Press. Canter, D. (1979). Y a t-il Des Lois d’Interaction environmentales? Presented to 4th IAAP, Louvain. Canter, D. (1982). Facet approach to applied research. Perceptual and Motor Skills, 55, pp. 143–154. Canter, D. (1983a). The potential of facet theory for applied social psychology. Quality and Quality, 17, pp. 75–67. Canter, D. (1983b). Putting situations in their place. In Furnham, A. (ed) Social Behavior in Context. London: Allyn & Bacon. Canter, D., and Ambrose, I. (1981). Prison Design and Use. Guildford: University of Surrey. Canter, D., and Kenny, C. (1981). The multivariate structure of design evaluation. Multivariate Behavioral Research, 16, pp. 215–236. Canter, D., and Rees, K. (1982). A multivariate model of housing satisfaction. International Review of Applied Psychology, 31, pp. 185–208. Canter, D., and Walker, E. (1980). Environmental role and conceptualizations of housing. Journal of Architectural Research, 7, pp. 30–35. Craik, K. H., and Appleyard, D. (1981). Streets of San Francisco: Brunswick’s Lens Model Applied to Urban Inference and Assessment. Berkeley: IPAR. Craik, K. H., and Zube, E. H. (1976). Perceiving Environmental Quality. New York: Plenum Press. Degerman, R. L. (1972). The geometric representation of some simple structures. In Shepard, R. N. et al. (eds) Multidimensional Scaling: Vol. 1, Theory. New York: Seminar Press. Donald, I. (1983). The Multivariate Structure of Ofce Evaluation (Unpublished M.Sc. dissertation, University of Surrey). Downs, R. M., and Stea, D. (1977).  Maps in Minds: Reflections on Cognitive Mapping  (pp. 264–272). New York: Harper & Row. Friedmann, A., Zimring, C., and Zube, E. (1978). Environmental Design Evaluation. New York: Plenum. Guttman, L. (1957). Empirical verification of the radex structure of mental abilities and personality traits. Educational and Psychological Measurement, 17, pp. 391–407. Guttman, L. (1965). A Faceted definition of intelligence. In Studies in Psychology. Jerusalem: Hebrew University Scripts Hierosolymitana, vol. 14, pp. 166–181. Guttman, L. (1968). A general non-metric technique for finding the smallest co-ordinate space for a configuration of points. Psychometrika, 33 (4), pp. 469–506.  https://doi. org/10.1007/BF02290164

104

David Canter

Ittelson, W. H. (1973). Environment perception and contemporary perceptual theory. Environment and Cognition, Princeton: Seminar Press, pp. 1–19. Kenny, C., and Canter, D. (1981). Multivariate structure of nurses’ evaluations of hospital wards. Journal of Occupational Psychology, 54 (2), pp. 93–108. Lazarus, R. S. (1966). Psychological Stress and the Coping Process. New York: McGraw Hill. Lef, H. L. (1978). Experience, Environment and Human Potential. New York: Oxford University Press. Levy, S. (1981). Lawful roles of facets in social theories. In Borg, I. (ed) Multidimensional Data Representations: When and Why. Ann Arbor, MI: Mathesis Press. Levy-Leboyer, C. (1978). Etude Psychologique Du Cadre De Vie Pari. Paris: Editions du Centre National de la Recherche Scientifique. Lewin, K. (1951). Field theory in social science. In Cartwright, D. (ed) Social Science Paperbacks. New York: Harper & Brothers. Lingoes, J. C. (1973). The Guttman-Lingoes Non-metric Program Series. Ann Arbor, MI: Mathesis Press. Lingoes, J. C., and Borg, I. (1977). Identifying spatial manifolds for interpretation. In Lingoes, J. C. (ed) Geometric Representation of Relational Data. Ann Arbor, MI: Mathesis. Marans, R. W., and Spreckelmeyer, K. F. (1981). Evaluating Built Environments: A Behavioral Approach. IL: University if Michigan: Survey Research Center. Rapoport, A. (1977). Human Aspects of Urban Form. Oxford: Pergamon Press. Relph, E. (1976). Place and Placelessness. London: Pion. Russell, J. A., and Ward, L. M. (1982). Environmental psychology. Annual Review of Psychology, 33, pp. 651–688. Seamon, D. (1979). A Geography of the Lifeworld. London: Pion. Shapira, Z., and Zevulun, E. (1978). On the use of facet analysis in organizational behavior research: Some conceptual considerations and an example. Working Paper No. 10. Pittsburgh: Carnegie-Mellon University, pp. 78–79. Shye, S. (ed) (1978). Theory Construction and Data Analysis in the Behavioural Sciences. London: Jossey-Bass. Stokols, D. (1976). The experience of crowding in primary and secondary environments. Environment and Behavior, 8, pp. 49–86. Stokols, D. (1978). Environmental psychology. Annual Review of Psychology, 29, pp. 253–295. Stokols, D. (1979). A congruence analysis of stress. In Sarason, I. and Spielberger, C. (eds) Stress and Anxiety, vol 6. Washington, DC: Hemisphere Press. Stokols, D. (1981). People in places: A transactional view of settings. In Harvey, J. H. (ed) Cognition, Social Behavior, and the Environment, Hillsdale, NJ: Lawrence Erlbaum. pp. 441–488. Stokols, D., and Shumaker, S. A. (1981). People in place: A transactional view of settings. In Harvey, J. H. (ed) Cognition, Social Behavior and the Environment. Hillsdale, NJ: Lawrence Erlbaum. Tuan, Y-F. (1977). Space and Place. London: Arnold. Toulmin, S. E. (1958). The Uses of Argument. London: Cambridge University Press. Wohlwill, J. F. (1973). The environment is not in the head. In Preiser, W. F. E. (ed) EDRA 4. Philadelphia, PA: Dowden, Hutchinson, and Ross.

9

Putting situations in their place1 David Canter

A role for the physical context We take for granted in our daily lives that human beings have a physical existence and require such environmental resources as space, heat and light for their social interactions and individual activities. The scale of resources society assigns to the physical environment is illustrated, for example, by the fact that in most countries, the building industry, which provides these facilities, is usually a major industry. Furthermore, we expect any account of the social activities in a little-known culture, either a popular account through the media or a learned one from an anthropologist, to provide some information on the physical environment, architecture and objects of that culture. The well-established discipline of archaeology, which studies social processes, is heavily based on a culture’s physical artifacts – its deserted monuments, pottery shards, ground plans and the like. Yet it would be hard to demonstrate that archaeology is less scientific than psychology or that its insights into the human condition are less valid. History books and museums, weekly magazines and daily papers, novels and films all describe and explain the physical context in which human activities occur. For example, in the opening paragraph of Time Regained, Proust (1957) describes in detail the curtains in the room in which his central character finds himself, contrasting them with the decorations in other rooms he has known. Yet Proust’s interests and purposes in writing are as abstract as those of any psychologist. He is not concerned simply with external objectivity but with human experience and the social interactions of which it is composed. The reasons for changing and developing the physical surroundings, as well as for describing the physical context of human activities, go beyond the merely functional. In other words, the role the physical environment plays in human experience, how it contributes to social interactions and ways of life, is a dominant reason for popular concern with it. To survive, human beings must be warm and dry and have a place to sleep. It is also clear that for at least the last 10,000 years, and possibly longer (Preziosi, 1979), these minimum functional requirements have been overlaid by many other roles, both psychological and social, for the physical surroundings. As part of the attempt to understand the role physical environment plays in social behaviour the cross-disciplinary subdiscipline of environmental psychology has, in DOI: 10.4324/9781003313052-11

106

David Canter

recent years, brought together psychologists, geographers, sociologists, architects, planners and others (see Canter and Craik, 1981, and Russell and Ward, 1982, for recent reviews). It is surprising that the emergence of environmental psychology is such a recent development but possibly more surprising that its links to activities in other areas of social psychology have been so tenuous (as Spencer, 1981a, discussed). The time is ripe for developing more efective links between the social and the environmental perspectives within psychology, so both can benefit from, as well as contribute to, understanding the significance of our physical surroundings. Thus, this chapter explores some bridges made possible by recent developments in both social and environmental psychology. Until recently, despite the high probability of the social psychological significance of the physical surroundings, as revealed by both popular accounts and other social sciences (see, for example, Michelson, 1970; Knox, 1975; Jakle et al., 1976; Rapoport, 1977), neither social psychology nor most other areas of psychology successfully accommodated what Margulis (1980) called ‘the objective physical environment’ into their theoretical formulation. Margulis argued, for example, in reviewing learning theory, personality theory, social psychological theory and perceptual theory that ‘in general the objective environment in psychological theory has a null status’. He argued that until recently, even perceptual theory presented a view of perception that ‘mirrored the internal world of values’. So that ‘psychologists, as a rule, have favoured points of view that stress what is inside the head, not what the head is inside of ’. Even within environmental psychology, authors noted that the physical surroundings are often given only a minimal role in the issues studied. The symposium at the APA convention from which Margulis is quoted explored this weakness in environmental psychology. Archea (1977), in discussing privacy studies, made the same point: ‘The environment presented in this literature tends to lack enduring properties which set it apart from the behaviour to which it is presumably related’ (117). It is against this background that social psychologists have directed little attention to the role of the physical surroundings (Spencer, 1981b). The sins of the forefathers

How, then, did social scientists get into this situation in which many around us appear to notice a physical environment, but we have difculty talking about it? Can the emerging situational approach help us out of this situation? One answer suggests that, almost like the biblical curse, the sins of the forefathers are being visited on their grandchildren. This curse, stemming from the laboratory tradition, has been transmitted to many environmental psychologists who have tried to shake of that tradition; there is a danger that it is finding its way into studies of situations. When psychologists in general, and social psychologists in particular, embraced the experimental laboratory approach, they also adopted the view that the laboratory is a neutral setting, playing no part in the processes to be studied. Within the laboratory framework, the actual laboratory itself – its size, shape, color, location and

Putting situations in their place

107

so on – is deemed to play no role in the experimental processes. Thus, the literature of laboratory psychology did not develop a language for describing (or a set of ways of thinking about) the physical setting’s role in the processes being studied. By the time psychologists concerned with the environment dragged their studies out of the laboratory (if not into the open air, at least to the ofce down the corridor), a set of problems and a language of study had been developed for discussing psychological issues. The new breed was stuck with attitudes, personalities, perceptions and the like, both literally and conceptually. They were stuck with thinking of the environment as another set of variables to be manipulated or controlled, variables they thought happened to be difcult to deal with in the laboratory. There thus first emerged ecological psychologists, led by Barker (1968). Environmental psychologists such as Craik (1970) and Proshansky et al. (1970) then tried to understand human activities in their naturally occurring contexts and the role the environment might play in those activities, but they had to draw heavily on conceptual tools from the disciplines from which they were trying to break away. Barker (1965) observed children’s behaviour, profoundly aware of the artificiality of the operant traditions then current. Within the accepted behaviourist perspective on the nature of scientific psychology, he studied what children did in their natural habitats. The experimenter thus had a minimal efect on what was being studied. Barker believed any overt interaction between psychologist and subject, such as asking questions, would contaminate results. Barker and his colleagues (1978), as a consequence, restricted their activities to observing behaviour in situ. Yet they were apparently unaware that this fear of contamination from unwanted interactions had driven psychologists into the laboratory in the first place. Thus, Barker, his associates and students shared with laboratory-based psychologists a narrow focus on what was to be studied and on tactics considered appropriate for such studies. Of course, not all of Barker’s students or their students were so orthodox. Recent developments in the approach have introduced questionnaire and other procedures less focused on observations (Wicker, 1979; Bechtel, 1980). Even with these developments, however, Barker’s ecological psychology still remains isolated from broader ecological perspectives within the social sciences at large (see, e.g., Richerson and McEvoy, 1976) because its conceptual tools are so restricted to the behaviourist tradition. Unlike Barker, Craik (1976) was motivated to environmental concerns by what he saw as the strengths of existing approaches within psychology and their potential for use in the untried realm of the physical surroundings. In particular, he saw the study of the environment as a development in personality measurement and developed scales for carrying out environmental assessment and inventories for distinguishing between people in their environmental responses. Craik (1981) described environmental psychology as ‘the invasion of an array of relatively distinct and mature research traditions, currently viable within scientific psychology, into the domain of person-environment relations’. Although Craik’s formulations went beyond Barker’s by including the subject’s descriptions of their environmental experiences, he still focused on ‘what is inside the head, not what the head is inside of ’. Neither Barker nor Craik’s influential writings nor the writings of many other early environmental psychologists (cf. Stokols, 1978) described actual physical

108

David Canter

settings. Proshansky et al. (1970) made the important contribution of pointing to the need actually to map places and the behaviour within them if they were to be understood. The book of readings he and his colleagues produced was one of the first to include plans and axonometric drawings of buildings as well as photographs of actual settings. But this innovation, although frequently quoted, has yet to be fully integrated into the mainstream of environmental psychology research. For example, the epitome of Proshansky’s approach as demonstrated in the detailed study carried out by his colleagues finds only few parallels in other research (Richer, 1979; Canter, 1972; Hart, 1979). A number of people active in environmental psychology for some time have questioned the initial bases of their activities. Sommer (1974), for example, questioned the practical value of the notion of ‘personal space’. Wohlwill (1973) asked whether the environment is in the head. Only a few psychologists looked directly at the role the physical environment in all its shapes and forms plays in human experience. This has meant that the applicability gap perceived to exist between designers and researchers is still an issue for discussion (Hill, 1979) and that some psychologists are not convinced that environmental psychology has any future (Taylor, 1980). The consequence of this state in environmental psychology is now beginning to leave its mark as other developments in psychology turn to this new field for assistance. A number of developments within social psychology now encourage textbook writers to include at least a section on environmental psychology. Yet it is understandable that their view of environmental psychology, based almost inevitably on its early faltering steps, should be so uninspiring. Thus, recently, Argyle et al. (1981) characterized environmental psychologists as having ‘a behaviouristic approach to experimentation’, as being ‘closely aligned to the pure social learning theory position, which holds that individual diference variables are a function of the environment’ and are underplaying ‘the role of the subject, both in selecting, avoiding, negotiating or defining social situations’ (23). These weaknesses are true of some work of some environmental psychologists, but this is not their fault. They have drawn on the same sources as other psychologists and fallen into the same traps. The reason these pitfalls are so important here is that if social psychologists buy back from environmental psychologists the concepts and procedures they bought from social and experimental psychologists a generation earlier, the same problems are likely to be recycled under a diferent label. Thus, in their recent overview of social situations, Argyle and his colleagues make the confusion of devoting a chapter to ‘environmental setting’. The opening sentence of that chapter refers to the physical environment as a ‘feature of situations’. They thus get caught in the confusion referred to earlier, of dealing with the physical environment as another set of variables to be considered in a situation, rather than recognizing that human beings inevitably have a physical presence and that the physical must, therefore, be an integral component of any situation, not simply a cause of or an efect produced by behaviour. Of course, the integral, systemic nature of the physical environment’s contribution is implicit in the assertion by Argyle and his colleagues that transactions with the surroundings are stimulated as much by the individual as by the surroundings. This was also argued by Proshansky and his colleagues (1970) and enshrined in the

Putting situations in their place

109

book Environmental Interaction in which Canter and Stringer (1975) developed a similar argument. Furthermore, by pointing out, as Argyle et al. did, that the physical surroundings often reflect the goals of the users of those settings, they showed that human actions must play a more fundamentally causal role than the physical variables themselves. By emphasizing human agency rather than responses to environmental stimuli, the starting point for any activity is looked for in the actions of people. But again, the process is set in motion whereby the physical nature of our existence is lost sight of and processes within the individual again become superordinate. The problem here is a profound philosophical one that has, like most problems in philosophy, been debated at least since Plato’s time. In essence, if a dualist distinction is proposed between the physical and the psychological aspects of reality, then it is difficult to show how or where these dual realms of existence can afect each other. Any causal link between subjective and objective reality requires their coming into contact with each other. But how can they do that if they are distinct aspects of existence? On the other hand, if we take our own experience as paramount, recognizing the physical world as a product of our own perceptions and cognitions, then it is difcult to accept that modifications of those surroundings can be of any substantial consequence. This summary of the epistemological questions central to exploring the psychological significance of the physical environment does not begin to do justice to the subtleties and complexities of more than 2,000 years of philosophical debate and the vast libraries exploring these themes. The reader wishing a further understanding of this debate should read a classic introductory text such as Russell (1927), or a more recent review such as Ayer (1982), or Pirsig’s more popular account (1974), or even the script of Jumpers (Stoppard, 1972). Within the area of environmental research, Hillier and Leaman (1973) and, more recently, Teymur (1981) argued in some detail that the lack of philosophical clarity makes the bridge between psychology and the physical environment so difcult to build. The importance of this debate can be gauged from two conflicting facts. On the one hand, as noted, most societies assign a lot of their resources to shaping the physical surroundings. Many diferent professional groups, architecture, interior design, planning and so on have been established to create and manage these physical resources. On the other hand, the few psychologists concerned with the human consequences of the physical surroundings have had difculty in establishing any clear, direct causal efects of the physical environment on behaviour. This has been true since the days of the Hawthorne investigations (Roethlisberger and Dickson, 1939). Of course, preferences and diferences in evaluative descriptions of environments can be readily established (e.g., Nasar, 1981; Espe, 1981), and studies of diferences between groups in their reactions to the surroundings are frequently published (e.g., Webley, 1981; Schroeder, 1981). However, reports of the direct efects of the physical environment on human behaviour are rare, and those published tend to be in the psychological domain (e.g., Hawkins, 1981; Rohles and Munson, 1981). This lack of support for a dualist perspective is also found when attempts are made to act on the findings of studies of the physical surroundings. As Bechtel (1980) has shown from a major review of environmental evaluations, they do not connect with the issues confronting decision makers and rarely lead to implementation.

110

David Canter

There is thus growing evidence that as long as the human and the physical are treated as separate realms of discourse and brought together only at a notional level through general models of interaction, psychologists will continue to be caught in the various cul-de-sacs of dualism. They will continue to attempt to get out of these awkward situations by studies focusing on the human components, preferences, individual diferences, conceptualizations, expectations and the like, without connecting directly with the physical aspects of experience and with a consequent weakness of potential for application. These difculties exist in any psychology still struggling with the remnants of the dualist framework of stimulus and response. It follows that they are central to the attempts of social and environmental psychologists to incorporate the physical surroundings into their theoretical formulations. The concept of place A brief review of the recent history of our understanding of the environment’s role in social behaviour provides four points on which to build the foundations of bridges between social and environmental psychology. 1 There is a prima facie case for the relevance of the physical surroundings to social behaviour. 2 Both social and environmental psychology have had difculty accommodating physical variables directly into their theoretical formulations. 3 The difculty of integrating physical variables into environmentally oriented studies has its roots in part in the experimental laboratory tradition and the epistemological dualism typically assumed within that tradition. 4 As Argyle et al. (1981) pointed out, many environmental studies, especially in North America, have been in the behaviourist tradition. People’s goals, objectives or reasons for being in any particular location are not normally considered. Following from these four points, I propose to identify units of environmental experience for which some people use the term place. The first detailed formulation of place as a technical term describing an approach to located action was detailed in Canter (1977). This approach has many general parallels to the writings of Tuan (1977) and Relph (1979) and other geographers (e.g., Buttimer and Seamon, 1980; Seamon, 1979), but there are some important diferences in emphasis. The most fruitful way to think of places in a psychological framework is as quanta of experience. As in physics, the smallest units of energy normally available are quanta of energy, and all energy can usually be released only in multiples of these quanta; it is proposed there are similarities in human experience. We think of ourselves as being in one place or another. A distinct locational component to experience is always present. However, an immediate caution is necessary; the parallels with physics are not precise. The molecules of chemistry have some properties this theory assigns to places. In other words, components of places can readily combine with each other to form new quanta of experience. Similarly, attention can be directed to larger or

Putting situations in their place

111

smaller place scales. We can focus our experience on big molecules or small ones; a city can be a place psychologically just as much as a corner of a room can. Places, then, are the major building blocks for understanding human actions in their naturally occurring context. Both behaviour settings and situations occur within places. One place may house many of Barker’s behavioural settings or Argyle’s situations at the same or diferent times. It must be emphasized, though, that places are part of experience. They cannot be specified independently of the people experiencing them. The central postulates are that people always situate their actions in a specifiable place and that the nature of the place so specified is an important ingredient in understanding human actions and experience. One goal for any science of situated human action is classifying the places in which those actions occur. How, then, are we to describe and classify places? Again, a clue can be found in chemistry. Consider the analogy of elements and how they are classified and described. Distinct elements can, of course, be named for their special properties (for example, hydrogen was so named by Lavoisier because it was seen as a maker of water). In the same way, if we have evidence for their distinctness, we can specify places, such as bedrooms, park benches, hospitals or inner cities. Like the early chemists, we may often confuse similar places and not distinguish diferent ones in our labeling. But descriptions are the key to any future improvement of labeling. In chemistry, the growth in understanding of the nature of atomic weight as a crucial descriptor of elements helped stabilize the classification system. For places, it is proposed that three sets of properties are the key to their distinctness. Taken together, these sets describe much of what is psychologically significant about a place. They are not variables to be dealt with independently and intercorrelated but aspects of places to be explored. The three constituents of places are the following: 1 The activities understood to occur at a location and the reasons for them. 2 The evaluative conceptualizations that are held of the occurrence of those activities. 3 The physical properties of the place as they are evaluated in relation to the activities. Let me emphasize again that places can be readily distinguished from behaviour settings and situations. Unlike behaviour settings, places (a) are not created by the investigator on the basis of observing behaviour and (b) have distinct evaluative and physical components. Unlike situations, at least as identified by Argyle and his colleagues (1981), places have a distinct enduring existence as well as being inevitably intertwined with the location’s physical properties. Some of the best examples of place descriptions are in the work of novelists rather than scientists. Reference has been made to Proust. At the other end of the writing spectrum, a delightful example of a place description is in the closing chapter of Milne’s The House at Pooh Corner. They walked on, thinking of This and That, and by-and-by they came to an enchanted place on the very top of the Forest called Galleons Lap, which is

112

David Canter

sixty-something trees in a circle; and Christopher Robin knew that it was enchanted because nobody had ever been able to count whether it was sixtythree or sixty-four, not even when he tied a piece of string round each tree after he had counted it. Being enchanted, its floor was not like the floor of the Forest, gorse and bracken and heather, but close-set grass, quiet and smooth and green. It was the only place in the Forest where you could sit down carelessly, without getting up again almost at once and looking for somewhere else. Sitting there you could see the whole world spread out until it reached the sky, and whatever there was all the world over was with them in Galleons Lap. (Milne, 1928: 169–172) An analysis of Milne’s description of Galleons Lap shows the following: 1 It gives pointers to the direct experience of the place’s structure and location. The position of Galleons Lap in the forest, the texture of the floor, and even the experience of so many trees that a six-year-old cannot possibly hope to count them accurately all highlight the setting’s physical aspects characterizing the experience of it. 2 It provides reference to the activities associated with the place. It is a comfortable place to sit during a walk in the forest and to contemplate the world. 3 It gives an account of the feelings and conceptions associated with the place. It is enchanted and quiet and a place where ‘all the world over’ is with you. These constituents have been isolated for identification under three broad headings here, but the experience of the place clearly is a unique blend of them all. Galleons Lap as a place can be seen as a system containing the physical location ‘at the very top of the Forest’, which is thought of as being ‘enchanted’ and where it is possible to indulge in the activity of ‘sitting down carelessly’. For study and analysis, it is necessary to divide the places’ components into their constituent parts, but the essence of the argument is that they are always components of an integrated system. It is thus always possible, in literature, at least, to re-create one, or possibly two, components of a place from others specified. For example, Drabble’s (1979) book on landscape in literature is about the conceptions and evaluations associated with particular ways of describing landscapes and how these conceptions in their turn are a function of the uses of and culturally embedded approaches to thinking about and experiencing landscape. Girouard (1978), an architectural historian, shows that the physical form of English country houses and their changes from the sixteenth century to the present can only be understood as part of changes in the pattern of activities in the house and the changing symbolic qualities important to associate with those activities. For example, Girouard argues that one of the strongest influences leading to design changes in English country houses was the increasing gap between the upper and lower strata in society, accentuated by the disappearance of the intermediate ranks, who no longer needed to put themselves under the protection of the great by entering their service.

Putting situations in their place

113

The large baronial halls of medieval times, in which all members of a vast household ate together, gave way to quarters associated with each social stratum in the household. The location, style and furnishing of each section of the house responded as much to what went on in those places as to what was believed to be the appropriate symbolic representation of their place in the household. The size and pretensions of the houses overall were ‘an accurate index of the ambitions – or lack of them – of their owners’. Thus, together with the changes in the society’s structure and in people’s expectations of their physical surroundings are parallel changes in the physical forms housing social behaviour. But note that the parallels between the physical components of the English country house and its behavioural components would not be possible, at any time, were it not for the distinct roles existing in a household and the social norms and rules guiding who does what where. The influence of role diferences and guiding rules is not limited to historic houses. Ellis (1982), for instance, showed that these social processes can be found in controlling the use of public spaces on council housing (publicly owned) estates. They provide a basis for understanding how places come into being and change. Place roles and rules Literary and historical illustrations, by their nature, do not provide any specific indications as to what psychological processes generate and maintain the experience of places. Some powerful roots in psychological theory do clarify how places form units of experience. The notion that people can cope with their life experiences insofar as they can develop an articulated structure for conceptualizing those experiences has origins in Mead (1967) and has since been elaborated by many authors (e.g., Brittan, 1973). For psychologists, the stimulus of Kelly’s (1955) personal construct theory provided the impetus for harnessing this model to explain social behaviour. A quarter of a century after Kelly’s major publication, it is still instructive to read. His statement of the communality and sociality corollaries of his theory gives direct understanding of how the personal experience of places can still be part of something as public and socially shared as the physical environment. In his communality corollary, Kelly states that the similarity of the psychological processes of any two people is a function of how similar their construction of experience is. It follows that if two people expect the same relationships between physical and social events (construing their experience of these events in a similar way), then the meaning to them of that combination of events will be similar. The sociality corollary emphasizes that the possibility of one person’s playing a role in the social processes of any other person derives from the first person’s ability to construe the construction processes of the other person. Thus, two people must realize that they expect the same associations between physical and social events for them to be able to use a place in relation to each other. If a lecturer sits at the back of the lecture hall anticipating that the students will sit on their desks and face him, then until they share an understanding of his

114

David Canter

conceptualization of the place, they will not be able to use it appropriately. It follows from the sociality corollary that places are shared aspects of experience. They are components of our conceptual systems deriving validity from their similarity to those of other people who experience the same places. This is how it is possible for places to be part of public discourse and awareness yet essentially components of individual, personal conceptual systems. Although personal construct theory provides a useful starting point for understanding places, it still relies on the relatively primitive mechanisms of people recognizing associations between the constituents of places in order to build up a conceptualization of those places. Primitive though it is, this lays the framework for a powerful self-fulfilling prophecy. To use any place, an individual must be aware of the pattern of activities that might be expected in that location and of how satisfactory those activities are likely to be. The person is thus likely to use the place in accordance with those expectations, thereby creating the basis for others to recognize and support the expectations the individual had. However, if this were the only mechanism available for developing conceptualizations and the related actions of places, then our experience would be hard and painfully won and would not be subject to the changes that are apparent. The elaboration of these primitive processes comes fruitfully from another aspect of Kelly’s theory that evolved within the framework of the ethogenic approach as espoused by Harré (1979). This approach recognizes the significance of people’s goals or purposes. Two consequences of taking a goal orientation have implications for understanding place experience, as they do for studying situations in general. One is that the people’s roles in any setting will influence how they conceptualize that setting. The other consequence is that systems of social rules will be drawn on and transmitted so the use of places can be an efective part of the society’s fabric. Whether an open area is regarded as a public path or a private garden is in part determined by the accepted rules governing that place and the ease with which the people who experience that place can recognize or learn those rules (cf. Ellis, 1982). The stability of place rules

The existence of socially negotiated expectations of what happens in places leads to the predictions that relatively stable rules of place use can be identified, that these rules will relate to the place users’ interpretation of the physical clues a place provides of its use. Such predictions are, of course, open to empirical test. Furthermore, only through empirical investigation can the particular existing place rules be identified and classified. With the enhanced hindsight this view of place provides, we see that many empirical studies often presented under the heading personal space are, in fact, demonstrations of the rules associated with particular places. Much relevant empirical evidence in the pioneering reports of such studies by Sommer (1969) noted how regularly the sequence of seat selection at library tables was governed by rules of maximizing distance and minimizing potential eye contact. Other studies, such as

Putting situations in their place

115

Delong’s (1970) account of seating in a committee room, also show place rules clearly in operation. Rosengren and DeVault’s (1963) fascinating account of how a mother-to-be in a maternity hospital is treated diferently as she moves closer to the delivery room also clearly shows how consistent rules of social interaction relate directly to physical location and its associated design and layout. However, these studies, evolving within the social psychological framework, pay little attention to the salient features of the physical surroundings in which the behaviour being studied actually occurs. We are told little of the location, shape or size of the library tables, or of the physical details in many other studies of personal space. Even Altman (1975), in his later summary of these and related studies, with an eye to the significance of the physical surroundings, finds it difcult to indicate how the surroundings actually contribute to the development and reflection of role differences and the establishment of rules of place. Intriguing studies by Sommer et al. (1981) have begun to rectify this weakness. He noted how the physical structure and organization of food shopping facilities, notably supermarkets and farmers’ markets, create a context where diferent social interaction patterns are expected both among customers and between customers and the salespeople. These empirical studies reveal the existence of a rich set of place rules, but research is needed to address this issue more directly. Such research need not be only anthropological; various forms of field experiment are also possible. One example of such a study going further to show how the rules of a place can be established and modified in relation to physical form can be drawn from a study presented in detail in Canter (1969). In that study, students were invited into a lecture room in which chairs were arranged in rows. As the students entered, a lecturer standing at the front of the room, with his back near the wall, gave them a questionnaire and asked them to sit down and take part in an experiment. A second group of students were invited in to the room; this time, the lecturer was standing near the front row of desks. In both cases, the students sat at about the same distance from where the lecturer was standing. They sat farther back in the lecture room when he was standing near the front rows of seats than when he was standing a few feet away from those seats. Many readers have experienced the phenomenon, especially with students in their first year at college. They use the room layout to distance themselves both physically and metaphorically from what they presumably find a threatening prospect for formal interaction. Particularly interesting about this study of student seat selection, beyond the distancing mechanism, was that the investigator went on to explore the consequences of altering the room’s furniture arrangement. The study was repeated with the chairs arranged in a semi-circle. In this case, the lecturer’s location did not have any measurable efect on the seats the group chose. In other words, the mechanism seemingly operating here was that what students believed would occur in the room led them to apply an appropriate set of rules to guide their use of space, in accordance with what they wished to achieve in the room. When they thought the activity was a formal one in which they could, in a sense, get away from the lecturer, they tried to optimize their distance from him. When they saw from the layout of the furniture

116

David Canter

that a more informal interaction was expected, they used diferent principles to choose seats. Note that the change in the physical structure did not directly afect behaviour. A more subtle role must be conceived for the physical surroundings than as a mere stimulus. The most interesting findings here are how the interaction between the use of the space by a person in one role group and the layout of that space is related to the use of the space by other role groups. Unfortunately, like so many intriguing studies in psychology, no replications have been published. Other parallel studies in diferent contexts suggest themselves. For example, how does a customer act if a bank manager calls him or her into the ofce (or to his or her desk) as opposed to getting up to lead the customer over? How do rowdy crowds behave in response to diferent types of movement of groups of police ofcers when areas have been demarcated for diferent groups as opposed to when they have not? The reliability of place rules

One reason for following through on the place approach to situations is to establish a level of analysis of the contexts of human action consistent and stable enough to allow classification. Beyond the evidence that the rules for actions in places are stable, it is necessary to show they do not change, except in relation to properties of the use of places themselves. This proposition can have remarkable implications, many of a practical consequence. The strongest example can be derived from looking at what happens in a building on fire. Even in extreme circumstances, well-established place-related rules continued to operate (cf. studies reviewed in Canter et al., 1980). The most dramatic example is the fire at the Kentucky Supper Club in which more than a hundred people died (Best, 1977). While this fire was in progress, waitresses showed patrons out of the building. Later interviews revealed that many waitresses only showed out people sitting at their stations. Patrons unlucky enough to be at a table where the waitress was trapped by smoke sat as smoke filled the room, while people at the next table had been shown out by their waitress. At a more general level, studies by Canter et al. (1980) of behaviour in fires revealed a consistent diference in the sequence of actions of husbands when compared with wives. Typically, the husband attempts to fight or investigate the fire, whereas the wife is more likely to inform others and get away from the fire. These diferent patterns clearly relate to the diferent role/rule structures within the average household. They have the consequence that if anybody is likely to get hurt in a fire through inappropriate actions, it is the husband rather than the wife. On the other hand, studies in hospitals, where the nursing staf is predominantly female, do not reveal a sex bias in the activities carried out in a fire. Diferences relate closely to a person’s position in the hierarchy and how that hierarchy normally functions in relation to locations within the hospital. The contribution of the physical component of places

The role/rule perspective on places and their experience explains why it is difcult to establish direct and simple efects of the physical environment on behaviour. If

Putting situations in their place

117

the physical component reflects, facilitates or inhibits the evolution of rules for the diferential pattern of place uses by varying role groups, then changes in the role structure or developments in the rule system would produce diferent patterns of place use in the same places. But this does not mean that the physical form of place is irrelevant. The reverse is true; it always plays a role in the definition and evolution of a place. Transposing role/rule relationships to new physical forms is likely to start a process resulting in a diferent pattern of role/rule relationships. It follows that challenges to the researcher are to (a) identify consistent placerelated rules, (b) classify the varieties of places commonly occurring and (c) establish the roles that the components of places play in developing place use and place experience. We now turn to the first of these challenges, taking as an example a commonly experienced place – the home. Places in the home In considering the home, one question is whether the unit of place experience described as a home can be shown to have in it a patterned structure of place use reflecting the various rules constituting living in a house or apartment. Canter and Lee’s (1974) study provides a useful basis for this exploration. It used data from Japanese high-rise apartments and thus is of interest because the identification of rules is often easier for a culture obviously diferent from that of the researcher. The perception of patterns in a foreign setting is less likely to be dulled by familiarity. Japanese houses

Western visitors are often told that Japanese households are completely flexible: beds can be taken from cupboards and rolled out on floors; low, portable dining tables can be moved into a corner. Thus, no rules as to which space be used for what purpose seem to apply. As Nishihara (1967) argues: The notion of continuous and uninterrupted spatial flow pervades even the storage spaces, which are treated much as the actual living spaces. . . . Since these interiors conform to a number of needs they actually serve no function at all. When a bed is needed, the Japanese bring it in, and when a table is required, they bring that in too. In other words the Japanese house is functionally flexible. This view is antithetical to the theory of place use outlined earlier, so Canter and Lee’s (1974) study is instructive. This study used plans of 120 Japanese apartments, with the furniture in the apartment recorded on the plans. By content analyzing the furniture, it was possible to see which pieces of furniture were likely to be found together in the same room across the plans. In efect, two stages in the process are being explored. The first stage assumes that people buy furniture and arrange it in their rooms in accordance with the activities they wish to occur in that room. The second stage assumes that to facilitate

118

David Canter

activities within a household, it is necessary to avoid housing conflicting activities by designating separate places for incompatible activities. Thus, by looking at the types of furniture grouping, it is possible to identify what sorts of conflicts are being avoided and to find out what place rules exist. Put another way, the study allows the listing of patterns of objectives people have in their house – the goals or range of intentions they are striving towards when they move into and furnish a house. This study of Japanese furniture shows, in efect, an archaeological account of the relationships between the activities in the home. An MDS analysis of the data showed first that some pieces of furniture are more general or possibly flexible in their use. Such furniture as a large Japanese cushion or charcoal brazier has high correlations across all rooms, on average, with all other pieces of furniture, showing that they are not special to any particular locations in the house. On the other hand, some furniture is more specific in its use and more likely to be found in only one location – such as a dining table or a bed. This finding gives a first clue to underlying principles of space use in a Japanese household. Some activities are considered general, open ended and compatible with other activities. In contrast, some furniture seems specific in the activities associated with it. This concept of an activity’s specificity has many implications for looking at situations. Some situations have clear rules; other situations have only loose, general rules. It seems probable that people learn to cope with these diferent sorts of situations in diferent ways. Looking at the more specific situations, using room labels for simplicity, a living room, kitchen, bedroom, study and dining room emerge from the results. It is possible also to recognize a distinction within the kitchen of a utility room and a cooking facility. In the study area, desk-related activities can be distinguished from activities more related to recreation and music. Nishihara’s statements are, therefore, not found to have empirical support when modern Japanese apartments are used to generate data. His ideas are thus either an abstraction based on general ways of thinking about Japanese domestic life or relate only to special houses used by the rich (or poor) in the past or both. Also note that although the actual furniture in Japanese houses difers from that in the West, the structure of activities it reveals is recognizable. Can similar patterns be established with studies carried out in Britain? To answer this question, a number of studies have been conducted. The structure of living room situations

Tagg (1974) asked undergraduates in Scotland what activities they did in which room in their house. He obtained a similar structure to that found in Japan. In a more recent study, Kimura (1982) looked at English houses. In particular, he became interested in the curious British institution of the living room. He approached 100 households in the Guildford area, gave them a list of activities and asked how frequently they were likely to occur in their own living rooms. Using the same procedure as for the Japanese data, Kimura found a clear distinction between general activities likely to occur in most living rooms and thus seemingly compatible

Putting situations in their place

119

with other activities and more specific activities likely to diferentiate between living rooms. He thus found evidence for a pattern of place rules being used. It was possible to distinguish living rooms on the basis of whether they were more likely to house such activities as eating, or studying, or listening to stereo, or cleaning, or making things. A generalized schematic representation based on the MDS analyses referred to is presented in Figure 9.1. The basis of this representation is that the closer together any two activities are in the diagram, the more likely are they to occur in the same locations – that is, the more compatible are their place rules. The diagram’s circular structure is a product of the essentially qualitative distinction between the activities rather than a distinction based on quantitative dimensions. Figure 9.1 thus parallels such representations as Munsell’s color system (see, e.g., Boyce, 1981: 12) in which hues of colors form a circle going through blue, purple, red, yellow, green and back to blue. In Munsell’s circular representation, the direction from the center of the circle to its periphery represents increase in chroma, the grey combination of all colors being at the center and the most colorful colors at the edge. In the schematic representation of activity compatibilities in Figure 9.1, the general activities that can occur anywhere are at the center and the highly place-specific activities are around the circumference of the circle. The schematic summary of the ways in which activities are distinguished from each other by rules relating to where they are likely to occur has interesting implications for which combinations of activities are most feasible. Activities close together in the diagram are potentially more place compatible than are those farther apart. For example, certain arrangements can be easily accommodated: the living/ dining room, the study/living room, the study/bedroom. The farther away any two

Study (Workroom/Den)

Relaxation (Bedroom)

Recreation (Living Room)

General Eating (Dining Room) Prepartion (Kitchen)

Figure 9.1 Schematic representation of domestic activities and their place compatibilities derived from MDS analyses.

120

David Canter

situations are from each other in this model, the more likely they are to be difcult to coalesce. It can thus be hypothesized that sleeping and eating or studying and cooking in the same space are more likely to be problematic. These hypotheses are open to direct empirical test. The schematic diagram has other uses, especially when seen as a summary model of a social system. For example, does the nature of the general activities in the center change across cultures? Are there special patterns of coping when the notionally incompatible activities are forced together? How do participants’ understanding of the physical structure of place-specific activities shape their use of spaces? One problem now is to find situations in which the model does not hold; these are likely to be situations governed by diferent rule structures. The structure of building interpretations Similarities between the structure of place rules have been found for living rooms, apartments and houses. The question, therefore, follows whether diferences at the scale of the buildings within a city also relate to a general conceptual structure and, second, whether this structure bears any relationship to that found for domestic experiences. It can be argued that buildings, like rooms in a house, can accommodate combinations of some activities more readily than others. A school may not be expected to house activities associated with blocks of apartments, whereas it might be expected that ofces and factories were similar to each other. One important issue emerging here is whether building designs represent the range of compatibilities of activities the buildings accommodate. Do architects make a university look more like an ofce block than a school because they are referring to a structured relationship between the rules governing the use of those places? One direct way to study this is to show people pictures of buildings and to ask them what the buildings are and then to derive correlations from the resulting agreements and disagreements. In other words, illustrations typically thought to represent similar types of buildings can be placed together in some multidimensional space. Young (1978) carried out such a study in Britain, with a number of refinements; complementary yet independent studies were carried out by Groat (1982) in California and by Krampen (1979) in Germany. These studies reveal an organized, structured relationships between what buildings look like and the types of activities and institutions those buildings are expected to house. In efect, these studies show that a building’s physical form can indicate the type of role/rule structures governing the patterns of behaviour in the building. A clear illustration of how the interpretation of buildings has a coherent structure can be derived from Young’s (1978) study. He asked people to produce drawings representing each of six typical types of buildings. He then showed these drawings to another group and asked them to guess which types of buildings they were. Each drawing was scored on the basis of the accuracy with which people recognized its creator’s intentions. The scores were used to produce an MDS configuration, as shown in Figure 9.2. In Figure 9.2, sketches of representative drawings have been placed at points in the two-dimensional, computer-generated space. In the original

Putting situations in their place

121

computer printout, every drawing was represented by a number. However, the results are more obvious when a sample of the numbers are represented by drawings. As in other MDS diagrams, the closer together any two drawings are, the more similar are they. In this case, the more similar are the assignment of building types to them by the respondents. Figure 9.2 thus reveals a pattern of expectation of the situations that buildings may house. The pattern is reflected in the physical form those buildings take. Like the domestic activities in Figure 9.1, a qualitative order is also apparent in Figure 9.2, from houses to ofces to factories to schools to churches and back to houses. Young used only these six building types in his study, so it is an open question as to the location of such other places as hospitals, shops, trafc terminals and so on. Note, however, that in using the same six building types in Germany, Krampen (1979) produced a similar structure. The only diference was that schools were closer to ofces than to factories, presumably indicating something about cross-national diferences in attitudes to education. A question for future research is whether the structure emerging from the studies of buildings bears any relationships to the patterns found for room activities. Two general hypotheses underlie this question: 1 Does the central focus of each pattern of activities in a place relate to each component of the overall structure of building types? 2 Does the pattern of building types reflect, or parallel, the pattern of activities in the home? Support for each hypothesis would have wide-ranging implications for both theory and practice. The experience of place Activities and physical description are only two components of places. A third component can be described as the feelings a person has about a place. The most direct way this can be articulated is to ask a person how satisfactory a place is, the extent to which he or she is attracted towards it. Despite the limitations to what people can put into words, either due to their ability to verbalize their reactions or their willingness to publicize private feelings, using language to describe places is such a developed activity in our culture (as revealed by earlier reference to Proust, Drabble and Milne) that asking people about their delight or distaste with particular places is often the most fruitful starting point for exploring the success or failure of place uses or place forms. But even if a nondirective, less interventionist approach is taken to place evaluation, as in Bechtel’s (1980) ecological explorations or the studies of room furniture discussed earlier, the afective component of place experience is still implicit. Place satisfaction provides a valuable key to understanding the structures described so far. The patterns of rules existing for places are rules created to enable people to live a pleasant, acceptable, comfortable existence. If the rules did not

122

David Canter Churches Schools

Houses

Factories

Apartments

Offices

Figure 9.2 Representation of MDS analysis of building types. Source: Young (1978)

Putting situations in their place

123

facilitate the satisfactory use of a place, they would not persist. We, therefore, need to know more about the structure of satisfactions people have with their places because these attitudes are a guide to the motivations leading people to enter into the sorts of informal social negotiations and contracts that make place rules possible. In a sense, any discussion of the social psychological principles making up the structure of people’s experiences of places must be complemented by the individual motivational components that provide the organizations of the objectives or intentions people have in using places. A number of studies carried out within the framework considered here (Kenny and Canter, 1981; Canter and Rees, 1982) indicated the consistency with which people use the physical surroundings to help them achieve their objectives within that situation. Also consistent is that the core objectives characterizing an individual’s interaction pattern in any place are distinct from one place to another. Focal processes thus seem central to place satisfaction. For example, in housing, the general pleasantness and welcomingness of the house frequently enshrined in the quality of the living room appears to be crucial (Canter and Rees, 1982). Other facilities, such as the heating system or the amount of space for parking, are ancillary to this core construct. If, on the other hand, a hospital is considered, it is found that the patient-nurse contact, particularly at the bedside, forms the central core for a nurse’s satisfaction with a hospital ward (Kenny and Canter, 1981). A second recurrent finding is that the structure of the facilities ancillary to this central core is at least threefold, dealing with the following: 1 social issues, in other words, contacts with other people or displays made for public consumption; 2 issues to do with the amount of space available; and to do with 3 services provided, not necessarily based in one location, the fabric or infrastructure making the place possible. These aspects of satisfaction with a place imply a model of people as both social and having a physical existence. They must be accommodated in a certain size space with particular properties and a particular environment created by the services. There is thus an implication that people are motivated both by their social existence and their existence as physical objects. Roles and satisfaction

If it is accepted that people are part of a social context in which they are trying to achieve acceptable levels of satisfaction by interacting with other people in particular places, then it is clear that their roles in those places and how they interact within them will difer from person to person. It is, therefore, feasible that the type of satisfaction they will have with a place will difer according to their role in that place. The term role is potentially ambiguous. I have attempted to clarify it by writing about an environmental role (Canter, 1977) – the pattern of interactions a person has in any particular place. Elsewhere, I summarized a number of studies clearly

124

David Canter

showing that the pattern of interactions a person has in a place will relate to his or her place satisfaction (Canter, 1979). One study, for example, illustrated the similarity and diferences between diferent male inhabitants of a school who were asked to evaluate that school (Gerngross-Haas, 1982). Her results showed that the school parts to which the individual has access or his responsibilities within the school distinguish him from others in terms of his evaluation. Indeed, the more diference between people in their environmental role, the more likely they are to have difering satisfactions with their settings. What, then, are the implications of role diferences in satisfaction for the structures of rules in diferent places? One type of answer can be drawn from a study by Canter and Walker (1980). In England, Walker interviewed all people involved in creating a public housing estate, asking them what their major concerns about the housing estate were. She then examined the similarities and diferences among individuals on the basis of their concerns. Some people were concerned with the building’s administration, some with the building fabric itself and some with it as an aspect of the suburb’s housing requirements. People also difered in terms of how much interaction they had with the actual building. Clearly, the rules they emphasized about the housing facility’s creation difered because of their diferent types of interaction with the entity being produced. The whole notion of housing meant something diferent to each role group. Bridging social and environmental psychology In summary, this chapter argued that both environmental psychology and social psychology have sufered from an inability to accommodate efectively the physical surroundings in their formulations. The developing importance within social psychology of the study of situations, when taken together with the models of place experience presented here, probably provides fruitful bases for developing bridges between social and environmental psychology and, in so doing, provides a more efective role for the physical components of place within social and environmental models. The experience of place has been characterized as having three integrated components: activities, evaluations and physical form. The system of experiences of any given place brings these components into interaction with each other because of the rules people recognize for the use of places. These rules are both a function of the role a person has in any given place and are motivated by a person’s desire to achieve role-related objectives. The extent to which people feel able to do this is indicated by their evaluation of a place. The framework outlined here, then, draws attention to the interrelationships between the following: 1 2 3 4

rule-guided place use; the roles people have in those places; place evaluations; and the structured meanings of the physical forms of places.

Putting situations in their place

125

Much empirical and theoretical work must be done to elaborate fully and to clarify this framework, but the consistencies to which it has already pointed are encouraging. To conclude, this chapter considers the basis this framework provides for integrating developments in social and environmental psychology. Social skills and place experience

Before considering some of the more general theoretical bridges now possible, a more specific link to understanding social skills is of interest. Argyle et al. (1981), in particular, argued that understanding situations has much to ofer to social skills training. He makes it clear that being able to perform social skills efectively involves recognizing a particular situation’s structure and the roles within it. One difculty with a strictly situational approach is that the variability and changes characteristic of situations have led to an inability to provide an acceptable classification scheme for situations. Thus, much social skill training has focused on particular and specific events, such as how to be assertive when told to work during a lunch break, or how to apply for a job on the telephone (cf. Wilkinson and Canter, 1982), or how to greet people. Of course, our understanding of social skills has developed rapidly over the last few years, as Trower discussed in Chapter 7. Until some superordinate categorization is available of the contexts in which social skills occur, there is the risk that people’s training to cope in one place will not necessarily transfer to other places. It is, therefore, tempting to ask if the overall structures of places illustrated in Figures 9.1 and 9.2 might provide the basis for classifying the broad types of social skills people must master to survive in society. Indeed, have social skills trainers tended to concentrate on living room skills at the expense of bedroom skills? Or at a larger scale, have domestic social skills been emphasized to the detriment of social skills in ofces or factories? In raising these questions, it must be emphasized that the integrated systemic nature of our place experience precludes the notion that people are diferent types of beings in diferent places. Analogous social and psychological processes run through all place experience. But even at an elementary level, it is apparent that a person’s ability to comprehend a place’s social structure from its physical form and the location of people within it, as well as to recognize the role possibilities provided by how the space is used, contributes directly to the person’s skill in performing efectively in that place. This chapter suggests two more specific sets of hypotheses beyond the general guidance that might be taken from a place perspective. First, if the qualitative relationships represented in Figures 9.1 and 9.2 are valid, then is the transfer of social skills more efective if the places between which the transfer is made are qualitatively similar? For example, are people who are socially efective in a domestic setting more likely to be efective in an ofce than in a school? Second, are people’s skills general or place specific? Two subsidiary questions arise here. One is whether the development of skills grows from one specific place, such as the home, slowly encompassing all the types

126

David Canter

of places of which a person gains experience (moving around the periphery of Figure 9.2, say), or whether people develop a general level of social skill that becomes diferentiated over time (starting from the middle of the circle). The second subsidiary question is whether there is a radical diference in gaining a command of general social skills (at the center of Figures 9.1 and 9.2) when compared to how people gain a command of place-specific social skills. Whatever the answers to these questions, and whether the place perspective is shown to be valuable in helping us understand social skills and develop ways of training people in them, it is clear that exploring the possibilities for such links will result in a clarification of both social skills and the psychology of place. Some bridges

As recent books have revealed, many students of personality diferences now accept that much of human behaviour can be explained only if the interaction between a person and the situation in which he or she finds himself or herself is also considered – the P × S debate (Forgas, 1979; Ginsburg, 1979; Harvey, 1981; Krasner, 1980; Magnusson, 1981). That this development is facilitating the broader uptake of an environmental perspective is shown directly in chapters in these books by researchers recognized in environmental psychology (e.g., Craik, 1981; Stokols and Shumaker, 1981). This chapter is, therefore, one of a growing number of bridges, and it is, therefore, probably valuable, finally, to identify the links that the place perspective outlined here appear to ofer for future research. 1 The social texture of places. In his overview of emerging strategies in social psychology, Ginsburg (1979) emphasized the ‘meanings of actions, events and settings’. Harré (1979) brings this closer to the place perspective by discussing the ‘social texture of space and time’. There thus appears to be a growing acceptance that our experience of the physical world is characterized by the assignment of meanings to locations. These meanings incorporate expectations about role/rule relationships and the resulting pattern of activities. It is also clear that to operate within these places, we must be able to recognize the place rules and roles guiding patterns of place use. Future research could explore what is expected of diferent classes of places and how understanding diferent physical forms contributes to what is expected of various role performances. Further, if the setting appears at odds with the patterns of behaviour anticipated for other reasons, what coping strategies do people adopt for dealing with these discrepancies? Stokols’s (1979) congruence analysis of stress showed some theoretical developments that may help answer this question. 2 Place cognition. Harvey (1981) saw the integration of environmental concepts into many areas of psychology as inextricably linked to understanding cognitive processes. This viewpoint can be supported directly from the environmental psychology literature by how environmental cognition studies played a predominant role in the field’s development (Russell and Ward, 1982). The place framework

Putting situations in their place

127

adds impetus to this development by highlighting the contribution an understanding of physical form and location makes to understanding the social behaviour that physical form houses. If a place always has a social as well as a physical component in cognition, then the student of place experience cannot aford to ignore developments in either social or cognitive psychology. 3 Place and situation prototypes. More specifically, the suggestion that places form units of experience has interesting parallels in Nancy Cantor’s (Cantor, 1982) argument that most people understand behaviour by drawing on a categorical structure of situations. In essence, her argument, drawn from cognitive models in psychology, is that people use prototypes of situations to the extent that a list of consensual situation prototypes can be presented. This suggests two questions in our current framework. Are there identifiable consensual place prototypes? If so, how do they relate to situational prototypes? The question of the structural relationships in each prototype group and between groups is also of interest. The importance of answering these questions can be gauged from the paradox that she recognizes of people being able to articulate situational prototypes but not appearing to draw on them directly when interpreting others’ behaviour. Perhaps by changing the level of analysis from situations to places, both in terms of prototypes and evaluated patterns of space use, a coherent perspective complementing her ‘person-oriented’ resolution of the paradox can evolve. 4 Places and social episodes. When focusing on behaviour patterns in a place, the significance of the sequence and structure of those patterns (their episodic nature) becomes paramount. Forgas (1979) mapped out the issues to consider to understand social episodes. He recognized directly that certain ‘physical components of the environment have a disproportionally important role in the definition and regulation of the interaction episodes occurring within that milieu’ (46). He also saw the environment’s symbolic meaning as a major mechanism by which it plays its role. However, he virtually restricted his conceptualization of the physical environment to ‘physical props and furnishings’. He moved from what he called ‘global environments’, such as drugstores or residential treatment institutions, at one scale to white coats or chair arrangements at the other. As a consequence, he incorporated the organizational environment either as the contribution to social episodes from the global scale or as symbolic markers giving information about who is allowed to sit where. The examples of social episodes in buildings on fire illustrate how Forgas’s valuable insights can be extended. It was argued that in a home or a hospital, the pattern of rules as to who should do what where, together with the related demarcation of roles, gave rise to and were facilitated by consistent patterns of place use. Thus, during a fire, the social episodes were a direct product of the existing patterns of place behaviour. We must know more, of course, about how place rules and social episodes complement each other. The question of whether different types of places tend to house episodes with diferent structure is also raised. In some types of places, the appropriate physical form of a place may directly parallel the structure of its characteristic social episodes. For example, as seen in Rosengren and DeVault’s (1963) consideration of an obstetrics hospital, which

128

David Canter

characteristically houses episodes having a distinct sequential order, it may be that understanding how the physical layout represents and facilitates this sequence helps us understand the stages in the social episodes occurring there. This overview of a few directions in which bridges between environmental and social psychology can be built by using the place perspective shows that much must be done. Perhaps the strongest motivations for carrying out these tasks are that the curious distinction between environmental and social concerns will cease to exist and that some other divisions in the psychological literature, such as between the social and cognitive areas, may also be challenged. Note 1 Copyright David Canter. Used by permission.

References Altman, I. (1975). The Environment and Social Behavior. Monterey, CA: Brooks/Cole. Archea, J. (1977). The place of architectural factors in behavioral theories of privacy. Journal of Social Issues, 33, pp. 116–137. Argyle, M., Furnham, A., and Graham, J. A. (1981). Social Situations. Cambridge: Cambridge University Press. Ayer, A. J. (1982). Philosophy in the Twentieth Century. London: Weidenfeld and Nicholson. Barker, R. B. (1968). Ecological Psychology. Stanford, CA: Stanford University Press. Barker, R. G. (1965). Explorations in ecological psychology. American Psychologist, 20, pp. 1–14. Barker, R. G., and Associates. (1978). Habitats, Environments and Human Behavior. London: Jossey-Bass. Bechtel, R. B. (1980). Contributions of ecological psychology to the evaluation of environments. International Review of Applied Psychology, 31, pp. 153–168. Best, R. L. (1977). The Beverley Hills Supper Club Fire. Washington, DC: National Bureau of Standards. Boyce, P. R. (1981). Human Factors in Lighting. London: Applied Science Publishers. Brittan, A. (1973). Meaning and Situation. London: Routledge and Kegan Paul. Buttimer, A., and Seamon, D. (eds) (1980). The Human Experience of Space and Place. New York: St. Martin’s Press. Canter, D. (1969). Should we treat building users as subjects or objects? In Canter, D. (ed) Architectural Psychology (pp. 11–17). London: RIBA Publications. Canter, D. (1972). Royal hospital for sick children: A psychological analysis. The Architects’ Journal, 156, pp. 525–564. Canter, D. (1977). The Psychology of Place. London: Architectural Press. Canter, D. (1979). Y à t-il des lois d’interaction environmentales? In Simon, J. (ed) Proceedings of 4th IAAP at Louvain-la-Neuve. Louvain: University Catholic. Canter, D., Breaux, J., and Sime, J. (1980). Domestic, Multiple Occupancy and Hospital Fires. D. Canter (Ed.). Chichester: Wiley. Canter, D., and Craik, K. H. (1981). Environmental psychology. Journal of Environmental Psychology, 1, pp. 1–11. Canter, D., and Kenny, C. (1975). The spatial environment. In Canter, D., and Stringer, P. (eds) Environmental Interaction. London: Surrey University Press.

Putting situations in their place

129

Canter, D., and Lee, K. H. (1974). A non-reactive study of room usage in modern Japanese apartments. In Canter, D., and Lee, T. (eds) Psychology and the Built Environment. London: Architectural Press. Canter, D., and Rees, K. (1982). A multivariate model of housing satisfaction. International Review of Applied Psychology, 31, pp. 185–208. Canter, D., and Stringer, P. (1975). Environmental Interaction. London: Surrey University Press. Canter, D., and Walker, E. (1980). Environmental role and perceived housing quality. Architectural Research Paper, Symposium on Housing Quality, IAAP, Munich, 31 July 1978. Cantor, N. (1982). Perceptions of situations: Situation prototypes and person-situation prototypes. In Magnusson, D. (ed) Toward a Psychology of Situations: An Interactional Perspective. Hillsdale, NJ: Lawrence Erlbaum Associates. Craik, K. H. (1970). Environmental psychology. In Craik, K. H., et al. (eds) New Directions in Psychology. New York: Holt, Rinehart and Winston. Craik, K. H. (1976). The personality research paradigm in environmental psychology. In Wapner, S., Kaplan, B., and Cohen, S. (eds) Experiencing the Environment. New York: Plenum. Craik, K. H. (1981). Environmental assessment and situational analysis. In Magnusson, D. (ed) The Situation: An Interactional Perspective. Hillsdale, NJ: Lawrence Erlbaum Associates. Delong, A. J. (1970). Dominance – territorial relations in a small group. Environment and Behaviour, 2, pp. 179–191. Drabble, M. (1979). A Writer’s Britain. London: Thames and Hudson. Ellis, P. (1982). Shared outdoor space and shared meaning. International Review of Applied Psychology, 31, pp. 209–222. Espe, H. (1981). Diferences in the perception of national socialist and classicist architecture. Journal of Environmental Psychology, 1, pp. 33–42. Forgas, J. P. (1979). Social Episodes: The Study of Interaction Routines. London: Academic Press. Gerngross-Haas, G. (1982). Organizational role diferences in the evaluation of an experimental school. International Review of Applied Psychology, 31, pp. 223–236. Ginsburg, G. P. (ed) (1979). Emerging Strategies in Social Psychological Research. Chichester: Wiley. Girouard, M. (1978). Life in the English Country House: A Social and Architectural History. London: Yale University Press. Groat, L. (1982). Meaning of post-modern architecture: An examination using the multiple sorting task. Journal of Environmental Psychology, 2, pp. 3–22. Harré, R. (1979). Social Being. Oxford: Blackwell. Hart, R. (1979). Children’s Experience of Place: A Developmental Study. New York: Irvington Press. Harvey, J. H. (ed) (1981). Cognition, Social Behavior, and the Environment. Hillsdale, NJ: Lawrence Erlbaum Associates. Hawkins, L. H. (1981). The influence of air ions, temperature and humidity on subjective well-being and comfort. Journal of Environmental Psychology, 1, pp. 279–292. Hill, M. R. (1979). Misunderstanding the goals of science: Myths, reconciliation, and an example. In Seidel, A. D., and Danford, S. (eds) EDRA 10 Proceedings. Washington, DC: EDRA. Hillier, W. R. G., and Leaman, A. (1973, August). The man-environment paradigm and its paradoxes. Architectural Design. Jakle, J. A., Brunn, S., & Roseman, C. C. (1976). Human Spatial Behavior: A Social Geography. North Scituate, MA: Duxbury. Kelly, G. A. (1955). The Psychology of Personal Constructs. New York: Norton. Kenny, C., and Canter, D. (1981). A facet structure of nurses’ evaluations of ward designs. Journal of Occupational Psychology, 54, pp. 93–108.

130

David Canter

Kimura, M. (1982). A Cross-Cultural Comparison of Living Room Use and Evaluation (unpublished Ph.D. Thesis, Surrey University). Knox, P. L. (1975). Social Well-being: A Spatial Perspective. London: Oxford University Press. Krampen, M. (1979). Meaning in the Urban Environment. London: Pion. Krasner, L. (ed) (1980). Environmental Design and Human Behavior: A Psychology of the Individual in Society. New York: Pergamon. Magnusson, D. (ed) (1981). Towards a Psychology of Situations. Hillsdale, NJ: Lawrence Erlbaum Associates. Margulis, S. T. (1980). An overview of the status of the objective physical environment in psychological theory. Paper presented to APA Convention, Montreal, September 4, in the Symposium The Status of the Objective, Physical Environment in Environmental Psychology, 1980. Mead, G. M. (1967). Mind, Self and Society. Chicago: University of Chicago Press. Michelson, W. (1970). Man and His Urban Environment. Reading, MA: Addison-Wesley. Milne, A. A. (1928). The House at Pooh Corner. London: Methuen. Nasar, J. L. (1981). Visual preferences of elderly public housing residents: Residential street scenes. Journal of Environmental Psychology, 1, pp. 303–314. Nishihara, K. (1967). Japanese Houses: Patterns for Living. Tokyo: Japan Publications. Pirsig, R. M. (1974). Zen and the Art of Motorcycle Maintenance: An Enquiry into Values. London: Bodeley Head. Preziosi, D. (1979). Architecture, Language, and Meaning: The Origins of the Built World and its Semiotic Organization. Berlin and Boston, MA: De Gruyter Mouton. https://doi.org/ 10.1515/9783110808674 Proshansky, H. M., Ittelson, W. H., and Rivlin, L. G. (eds) (1970). Environmental Psychology: Man and His Physical Setting. New York: Holt, Rinehart and Winston. Proust, M. (1957). Time Regained. London: Chatto and Windus. Rapoport, A. (1977). Human Aspects of Urban Form. Oxford: Pergamon. Relph, J. (1979). Place and Placelessness. London: Pion. Richer, J. (1979). Physical environments for autistic children – four case studies. In Canter, D., and Canter, S. (eds) Designing for Therapeutic Environments. Chichester: Wiley. Richerson, P. J., and McEvoy, J. III. (eds) (1976). Human Ecology: An Environmental Approach. North Scituate, MA: Duxbury. Roethlisberger, F. J., and Dickson, W. J. (1939). Management and the Worker. Cambridge, MA: Harvard University Press, 1939. Rohles, F. H., and Munson, D. M. (1981). Sleep and the sleep environment temperature. Journal of Environmental Psychology, 1, pp. 207–214. Rosengren, W. R., and DeVaults, S. (1963). The sociology of time and in an obstetrical hospital. In Freidson, E. (ed) The Hospital in Modern Society. London: The Free Press of Glencoe. Russell, B. (1927). Outline of Philosophy. London: Allen and Unwin. Russell, J. A., and Ward, L. M. (1982). Environmental psychology. Annual Review of Psychology, 33, pp. 651–688. Schroeder, H. (1981). The efect of perceived conflict on evaluations of natural resource management goals. Journal of Environmental Psychology, 1, pp. 61–72. Seamon, D. (1979). A Geography of the Lifeworld: Movement, Rest and Encounter. London: Croom Helm. Sommer, R. (1969). Personal Space: The Behavioral Basis of Design. Englewood Clifs, NJ: Prentice-Hall. Sommer, R. (1974). Looking back at personal space. In Lang, J., Burnette, C., Moleski, W., and Vachon, D. (eds.), Designing for Human Behavior. Stroudsburg, PA: Dowden, Hutchinson and Ross.

Putting situations in their place

131

Sommer, R., Herrick, J., and Sommer, T. R. (1981). The behavioural ecology of supermarkets and farmers’ markets. Journal of Environmental Psychology, 1, pp. 13–19. Spencer, C. (1981a, June). Physical determinism and environmental cognition: Two major themes of early environmental psychology which served to keep it apart from social Psychology. Paper presented to the colloquium on The Position of Environmental Psychology in Relation to Social Psychology, Paris. Spencer, C. (1981b). The new social psychology and its relation to environmental psychology. Journal of Environmental Psychology, 1, pp. 329–336. Stokols, D. (1978). Environmental psychology. Annual Review of Psychology, 29, pp. 253–295. Stokols, D. (1979). A congruence analysis of stress. In Sarason, I. and Spielberger, C. (eds) Stress and Anxiety, vol. 6. Washington, DC: Hemisphere Press. Stokols, D., and Shumaker, S. A. (1981). People in place: A transactional view of settings. In Harvey, J. H. (ed) Cognition, Social Behavior and the Environment. Hillsdale, NJ: Lawrence Erlbaum Associates. Stoppard, T. (1972). Jumpers. London: Faber and Faber. Tagg, S. K. (1974). The subjective meanings of rooms: Some analyses and investigations. In Canter, D., and Lee, T. R. (eds) Psychology and the Built Environment. London: Architectural Press. Taylor, R. B. (1980). Is environmental psychology dying? Population and Environmental Psychology Newsletter, 1, pp. 14–15. Teymur, N. (1981) Environmental discourse. London: Question Press. Tuan, Y. (1977). Space and Place: The Perspective of Experience. Minneapolis: University of Minnesota Press. Webley, P. (1981). Sex diferences in home range and cognitive maps in eight-year-old children. Journal of Environmental Psychology, 1, pp. 293–302. Wicker, A. W. (1979). An Introduction to Ecological Psychology. Monterey, CA: Brooks/Cole. Wilkinson, J., and Canter, S. (1982). Social Skills Training Manual. Chichester: Wiley. Wohlwill, J. F. (1973). The environment is not in the head. In Preiser, W. F. E. (ed) EDRA 4. Stroudsburg, PA: Dowden, Hutchinson and Ross. Young, D. (1978). The Interpretation of Form: Meanings and Ambiguities in Contemporary Architecture (unpublished M.Sc. thesis, University of Surrey).

10 Intention, meaning and structure Social action in its physical context David Canter

Actions and behaviour One of the advantages of focusing on the term action rather than the term behaviour is that a clear distinction is implied between the conventional laboratory-based psychological study of responses and motor movements (what has usually been termed behaviour) and the study of situated sequences of human activity (what I take to be the starting point for the study of social action). The great value of providing this distinction is that it enables psychologists to embrace the fact that what people do occurs within a specifiable context. It is no longer necessary to focus on the limited behaviours of an isolated organism or to develop a stilted social psychology based upon the rudimentary exchange of stimuli and responses between individuals. Instead, the locus of psychological attention can legitimately be human activities and experiences within their natural settings. From this perspective, the study of social action can be seen as a development which is very much part of a widely ranging evolution of social psychology. This evolution in which American authors such as Ginsburg (1979) have joined can be seen occurring in Britain through the work of Argyle and his associates (Argyle et al., 1981), as well as through U.S. psychologists in collaboration with Polish (e.g., Nowakowska, 1981) and Soviet psychologists (Wertsch, 1981). The studies in Switzerland by von Cranach and Harré (1982) and his German colleagues also serve to show how international are the changes sweeping through social psychology. All these authors and others have written at some length about the value of exploring the meanings of naturally occurring purposive human actions and the importance of elaborating the psychological significance of the situations in which those actions occur. Because of the complexity of the interrelated issues that emerge once the decision is made to take human action as the focus of study, it is understandable that diferent psychologists typically choose to focus on diferent aspects of the full range of important issues. Some focus on the significance of human agency (Harré, 1979), others on the consequences of actions as sequences of events over time (Forgas, 1979). Yet others focus on definitions of the contexts (or situations) within which actions occur (Cantor, 1981). However, any integration of human action must recognize that all the diferent components of actions interrelate with each other. DOI: 10.4324/9781003313052-12

Intention, meaning and structure

133

Environmental links One group of researchers within psychology that has always been faced with the problem of understanding human actions in their naturally occurring contexts is environmental psychologists (cf. Canter and Craik, 1981; Russell and Ward, 1982). Their focus has been on the role of the physical environment in human activities and experience. Even when using laboratory-based experimental procedures, they have needed to keep clearly in view the representativeness of their conditions, regarding them as simulations rather than merely as interesting stimuli to be presented under controlled conditions. The majority of environmental psychologists have deliberately eschewed the problems of validating simulations by operating within such existing settings as hospitals, schools, ofces and recreation areas. The present chapter follows these developments by focusing on those processes that are integral to social action by virtue of their physical contexts. In other words, we are asking what can be learned about social action by considering directly the physical context in which it occurs. In the present chapter, then, my intentions are (a) to discuss the role that the physical environment plays in the organization and development of social actions and (b) to illustrate that even when the physical-environmental aspect of social actions is the focus of attention, the other non-physical components of human activities must still be integrated into any account of the role of the physical context. Before moving on to consider directly the way in which the physical environment plays its roles, it is of value to emphasize three processes that are fundamental to efective social action. It is the recognition of the existence of these processes which makes it possible to understand the role of the physical surroundings. Orientation to goals intention

The purposive, active, goal-oriented nature of human activity serves to emphasize that in any exploration of what people are doing, it is important to understand why (from their points of view) they are doing it. This involves the recognition that people are in any given place for some purpose and, as van Cranach and Harré (1982) argue at some length, that they are able to draw upon some conscious awareness of what their goals are. The conscious, purposive nature of human action thus directs us to the individuals’ active utilization of their surroundings. The physical surroundings can be seen, in this light, as one of the tools available to a person for use in personal or social activities. The place significance – meaning

The understanding that any given individual assigns to the significance of his or her actions (their meaning) is a fundamental process (see the review by Groat, 1982). This makes it an especially fruitful starting point for considering the role of the physical surroundings. There is a possibility now, as there was not when architects did not accept that their buildings held meanings, to explore the way in which the

134

David Canter

meaning of our physical surroundings contributes to, or detracts from, the meaning of the actions we perform in those surroundings. The structuring of situations

Human action happens in a particular situation. In order for this to be possible, those situations require a structure. This structure exists both as a set of social rules and as a set of conceptualizations held by the participants in those situations about what actions occur where. A situation inevitably has a physical locus and physical properties. It is, as a consequence, no large step to claim that the meanings and reasons for actions can be understood and seen to derive, in part at least, from the physical properties of where they take place. The physical context as process integration

The three processes of intention, meaning and structure cannot be separated from each other in their consequences for human action at anything other than a logical level. Indeed, it is their relationship to the physical context that gives these processes their integrated form. Furthermore, this is not a coincidental agglomeration of processes; their interrelationship is essential to the human condition. To understand the integrating and integral roles of the physical surroundings, it is fruitful to recognize that one of the central questions of social psychology is how the individual, with all his or her private, subjective experiences, manages to function within, and to contribute towards, the development of a public, social world. That the actions of individuals are part of a social process is fundamental to the study of social action. Yet how are the individual and the social incorporated within each other? Part of the answer to this question comes from the fact that all individuals interact with a physical world that has a recognized existence independent of their own. This argument can be taken a stage further by accepting that the three processes summarized earlier each bear a diferent relationship to the personal and public perspectives of concern here. The intentions and objectives that a person has in any situation, although influenced by social processes, are essentially within the personal (often private) domain of that particular person. However, the meanings assigned to actions and the places in which they occur, whilst having a personal significance, nonetheless operate by virtue of social codes and frameworks. As is argued in more detail later, it is this assignment of meanings which brings into the social context the personal intentions of any given individual. One consequence of this translation of personal intentions into place meanings is that publicly available structures of place-related activities can be observed. The observation of what happens where will of course influence a person in what he or she wishes to do, thus completing an important feedback loop for the social control of individual aspirations. Some time ago, Canter (1977b) outlined a cruder version of the relationships between these three processes, citing them as examples

Intention, meaning and structure Individual Goals and Intentions

Located Patterns of Activity

Actions in Place A

Person 1

Person 2

SITUATIONAL MEANING Actions in Place B

Person 3

PERSONAL

135

SOCIAL

PUBLIC

Figure 10.1 Physical context as an integrator of personal intentions and social actions.

of a self-fulfilling prophecy that maintains the consistency of the patterns of activity found in any given place. The interrelationships implied for these processes is shown schematically in Figure 10.1. The schematic relationships in Figure 10.1 are intended only as a summary of a number of complex processes, most of which have still to be fully understood. For the present chapter, they are best seen as a set of interrelated research questions requiring empirical evidence. Environmentally located goals Social psychologists generally have recognized the importance of examining human goals, at least since Kurt Lewin pointed to the existence of ‘goals’ in the ‘life space’ and Aliport wrote about ‘becoming’. But it is those who are directly concerned with goal-directed action who have emphasized the importance of classifying the goals and intentions which are significant to people (van Cranach and Harré, 1982). The environmental perspective on the question of motives and intentions generates a diferent view of these concerns. It leads to questions such as the following: Are places distinguishable by the characteristic purposes of their occupants? Is the role of the physical setting a function of the uses that its users wish to make of it? There are many similar questions. What they all have in common is the search for an account of the organization and structure of the environmentally relevant purposes that people have. Put in other terms, these are questions about the classification of the objectives people have in their use of places. The multi-modal qualities of place experience attest to the fact that a person’s interactions within a place always have a number of related objectives, or referents. The essence of the argument here, which has been presented at length elsewhere (Canter, 1977b), is that although the experience of a place can be broken down

136

David Canter

into components for analysis, the experience itself is unitary. A person, for example, always has a physical and a social existence. Questions may be asked that direct attention to how a place contributes to the physical or social well-being of a person, without implying that the physical and the social are totally distinct systems and especially without assuming that they are orthogonal dimensions. It is this recognition that the physical and social aspects of experience can be treated as foci of attention, rather than independent domains that must somehow ‘correlate’ that frees a purposive model from the conundrums besetting environmental psychologists who follow, even if only implicitly, a stimulus (environment) response (behaviour) approach. Because the social and physical are obviously both fundamental aspects of place experience, they cannot be independent dimensions. Therefore, it is not surprising that the research literature, which has been based on factor-analytic and principal-component procedures, has not revealed this fundamental distinction. However, some researchers who have eschewed the factoranalytic approach have revealed interacting distinctions. For example, in his review of housing satisfaction, Rapoport found it of value to classify the components of evaluation under the headings of ‘social’ and ‘physical’. In their overview of approaches to environmental evaluation, Friedmann et al. (1978) place as central the study of what they label as (a) ‘the users’ and (b) ‘the proximate environmental context’. In her study of the quality of life of young men, Levy-Leboyer (1982), using correspondence analysis, identified what can be best translated as needs which she put under the headings of ‘full social life’ and ‘environmental comfort’. What also emerges from this literature is that the physical component of human experience of places usually has two identifiable aspects. One relates directly to the spatial component with all its connotations of demarcation and privacy; the other relates to the services and comfort that a place provides. Levy-Leboyer, for example, distinguishes between ‘environmental comfort’ and ‘secure personal space’. The fact that these areas of human physical experience have been dealt with quite independently by researchers, as revealed by the way they are handled in introductory textbooks (under separate chapter headings), supports the contention that they are important, distinct aspects of place experience. Yet they are always co-present in any place. In relation to goal-directed action, the importance of this distinction is that intentions always have physical as well as social referents. Furthermore, in any given situation, it is essential for these diferent varieties of referents to complement one another if the place is to be efectively used. Wanting to perform a particular act implies some knowledge of the physical conditions under which it can occur and the places available for it. Individuals’ intentions, then, have social and physical orientations. These both cohere in the efective conduct of actions. Yet unlike the essentially personal quality of intentions, actions can be observed by others. Thus, any consistent pattern of associations between intentions and locations will lead people to identify those locations with particular intentions. It is this which gives places their meanings. Situational meanings are goal driven. They depend on an individual’s understanding of the relationship between goals and the locations in which those goals can be achieved.

Intention, meaning and structure

137

Cognitive ecology The rich patterns of knowledge and understanding of who might be found where, doing what, might be called a cognitive ecology. This internal representation of an ecology of actions is seen as providing the basis for the meanings situations have. It should, therefore, be possible to provide empirical clarification of these internal representations. Such clarification would be expected to reveal consistent relationships between the physical form of places, whether they be buildings, parts of buildings or other natural settings, and the assumed functions of those places. The demonstration of such patterns has drawn heavily upon various methodological developments discussed in detail elsewhere (Canter, 1985). Because of the holistic structure that such methods present, it has been possible to compare very diferent data bases generated under varying circumstances. It is the consistency of the cognitive structures found in diferent studies that is so encouraging. Let us consider, as examples, some studies of buildings. It can be argued that any building, like a room in a house, can accommodate some combinations of activities more readily than others. A school may not be expected to house activities associated with blocks of flats, whereas it might be expected that ofces and factories are similar to each other. One of the important issues that emerges from this is whether the design of buildings, taken as a whole, represents the range of intended activities they usually accommodate and if the similarities in building form reflect similarities in assumed activities. Do architects make a university look more like an ofce block than a school because they are referring to a structured relationship between the uses of those places? One direct way of studying this is to show people pictures of buildings and to ask them what the buildings are and then to derive correlations from the resulting agreements and disagreements. In other words, illustrations that are typically thought to represent similar types of buildings can be placed together in some multidimensional space. Young (1978) has carried out such a study in Britain, with a number of refinements, and complementary yet completely independent studies have been carried out by Groat (1982) in California and Krampen (1979b) in Germany. What all these studies reveal is that there is an organized, structured relationship between what buildings look like and the types of activities and institutions that those buildings are expected to house. In efect, these studies show that the physical form of a building can be used to indicate the patterns of activity anticipated within them. The clearest illustration of how the interpretation of buildings has a coherent structure can be derived from Young’s (1978) study in which he asked people to produce drawings which represented each of six typical types of building. He then showed these drawings to another group and asked them to guess which types of buildings they were. Each of the drawings was thus scored on the basis of the accuracy with which people recognized its creator’s intentions, and these scores were used to produce a multidimensional scaling (MDS) configuration (discussed in more detail in Canter, 1985). A qualitative order was found in the MDS space: in efect, a circle from houses to ofces to factories to schools to churches and back to houses. Young used only these six building types in his study; the location of such other

138

David Canter

places as hospitals, shops, trafc terminals and so on is, therefore, an open question. But it is noteworthy that, in using the same six building types in Germany, Krampen (1979a) produced a very similar structure. The only diference was that schools were closer to ofces than factories, presumably indicating something about crossnational diferences in attitudes to education. Some limiting antecedents

Having indicated that there is a clear physical component to the psychological processes underlying human actions, it is surprising how long it has taken others to acknowledge such a literally obvious aspect of a setting. So before moving on to see how situational meanings facilitate and interact with patterns of activity in places, it is important to identify the antecedents, which have slowed the development of this perspective. Two antecedents to our present perspective help to explain why this awareness is still so novel. The first is the impact of the notorious Hawthorne Investigations (Roethlisberger and Dickson, 1939). We are still sufering from the popular misconception that these studies showed that the physical surroundings had no efect on performance. A closer examination of that piecemeal set of studies shows that the efective roles of the physical surroundings were in its symbolic properties and the types of social interaction the surroundings facilitated. Changing the lighting levels clearly carried meaning for the relay assembly inspectors, even though those changes did not have the anticipated efect on performance. Furthermore, the isolation of some workers in separate rooms clearly left its mark on the social processes. The acknowledgment of the symbolic role of the physical surroundings and of its consequences for social interaction has been substantiated by a number of separate investigators (see Levy-Leboyer, 1982, for an introductory review). Another reason why the role of the physical surroundings has not been examined by social psychologists, despite clues from the Hawthorne studies and subsequent research, can be traced to the impact of the second antecedent to our present perspective of ecological psychology. Barker (1978) and his colleagues of the Midwest field station in Oskaloosa set out to study human behaviour in naturally occurring settings. Their work was quite deliberately formulated as an attack on the domination of the laboratory-based Skinnerian approach, especially when used to study human development. The remarkable thing about all the work that emanated from the Midwest field station was that it gave no indication of the physical properties of the behaviour settings in which actions happened. Imagine Darwin writing his diaries on the Beagle without saying where he made his observations. What sense could be made of speciation in the Galapagos without knowing that they were islands? Yet it is a parallel limitation that has been imposed on our understanding of human activities. Ecology and conceptual systems Once the physical context of action reaches the level of the psychologist’s awareness, and its potential symbolic and social significance is recognized, then the structure of

Intention, meaning and structure

139

actions in diferent settings and their consistencies become topics of research interest. Indeed, the major single contribution to the environmental psychology literature by the New York group of Proshansky and his colleagues (1970) has probably been their adaptation of conventional observational techniques to include the locus of the observation, which they called ‘behavioural mapping’. The significance of this procedure has been to show quite clearly that the human use of space is not random. Furthermore, the observed patterns do have clear psychological significance, which is consistent across settings. The way in which the meanings associated with diferent places readily mirror, and thus presumably enhance, the actions that occur in those places can be illustrated by two published examples. The first example is drawn from the study by Canter (1977a) of wards in a children’s hospital. In those wards, the nurses’ station is in the middle of the ward, on one side of the aisle. Opposite it are four bedded bays. Behind it are preparation and utility rooms. The centre of the ward is definitely the nurse’s domain. At the far end from the ward entrance is a day-space/play area in which children were often observed to be on their own. Along the aisle from the entrance are cubicles large enough for a mother to stay with her child. These modern wards were certainly regarded as a great advance on the earlier Victorian wards. A questionnaire survey showed that nurses and many of the hospitalized children’s parents and the doctors expressed great satisfaction with them for their brightness and spaciousness. Nonetheless, the demarcation of domains, even though it was never made explicit, left its mark on the use of the ward and certainly contributed to parents’ anxieties about the ward. In efect, they were separated from the day-space by a region of nursing space. Behavioural mapping showed that it was very rare for parents to venture through the nurses’ domain or to be found in the day-space. Interviews revealed that parents were concerned about knowing what they could do when on the ward or where they could go. Clearly the design, and the cognitive ecology it invoked, had not facilitated communication between different groups. An obstetric example

Taking the larger scale of a hospital as a whole, a fascinating study of an obstetrical hospital carried out by Rosengren and Devaults (1963) throws some useful light on some of the more subtle ways in which the physical surroundings reflect and enhance certain activity patterns. Over a four-month period, they spent 150 hours observing what went on in a large ‘lying-in hospital’ in an eastern metropolitan area of the United States. They observed all the activities of the service and talked informally with staf in the lounges and work situations, making detailed records as they went along. Their initial concern was simply to record what occurred in the hospital. They found, however, that it was incomplete and of questionable utility to speak of the doctor-nurse relationship without specifying where those two persons interacted. Their observations of the same people in diferent settings revealed that there was a

140

David Canter

spatial distribution of types of interaction. In other words, just as animal ecologists have found that diferent species are to be found in diferent regions of an area, so Rosengren and DeVaults found that there were distinct regions of the hospital. Each region is itself set apart in several ways from the others. This segregation appears to be accomplished not only by space but also by rules of dress, of expected behaviour and of decorum. All of which indicate the dissimilarity of each place, as well as to present an image of the place that might cast both patients and staf into desired roles Thus, in the context of obstetrics, at least, a variety of distinct roles exist for the nurse. In the admitting ofce, in which no barriers such as doors existed, the staf was friendly and casual, the relationship between the medical and nursing staf being equally informal. By contrast, the delivery rooms was clearly under the control of the nurses, and consequently, the symbols of their ofce, uniforms, stainless steel and brilliant lighting were especially apparent. Examples drawn from many institutions, such as hospitals and schools, reveal similar results. Even in supposedly open-plan settings, diferent role groups are to be found, characteristically, in diferent locations. Hospital wards usually have areas where nurses are to be found, and open plan classrooms have places for the teacher. Such findings are especially interesting when taken together with the conclusions on environmental meanings presented earlier. In this light, the consistent patterns of space use can be seen as reflecting the meanings diferent places have for their users, which in turn relates to what they see as the primary purposes (or functions) of those places. Role and setting When we put together all the themes earlier, including the consistent yet multivariate part played by the physical surroundings in social action and the significance of understanding the purpose a person has in those surroundings, questions can be raised about the diferences between people in their use and conceptualizations of places. As it turns out, the query about individual diferences leads to some of the most valuable practical consequences of this approach and also emphasizes the concept of role. In this context, role can be seen as a major indication of the reason a person has for being in any given place. The pattern of activities in which a person partakes and the meanings assigned to those places and the activities they house are seen as being a function of what is required of the individual by the social system of which he or she is a part. The intentions a person has for any given setting is, as a consequence, a personal translation by the individual of the goals of the organization to which they belong. This view provides an interesting insight into the satisfaction a person has with the design and use of any given place. It suggests that satisfaction will be a function of the extent to which a person feels able to achieve his or her goals in that place. It follows, therefore, that people performing diferent roles in any given place will have diferent patterns of satisfaction with that place. Furthermore, it is precisely this

Intention, meaning and structure

141

diferentiation through the satisfaction which provides the feedback loop to modify goals and intentions in Figure 10.1. Roles and satisfaction The first empirical question to answer is whether there are, indeed, diferences between the satisfactions of people with diferent environmental roles. Canter (1979) summarizes a number of studies which show quite clearly that the pattern of interactions a person has in a place will relate to his or her place-satisfaction (Canter, 1979). One study, for example, illustrates the similarity and diferences between diferent inhabitants of a school who were asked to evaluate that school (Gerngross-Haas, 1982). Her results show with great clarity that the parts of the school to which the individuals have access, or the responsibilities they have within the school, distinguish their evaluations of the school building from others. Indeed, the more diference there is between people in their environmental role, the more likely they are to have difering satisfaction with their settings. Gerngross-Haas was able to order the users of a school building from the headmaster, through technical and teaching stafs to older pupils and then younger pupils, so that the farther apart they were from each other in succession, the less correlated were their evaluations of the diferent parts of the school. What then are the implications of role diferences in satisfaction for the structures of activities in diferent places? One type of answer can be drawn from a study by Canter and Walker (1980). Walker interviewed all the people involved in creating a local authority housing estate. She asked them what their major concerns about the housing estate were. She then examined the similarities and diferences between individuals on the basis of their concerns. Some people were concerned with the administration of the building, some with the building fabric itself and some with it simply as an aspect of the borough’s housing requirements. People difered also in terms of how much interaction they had with the actual building. The criteria they applied to the creation of the housing facility were diferent because of their diferent types of interaction with the entity being produced. Indeed, the whole notion of ‘housing’ meant something diferent to each role group. One reason why this type of finding is of such practical significance is that one of the most important role diferences with regard to the environment is whether a person is using, managing or creating that environment. The indications are that by virtue of their diferent relationships to their surroundings people will be thinking about and evaluating their environment in diferent ways. The architect consulting the manager will obtain, in the workers’ conceptualizations, a predictably inappropriate account of the role of the building. This poses organization and management problems for design briefing, which the psychologist is in a unique position to resolve. Place rules It has been argued that the spatial patterning of social action is based on the knowledge of what to do where. This appears to be related to each individual’s

142

David Canter

understanding of his or her role in a given setting. Yet there must be mechanisms for controlling these roles and not allowing personal satisfaction to become the sole feedback mechanism for shaping patterns of activities. The most viable social mechanisms for this are probably the rules associated with roles and, most notably, rules that relate to spatial behaviour. This role/rule model within the environmental framework owes much to the social psychological literature in which there is a long and rich history to the study of social rules (Douglas, 1973; Harré and Secord, 1972; Mischel, 1969). What was not anticipated is the power of this formulation when looking at the use and significance of physical surroundings. Here, there is growing empirical evidence about the stability of place rules. This evidence is possibly strongest when social systems are under threat. If an extant role/rule structure is maintained even in those situations, then its power in daily life is given even more credence. Examples of places under environmental threat, most notably houses and other buildings on fire, provide some interesting examples to illustrate these issues (Canter, 1979). For example, in one major fire, at the Summerland recreation complex, many of the staf left by the fire exit that they normally used for coming into work. They went to no efort to show members of the public out that same way. This reflected the rather distant and uninvolved nature of the relationship they had with the customers who used that place. On the other hand, in the Kentucky Supper Club fire, a quite diferent pattern of relationships between staf and patrons emerged. Here, staf in the small private dining rooms showed out people at the tables for whom they were responsible, in some cases, leading them out through the kitchens. In both cases, the form of relationships and the use of space, existing before the fire, were carried through into the disaster itself. In smaller domestic fires, an analogous continuation of role diferences is found. Men are more likely to search for others or fight the fire, whilst women are more likely to warn others and help them escape. Time, the forgotten dimension Having identified three processes that facilitate an integration of our understanding of the role of the physical surroundings in social action, it is important not to lose from view the one aspect of human action that is rarely considered in studies of the physical surroundings: the temporal quality of those actions. After all, our experience of environments beyond the scale of the room, through buildings to streets and cities, is inevitably temporal. However, one of the weaknesses of much research within the environmental psychology perspective has been the dearth of attention paid to temporal phenomena. This is quite possibly a weakness it has inherited from its forebearers in personality and social psychology. The whole notion of action involves the idea of movement and change. It is, therefore, very probable that as environmental psychologists more fully adopt models of human action, their theories will be more likely to encompass this almost forgotten dimension. Nonetheless, for the moment, the themes that link action to environment are generally concerned with stable structures and existing states, rather than with movement and change.

Intention, meaning and structure

143

Conclusions If there is one central point to be drawn from developments in social psychology as they pertain to the study of social action, it is that the person who is the focus of our study is the starting point for our inquiry. It is this person’s understanding and aspirations that we need to explore. In taking such a gift from our more social colleagues, we ofer in return the insight that by providing a physical context to the focus of our study, not only is the complexity of human action more likely to reveal its patterns but also, those patterns which do emerge are more likely to have some practical significance for human afairs. References Argyle, M., Furnham, A., and Graham, J.A. (1981), Social Situations, Cambridge University: Cambridge. Barker, R.G. and Associates (1978), Habitats, Environments and Human Behaviour, London: Jossey-Bass. Canter, D. (1977a), ‘Children in Hospital: A Facet Theory Approach to Person/Place Synomorphy’, Journal of Architectural Research, 20–32. Canter, D.V. (1977b), The Psychology of Place, London: Architectural. Canter, D.V. (1979), Y’A-T-i/des Lois D’/nteraction Environmentales?, in proceedings of 4th LP. at Louvain-ta-Neuve – Simon, ed., Louvain: University Catholic. Canter, D.V. (1985), ‘Putting Situations in their Place’, in A. Furnham (ed.), Social Behaviour in Context, New York: Allyn and Bacon. Canter, D.V. and Craik, K.H. (1981), ‘Environmental Psychology’, Journal of Environmental Psychology, 1(1), 1–11. Canter, D.V. and Walker, E. (1980), ‘Environmental Role and Conceptualization of Housing’, Journal of Architectural Research, 7, 30–35. Cantor, N. (1981), ‘Perceptions of Situations: Situation Prototypes in Person Situation Prototypes’, in D. Magnusson (ed.), Toward a Psychology of Situations: An International Perspective, pp. 229–244, Hillsdale, NJ: Erlbaum. Douglas, M. (ed.) (1973), Rules and Meanings, Harmondworth: Penguin. Forgas, J.P. (1979), Social Episodes: The Study of Interaction Routines, London: Academic Press. Friedmann, A., Zimring, C. and Zube, E. (1978), Environmental Design Evaluation, New York: Plenum. Gerngross-Haas, G. (1982), ‘Organizational Role Diferences in the Evaluation of an Experimental School’, International Review of Applied Psychology, 31(2), 223–236. Ginsburg, G.P. (ed.) (1979), Emerging Strategies in Social Psychological/Research, Chichester: Wiley. Groat, L. (1982), ‘Meaning of Post-Modern Architecture: An Examination using the Multiple Sorting Task’, Journal of Environmental Psychology, 2(1), 3–22. Harré, R. (1979), Social Being, Oxford: Blackwell. Harré, R., and Secord, P. (1972), The Explanation of Social Behaviour, Oxford: Blackwell. Krampen, M. (1979a), ‘Profusion of signs without confusion’, Ars Semiotica, 2, 327, 359. Krampen, M. (1979b), Meaning in the Urban Environment, London: Non. Levy-Leboyer, C. (1982), Psychology and Environment, Beverley Hills: Sage. Levy-Leboyer, C. (1982), Psychology and Environment, London: Sage Publications.

144

David Canter

Mischel, W. (ed.) (1969), Human Action: Conceptual and Empirical Issues, London: Academic Press. Nowakowska, M. (1981), ‘Structure of Situation and Action: Some Remarks on Formal Theory of Action’, in D. Magnusson (ed.), Toward A Psychology of Situations: An lnteractional Perspective, Hillsdale, NJ: Lawrence Eribaum. Proshansky, H.M., Ittelson, W.R. and Rivlin, L.G. (eds) (1970), Environmental Psychology: Man and His Physical Setting, New York: Ho] Rinehart and Winston. Roethlisberger, F.J. and Dickson, W.J. (1939), Management and the Worker, Cambridge, Mass: Harvard. Rosengren, W.R. and Devaults, S. (1963), ‘The Sociology of Time and Space in an Obstetrical Hospital’, in E. Freidson (ed.), The Hospital in Modern Society, London: The Free Press of Glencoe. Russell, J.A. and Ward, L.M. (1982), ‘Environmental Psychology’, Annual Review of Psychology, 33, 651–688. von Cranach, M. and Harré, R. (eds) (1982), The Analysts of Action: Recent Theoretical and Empirical Advances, Cambridge: Cambridge University Press. Wertsch, J.V. (ed.) (1981), e Concept of Activity in Soviet Isychology, New York: M.E. Sharpe. Young, D. (1978), The Interpretations of form: Meanings and Ambiguities in Contemporary Architecture, M.Sc Thesis (unpublished). University of Surrey.

11 Action and place An existential dialectic David Canter

Prelude1 A number of authors have recognized the limitations imposed by the assumed dichotomy between environment and behaviour. In this chapter, this argument will be developed to show how it can be replaced with a view of the dynamic interplay between action and place. The central argument is that there are dialectical processes which continuously create changes in the patterns of actions in relation to meanings of places. Central to these processes are conscious intentions shaped by a person’s awareness of self and role in a given context. Intentions and actions are themselves structured by place-related rules, negotiated with others who may have similar or diferent roles in those places. Expressed satisfaction with, or pleasure in, a given place is a consequence of the possibilities for achieving the role-related purposes for being there. The viewpoint advocated here has its origins in giving human agency pride of place. This was a central theme of many nineteenth-century thinkers. For example, Engels (Wertsch, 1981) wrote of human labour having the character of ‘premeditated planned action directed towards definite ends known in advance’. The British psychologist William McDougall (1908) made a similar perspective central to his writings. He saw the need to examine the goals that focus motivations if human activities are to be fully understood. Like many more recent psychologists, McDougall saw consciousness and the human mind as being central to this process of goal orientation. Unfortunately, he saw the essence of this goal orientation in innate needs, yet he did not underestimate the role of mindful processes in human activities. As he put it, ‘the essential nature of mind is to govern present action by anticipation of the future in the light of past experience’. This was taken up in the 1950s by George Kelly (1955), an American psychologist, whose ideas have been influential in both British environmental psychology and behavioural geography. Kelly said we should not have to invoke any special notions such as drives, or forces, to explain why people do not remain inert. He emphasized the dynamic, active nature of the human experience. Like McDougall, he saw the importance of our understanding the sense which people make of their past as well as their anticipation of their own possible futures.

DOI: 10.4324/9781003313052-13

146

David Canter

Allemande (Hegel’s dialectic) McDougall and Kelly and a number of present-day psychologists throughout Eastern and Western Europe all reflect an approach to humanity which was first clearly articulated in the far-reaching writings of Hegel (Singer, 1983). Hegel saw development and change as central not only to an individual’s lifelong experiences but also to the whole process of human civilization. His ideas have been so profound and influential that major themes in both scientific and political thought can be traced to his arguments; I am referring to the ideas of biological evolution and social revolution. I am not so courageous, or foolhardy, as to deliver here a disquisition on Hegel, so all I will draw from his vast and profound writings is the idea that in order to understand, explore and influence important individual, social and political processes, we must harness concepts which are fundamentally dynamic. It is important that we, self-consciously, use a language of human action in and on the world. Hegel drew on the philosophical tradition of dialectics, which goes back at least to Socrates. He showed that in order to understand any active, changing processes, it is necessary to consider at least two interacting systems. Within Hegel’s dialectics, these two systems have a natural relationship to each other such that each is a direct consequence of antagonisms (or as Mao calls them, ‘contradictions’ [1967]) inherent in the other. It is through the interplay of these symbiotic opponents that change and, in Hegel’s (and later Marx’s) terms, progress emerge. From this perspective, most of the major theories of twentieth-century psychology, sociology and anthropology have a dialectical quality to them. Just to stay within the realms of psychology, Piaget, Freud and Skinner all see development and human growth as a product of the interactions between two fundamentally distinct and inherently opposing systems. Russian psychologists, such as Leontiev (Wertsch, 1981), have taken an even more overtly dialectical perspective on human action and experience. We have much to learn from this approach to the problem of studying and shaping human actions in the world. After all, our central concern is precisely with the actions which decision makers perform on behalf of other people. Change and intervention are central to the study of the interactions between people and their surroundings. Yet an examination of the environment and behaviour literature reveals theories and findings that are stubbornly static. Our scientific metaphors, for example, are taken from the concrete and permanent. Systems of thought which are inherently concerned with variations over time, such as those within music (Karol yi 1981), are hardly ever drawn upon. The traditions of ecological psychology (Barker, 1968) with its behaviour settings and standing patterns of behaviour, the study of mental maps with their motionless two-dimensional qualities, the determination of acceptable performance for lighting or noise and the assignment of meaning to various symbolic forms, all these and the great majority of other studies are conceptualized and presented as phenomena which exist across an indefinite time period. With the notable exception of the writings of Lynch and some of the ‘chronogeographers’, such as Parkes and Thrift (1980), it is rare to find discussions of origin, modification, variations over time, evolution, development and decay.

Action and place

147

This omission has two important and related consequences. One consequence is that researchers tend to be chasing changes which others are bringing about. In the U.S., open-plan schools were studied once they were already an established part of architects’ vocabulary. In Britain, high-rise housing became a research focus when many decision makers had already recognized its failures. Even issues such as meaning and symbolism in built form only became dominant areas for research once the rhetoric of the architectural and planning professions had already accepted their importance. The second consequence is that from the practical point of view, this limitation way of thinking means that the design and planning professions stumble from one fashion to another. They are not driven by any evolving theory. An almost-random series of changes shapes towns and cities. Whether a city centre is destroyed and rebuilt from scratch, or rehabilitated and modified in relation to what already exists, is often an accident of interest charges and political whim. Yet the opposing forces, the dialectical colleagues, out of which growth and change emerge, are everywhere to be seen in the study of person-environment relationships. After all, the field is frequently identified by being called the study of ‘environment’ and ‘behaviour’. These two labels have generated their own realms of study but are seen to interact in some poorly defined way to create the world of experience. Yet this is precisely where the fundamental difculty in understanding the dialectical process lies. Hillier (1973) indicated it many years ago, when he pointed to the confusions inherent in the use of the term ‘environment’, but he did not challenge the term behaviour as well, and he did not emphasize the significance of dynamic person/place transactions. Furthermore, many of even the most avowedly interactionist students of environment and behaviour accept a Cartesian dualism in which there is the world of physical entities an ‘environment’ and there is a world of ‘behaviour’. This is a perspective which is part of the general metaphysical stance that Mao (1967) upbraids as mechanical materialism and vulgar evolutionism. This dichotomy between environment and behaviour, in which mental and social processes have no clear presence, or indeed location, is a confusion which benefits from clarification before an active framework of direct relevance to the realities of actions and decision making can be produced.

Courante (A rapid dance with forward and backwards movement) In order to produce a more active, dynamic range of explorations, I take as our starting point the fact that human experience is paramount. It is what each individual knows and does and feels within the world that creates the reality for that individual. But the problem is that such a reality cannot be brought into simple connection with a notional, scientifically measured, objective environment. The world as experienced has its own laws of objective existence. Thus, whenever we try to develop causal explanations which show environment and behaviour influencing each other, we find it difcult to proceed very far or very fast, and we frequently find

148

David Canter

very little empirical evidence for the environmental influences for which we search. Consequently, many researchers fall back upon statistical ‘levels of significance’. These often account for insignificantly small proportions of the measured variation. This is a form of hiding behind statistical tests for fear of facing the epistemological questions their work raises. The proposal here is that instead of dealing with environment and behaviour, it is more productive to deal with action and place. Action and place are both products of our experience of our transactions with the world. The notion of action is distinct from that of behaviour in many ways, but one of the most important is that actions integrate conscious objectives. A person’s acts include intentions of what it is wished to achieve with those acts. For example, it may often be more fruitful to classify actions in terms of their objectives rather than their content. ‘Walking’ may be an observed behaviour, ‘but going for a walk’ is a diferent act from ‘walking to school’. Because of the purposive nature of human actions, an act must always be directed to some entity or process outside of itself. In order to act, we must distinguish between ourselves as subjects and the objects of our acts. Action requires a distinction between the entity carrying out the act and the entity on which the action is carried out. In other words, as human beings, in order to be able to act on our surroundings, we need to make a distinction between ourselves and our surroundings. We must also, however, distinguish the active significance of difering surroundings. This is a theme central to Kaplan’s (1983) cognitive model of environmental transactions. Unfortunately, he does not point out that in order to know what is possible, we must experience the consequences of our acts. It is by acting on our surroundings that we make sense of them. This point is made quite clearly, and directly linked to a dialectical framework, in the study of the Meaning of Things carried out by Csikszentmihalyi and RochbergHalton (1981). They specify that in acting on the world, a diferentiation is being made between the world outside the individual and the world of the individual. This diferentiation is the first stage in a process which makes it more possible to integrate the world into further actions. Thus, the process of diferentiation and integration, which, in their ways, Piaget and Darwin saw as fundamental to growth and evolution, is also central to the human process of acting on the world. Grauman (1976) took this theme directly. He argued that diferentiation and integration are combined in the notion of appropriation, providing a fruitful way of exploring the nature of person/place transactions. He produced a long list of forms of appropriation, showing that actions, including everything from looking to emigrating, reveal qualities of environmental appropriation. One of the major research implications of this is that we need a fuller understanding of the categories and classifications that people use in order to be able to act on the world (Canter, 1986). We need to know what distinctions are between various building forms. How the personal taxonomies which people utilize for their location decision making are drawn upon to facilitate action, and pleasure in action/ inaction, in the world. Groat (1982) has carried out some interesting studies along these lines, but a great deal more is possible.

Action and place

149

One of the applied consequences is that passive participation in design is of little value. The border between participation and education by doing should be readily crossed. Action itself is a process of distinction and separation. But what we act on is a world which is experienced and understood through that action. Those qualities of it which are perceived as objective and independent of the individual are a product of the subjective, mindful activities of that individual. The value of dealing with human action, instead of behaviour, is that action is driven by an individual, a person. Action is the product of a particular person wishing to achieve certain objectives in a directed and intended way, even if the person is not aware of those intentions. Action thus encompasses both the concepts of objectives and goals as well as consciousness and intention. This view has remarkably old antecedents. In his Outlines of Psychology James Sully (1892) wrote: Besides the factor of active consciousness all the more complex processes of volitions have other ingredients as well. These consist of psychical antecedents. That is mental processes proceeding, as well as those accompanying the action. This antecedent factor may in general be described as a forecasting or prevision of the action itself and of some at least of its results under the form of an ‘end’. Self

Actions contain conscious components which relate to the anticipated outcomes of those actions. Harré and von Cranach (1982) have emphasized this point and shown the need to take these conscious components into account and to collect information on them in any explanation of human activity. At an even elementary tactical level, then, surveys which ask how much people like what they have, or bipolar adjectival ratings, lose contact with the fundamental questions about what people want to do in given locations and what cognitions relate to those intentions. As I have already argued, these actions are built upon a diferentiation of the individual self from the world in order to be able to integrate the world at a more general level. I decide what is a personal view and what is an audience for that view so that I can then express that view to the audience. This implies that all actions contain some component of self-clarification or self-definition. By acting on the world, we learn more of the nature of the people we are – our capabilities, our worth, our potential (Peled, 1976). It is in this sense that I see actions on the world as fundamentally existential (Valle and King, 1978). Proshansky (1983) and his colleagues were some of the first to write about the role of place identify for self-identity. The idea of the self has been elaborated by many writers on self-concepts, such as Gergen (1971). They have made clear the distinction between the self experienced in a subjective form, the ‘I’, and self as an objective, social entity having temporal and spatial existence that relates directly to other temporal and spatial entities, the ‘me’. We need to sort out the implications of these diferent facets of self for diferent person/place transactions because herein lies one clue to the strong emotional qualities places can elicit.

150

David Canter

The focus on self also helps to emphasize the critical importance of the social processes within which an individual acts. From Herbert Mead (1934) and beyond, it has been emphasized that the concept of self is in part defined by contact with other individuals. George Kelly echoes this and resonates with some of Hegel’s arguments when he points out that it is the individual’s awareness of and interactions with other selves which helps to clarify and formulate that individual’s self-concepts. Kelly sees the self as one of a number of roles of which the individual is aware. What I am arguing here is that the actions which an individual performs are structured by the possibilities made available through the role structure of which that individual is a part. Another way of thinking of this is that roles enshrine the dominant objectives which an individual feels required to achieve. As a consequence, in understanding people’s action on the world, we need to understand the role they see themselves playing and their understanding of those roles. This has direct research consequences. It shows that rather than focusing on individual diferences, on personality variations, in order to explain the diferent patterns of actions and understandings which people perform, we should be looking at their role within a given social and organizational context. Role diferences are reflected in diferent patterns of activity and contribute to diferent conceptualizations of transactions with the surroundings. Within this framework, for example, age, gender and membership of various special interest groups are all most fruitfully seen as aspects of role variations which shape person/place transactions. One of the most significant role diferentiations is between the individual who has responsibility for creating or changing places (the architect or manager) and the individual who is expected simply to sufer those changes (the user or resident). There is a superordinate relationship in which the designer, planner or architect, who has responsibility for place modification, is actually attempting to interfere with the existing person/place transactions. If the designer knows little about the patterns of action and objective, which are characteristic of existing transactions, then their design can only be inappropriate. If designers further misunderstand their own role in relation to place modification, we have a recipe for confusion and unproductive change. Rules

It is fascinating the extent to which design actually does work. The frequency with which people make use of building forms created by people with little understanding of those who will use them. The key to this consistency is not some notional ‘flexibility’ of buildings or ‘adaptability’ of human beings. It is the coherent and conscious social structuring and interpersonal negotiation which makes place use possible. What is crucial here is that daily use of places reveals that consistent patterns of human actions on the world are an observable product of the processes I have been discussing (Canter, 1984). We would expect rules and related role definitions to give rise to an observable set of consistencies. It is the short-term stabilities, the plateau which a process reaches before it changes, which can be observed in standing patterns of behaviour. Thus,

Action and place

151

ecological studies emphasize superficial qualities of human actions. The processes which generate these can be examined much more directly by determining what rules guide human actions and what objectives shape them. In doing these, remarkable consistencies across difering cultures can be found (Canter, 1986). The great interest in studying privacy in the United States can be seen as a function of the power which rules relating to privacy do have in structuring the acceptability of where people can act within that culture. More direct studies of rule formation, development and change, asking what may happen in which place, are already beginning to reveal the significance of difering building forms in diferent social contexts. The layout of a theatre requires an understanding of the roles being performed front of house as well as those backstage. Such explorations can also lay the foundations for participation in design guidance. In other words, it is now possible to turn away from identifying the amount of space needed for a particular class of behaviour, or only to allow participation in planning to focus on what activities will be housed in a given location, and turn towards consideration of the rules that will guide actions in places and to elaborate the pattern of place-role relationships which give places their structure and organization. Gavotte (A medium-paced French dance, popular in the eighteenth century) The process of diferentiation and subsequent integration through acting on the world is also mirrored in the creation by the individual of internal representations of the world as experienced. The famous quotation from Karl Marx serves to illustrate the critical relevance of this point to person/place transactions: what distinguishes the worst architect from the best of bees is this, that the architect raises his structure in his imagination before he erects it in reality. For designers as well as others, the emergent reality leads to a reshaping of the imaginative representation: bringing together an understanding of actions and the locations in which they occur. It is this integrated representation of actions within a physical context which I have called in earlier publications ‘places’. (Canter, 1977) Places are not only locations. They are categorizations of experience. They are diferentiations made by an individual in relation to possible acts to which that person may aspire. By distinguishing between a person’s actions and possible places for those actions, it is possible to create a more efective integration of the two. Consider a real example to illustrate this central theme. The managers of an organization come to the conclusion that they should increase their marketing activities. They examine their current place of work and decide that it is inappropriate for these new objectives. They have made a distinction between desired actions and

152

David Canter

current places, leading to the definition of a new place within which their new objectives can take root. However, once the new place has been created, it will take on a diferent definition both from the actions which are actually associated with it and from the social processes of place-rule negotiation. This will lead to the emergence of new objectives. These objectives, and their associated actions, will once again be distinguished from their context, requiring further modifications to the place as experienced. To the distant observer, such as an architect being briefed by a client, the previous process may look like an organization discovering it does not have enough space for marketing and commissioning an addition to its building. The architect may then be surprised to discover that the new extension, when built, is immediately assigned to central administration, not marketing at all. The dialectics of action and place, however, indicate that the amount of space involved is only a reflection of certain states of the person/place transactions. What is true of space in an ofce is true of other environmental resources, whether it is the Brazilian rainforest or urban open spaces. What we see is only a reflection of the current state of the diferentiation and integration of actions and places. Let me just reiterate the central message I am putting forward here. The research literature of person/place interactions over the past few years has shown that there are a number of consistencies and stabilities. Standing patterns of behaviour, mental maps, processes of control and space definition have all been identified. Now we move on to explore the processes of transaction and change which the stabilities identified so poorly captured. It has always been characteristic of our area of activity that our starting point is that we wish to change that which we study. We are not concerned with a solely neutral, objective stance. We are committed to improvement and development. We are committed to modification. But in the past, although we have accepted the desire to change what we study, we have done little to study what it is that changes. We must work harder at developing models that are inherently unstable. But in order to do that, we must go beyond the mechanical model of cause and efect. By taking active human agency as the point, we are logically required to see human actions as central to the change process. We have already seen that action requires the notion of a goal and intent, with their associated exploration of the internal model of the world we inhabit. In other words, by separating of a reality on which action operates, in order to help diferentiate the self from the environment, it is necessary to build up patterns of locations for potential actions and their likely consequences. Thus, just as we find consistent role-related patterns of behaviour with their associate rules and development mechanisms, we find consistent conceptualizations of places. These conceptual systems which people employ would, as a consequence, also be expected to have a structure and stability, one of which people are aware. Geographers, such as Relph (1976) and Seamon (1979, who have espoused a phenomenological framework, are some way along the road towards this perspective. The literature on mental mapping indirectly, and often in a very confused way, nonetheless

Action and place

153

demonstrates that if you ask people what happens where, they can reflect on their internal representations and generate symbolic configurations which summarize the understanding they have of the world in which they operate. We also know from many studies that physical locations are loaded with emotional connotations. The bookshops contain many general accounts of places (Blythe, 1981) that have had significance for individuals and from which meanings can be derived. Yet within our literature, there is remarkably little overlap between studies of behaviour in various settings and studies of the internal representations on which people can draw for various contexts. Yet at even a banal level, the dialectic proposed here indicates that we could fruitfully explore the relationship between behavioural maps and cognitive maps (Canter, 1977). In the writings of geographers such as Relph (1976), the concept of ‘place’ has been confused by the use of a romantic elaboration of the notion of place that is contrasted with ‘placelessness’. This is really an aesthetic judgement masquerading as a technical description. There are places that have the characteristic of anonymity, or even unprepossessing features. That is their nature as places. When I use the term ‘place’, I see it as a technical term to describe a component of experience that has distinct coexisting aspects to it. These aspects are drawn from the fact that by acting on the world, the individual makes a distinction between human activities and the world. Consequently, a place always contains, in an integrated form, an individual’s preconscious and conscious awareness of both likely actions and their outcomes and the physical form with which those actions may be integrated. We have thought too much about cognition and looked too closely at perception. Places are aspects of experience. It is important to emphasize that places can be neutral or emotionally charged. They can be experienced as highly relevant to a particular individual, such as in the writings of Lym (1980), or they can be the type of ‘placeless’ places which Relph discusses. Part of the task in influencing design is to understand the special qualities which particular places require for particular activities and goals. In exploring the way in which places are experienced, two important processes which link these experiences to human actions are worth noting. One is the symbolic qualities which places have – the way in which they represent both the individual who uses them and the type of activities which are possible there. Architects have always been aware of these qualities of places, frequently expressing this awareness through a discussion of ‘expressiveness’. Indeed, as in many other, probably all, forms of art, they have found joy and rapture in the way these expressions are actually generated. There is no opportunity here to discuss the profound and important questions of aesthetic reaction. However, I would suggest that Krampen’s (1979) arguments about the meaning which style itself carries is one step towards understanding the pleasures of place. Gigue (A lively baroque dance originating from the English jig, usually appearing at the end of a suite) Beyond the symbolic qualities of places, it is useful also to consider the way the structure and organization of places are integrated with role relationships of the people using them (Canter, 1984).

154

David Canter

Architects and planners have always been aware of the importance of understanding the relationships between places, within a building or within parts of a city. But what is clear now is that we are talking about relationships between the diferent roles which an individual might take on and the diferent roles which individuals play in relation to each other. By examining the patterns of shared objectives and the areas of conflict, we can identify the topological form which is appropriate for any given organization at any given point in time. By acting within a place, we demonstrate our separateness from it and thus increase our opportunity for controlling it. When I rearrange the furniture in a room, I reveal that the seminar I wish to hold in that room consists of something more than my own thought processes and involves a set of role relationships which have spatial implications. But in shaping the room, I create a reorganization of my experience of that place which, in turn, requires me to act on it in a particular way. Thus, my actions and my conceptualizations are always in dynamic tension. Insofar as I am sane in my experience of the world, and insofar as I share a common set of conceptualizations with others who share places with me, I will be part of a process which has short-term stability. However, this short-term stability has in it the seeds of long-term evolution, development and change. Part of the seeds of the change come from the fact that my definition of place use and action will not be identical to everybody else’s. Certainly, people who have diferent role relationships to any given place will have a requirement to negotiate the nature of that place and the actions which are permissible within it. The results of that negotiation produce a higher order stability which in its turn leads to a redefinition of the nature of the place. Such a redefinition will have biases to it and be more closely identified with some actions than with others. These biases will lead to further development and change. Further attempts to adjust the place or to redirect actions will eventually ensue. Where, then, does this leave us? A recognition that all is flux has more than methodological implications. It suggests that the results to date that have characterized person-environment transactions are only a reflection of dynamic processes at a particular stage in their evolution. That there are common structures which exist across changing circumstances is to be expected, but certainly, all those frameworks and perspectives that assume the possibility of a static set of transactions fall into doubt. This includes, notably, the concept of a brief for a design. The possibility that the form of a place can be specified at a given point in time and then designed must be questioned. Instead, the processes and structures of which the building, the neighbourhood or the city are a part must be articulated as a basis for creating a design process. The action research which Bob Sommer advocates can, in this light, be seen as far more than an adjunct to the design process but as an integral part of it. Also, the cry for conservation takes on a new light. Professor Galtung points to the need to monitor environmental modifications. But modifications are at the core of person/place transactions. They do have goals/objectives which direct them. These purposes will be reflected in the role relationships which manage them. They will accommodate conceptualizations of what is to be achieved at any given point.

Action and place

155

I am advocating an integrated set of person/place studies, which assimilate both an ecological framework and a cognitive one, which absorb the emotional components derived from pleasure in places and the possibility of conscious human volition. These studies will recognize that the places people experience have symbolic and representational components, which reflect the role relationships that their actions and those places play and the ways in which they play those roles (Canter, 1984). All this will be within a set of activities which take design decision making as an integrated part of normal events. This evolving process will lead to the overlap between design and research becoming greater, as both increasingly are seen as the same process of the interplay of action and place leading to the clarification and growth of self-definition. Some built form consequences2 of the existential dialectic 1 Relationships Between Places A strictly functional (efort-minimizing) pattern of links between locations will be less efective than one which responds to the pattern of user role diferentiations. This provides a comment, for instance, on space syntax analyses. 2 Symbolic Interpenetration The access from one place to another place is good when it reflects/symbolizes the distinctions between the activity objectives that users of those connecting places are likely to have. 3 Structured Flexibility That component of an action pattern which is quintessential to its existence should be central to a design conception. Elements of flexibility should be left to the design of those component places which are less significant. 4 Internal Objectives Within a place, the patterns of activities typical of that place will have a set of related objectives. These objectives will have appropriate meanings associated with them. The form of a place should thus reflect these activities and meanings. 5 Internal/External Relationships The identification by a person of what relates to them and what not is the starting categorization. This creates a division of what is inside a given place and what is outside. Thus, the ways in which a design handles the relationship between its inside and outside is critical. Some methodological implications 1 Conscious intentions are valuable in augmenting descript ions of behaviour. 2 Role diferences are key aspects of individual variation to be studied. 3 The rules which govern the use of places are critical aspects to examine in most studies.

156

David Canter

4 The use of descriptive and case-oriented studies is likely to be more fruitful than experimental research paradigms. Some research directions l The role of place diferentiation for those individuals whose sense of self may be less secure than average (e.g., children, the mentally ill) is worthy of study. 2 The consequences of efective action in specific places for later adjustment in other places requires exploration (e.g., Are place-linked actions in therapeutic settings predictive of later health?). 3 All the issues listed in the other tables are directly amenable to study. 4 Is a pattern language possible? It seems feasible that, at least within a given culture, there are place-action transactions which are so stable that particular physical forms can be suggested as most appropriate for them. Research can identify what these might be expected to be from an examination of individual’s taxonomic schemes and then look at the consequences of their presence or absence in actual buildings. Design practice 1 Transactions with Clients During the process of design, all parties are involved in a process of action differentiation and self-definition. 2 Beyond the Static Brief The basis for the design of a place is an analysis of the role/activity patterns which have given rise to the requirement for a place to be built and the inherent transactions out of which changes will emerge. 3 Participation Is Education Given that people come to understand the relevance of places for their actions by making decisions about those places, the process of participation in design is always one of self-education. Its planning and timing should, therefore, facilitate and accommodate that educational process. 4 Implicit Design Through Selection Many design decisions are made through selection rather than invention. It may be through the selection of an architect to do the design or selection of a place in which to live. Design practice should, therefore, enhance the diferentiation and integration on which efective selection can be founded. 5 Role/Rule Negotiation The design process should itself encourage the negotiation of place use, especially between diferent role groups.

Action and place

157

6 Symbol Sharing The expressive qualities appropriate to any place need to be explored with interested parties as much as the functional ‘brief ’ does. 7 Representations of Action In order to act on place proposals, the possible actions as well as the potential physical form should be represented. As much efort should go into representing what the experience of a place will be as in representing what it will look like. Notes 1 This paper emphasizes the dynamic interplay between what people do and the nature of the places where they act, challenging the static form of most human/environment research. In a somewhat-tongue-in-cheek approach to emphasizing that there are other more active ways of thinking about our being-in-the-world, and reflecting that unfolding set of related developments beyond the rigid form of academic arguments, I have structured this essay loosely around J.S. Bach’s most famous suites for solo cello, notably suite number six in D major BWV 1012, composed around 1720. The astute reader will find links to the theme and style of each movement in the content of that section. Although I have taken the liberty of leaving out direct reference to the fourth movement, the ‘Saraband’. 2 In this brief general paper there is not the space to elaborate all the specific consequences of the proposed approach. However, in the belief that an action theory should have direct implications for human acts, I am including some tables which indicate a few of the conclusions to which I see the perspective leading

Bibliography Bach, J.S. (circa 1720) Six Suites for solo ‘Cello Number 6 in D major BWV 1007–1012 Barker, R.G. (1968) Ecological Psychology Redwood City: Stanford University Press. Blythe. R. (1981) (ed) Places: An anthology of Britain Oxford: Oxford University Press. Canter, D. (1977) Psychology of Place New York: St. Martin’s Press. Canter. D. (1977) Children in Hospital: a facet theory approach to person/place synomorphy. Journal of Architectural Research 6 (2). Canter, D. (1979) Y a-t-il des Lois d’ Interaction Environnementale in J-G Simon (ed) Experiences Conflictuelle de l’Espace Louvain-LaNeuve: Universite Catholique, 391–398. Canter, D. (1986) “Putting Situations in their Place: foundations for a bridge between social and environmental psychology in A. Furnham (ed) Social behaviour in Context London: Allyn and Bacon. Canter, D. (1984) “Intention, Meaning and Structure: Social Action in its Physical Context” in M. Ginsburg et al (eds) Discovery Strategies in Psychology Social Von Cranach, M. and Harré, R. (eds) (1982) The Analysis of Action London: Cambridge University Press. Csikszentmihalyi, M. and Rochberg-Halton. E. (1981) The Meaning of Things London: Cambridge University Press. Gergen, K.J. (1971) The Concept of Self New York: Holt. Rinehart and Winston. Grauman, C.F. (1976) The Concept of Appropriation (Aneignung) and Modes of Appropriation of Space in P. Korosec-Serfaty (ed) Appropriation Space published by Universite Louis Pasteur pp. 113–125.

158

David Canter

Groat. L. (1982) Meaning in Post-Modern Architecture Journal of Environmental Psychology, 2(1). 3–22. Hillier, W.R.G. and Leaman, A. (1973) “The Man – Environment Paradigm and its Paradoxes” Architectural Design August.16 Kaplan, S. and Kaplan, R. (1983) Cognition and Environment New York: Praeger Karolyi, 0. (1981) Introducing Music Harmondsworth: Penguin, Kelly, C.A. (1955) The Psychology of Personal Constructs New York: W.W. Norton. Krampen, M. (1979) Meaning in Urban Environment London: Pion. Lym, C.R. (1980) A Psychology of Building Engelwood Clifs: Prentice Hall. Lynch, K. (1972) What Time is this Place? Cambridge: MIT Press. Mao, T-T. (1967) On Contradiction Peking: Foreign Language Press. McDougall, W. (1908) Introduction to Psychology London: Methuen. Mead, G.H. (1934) Mind, Self and Society Chicago: University of Chicago Parkes, DON. and Thrift, N.J. (1980) Times, Spaces and Places New York: Wiley. Peled, A. (1976) The Place as a Metaphoric Body Haifa: The Technion. Proshansky, H.M., Abbe, K e, Fabian, A.K. and Kaminof, R. (1983) Place-Identity: Physical world socialization of the self. Journal of Environmental Psychology 3(1) 57–83. Relph, E.C. (1976) Place and Placelessness London: Pion. Seamon, D. (1979 A Geography of the Lifeworld London: Croome Helm. Singer, P. (1983) Hegel New York: Oxford University Press. Sully, J. (1892) Outline of Psychology London: Longmans, Gren and Co. Valle, R.S. and King, M. (eds) (1978) Existential-Phenomenological Alternatives for Psychology New York: Oxford University Press. Wertsch, J.V. (ed) (1981) The Concept of Activity in Soviet Psychology New York: M.E. Sharpe.

Part 3

Methodology Studying place experience

Although we are always part of places, it is particularly challenging to explore the nature of the experience of places. Unlike other areas of psychology, it is difcult to reduce the experience of being somewhere to the sorts of components that can be examined in a laboratory. It was, therefore, possible to take advantage of a remarkable data set made available to me when I was on an amazing one-year fellowship to Japan. Another unexpected coincidence was that there was a Korean student working with me who read Japanese. I was, therefore, able to analyze the location of furniture across a range of diferent Japanese apartments. The importance of this analysis for the theory of place is that in the late 1960s, when a team of architecture students carefully prepared maps of the apartments and assiduously noted where furniture was placed in each space, Japanese appartmentss were notionally flexibly arranged. Sliding screens separated spaces from each other, and much of the furniture was small-scale and easily movable. Even the bedding used for sleeping was rolled up in the daytime and put in a cupboard. There was consequently potential for a great variety of room uses. Yet the psychology of place leads to the hypothesis that distinct patterns of place use would emerge. The novel methodology of carrying out multivariate statistical analysis of the cooccurrence of furniture in similar spaces within the apartments strongly supported that hypothesis. Furthermore, the pattern of place use overtly reflected a general model of the nature of the activities that make up a home. Other findings from the unique set of data revealed the interesting way in which the much more cumbersome Western-style furniture evolved in its use in Japanese apartments. It started as rather specific items for very distinct use but over time was absorbed into general patterns of place activities. This provides a general model of how one culture penetrates another, a finding still to be acknowledged by anthropologists. The power of furniture to reveal interesting aspects of the interactions between actions and places lead me to appreciate how much of human meanings and intentions are reflected in records of what people have done. Perhaps it is not a surprising idea for archaeologists but novel for someone schooled in the experimental laboratory tradition of cognitive psychology. This realization did encourage me to follow through on considerations of how place experience was best explored. One established methodology confronted by the idea of purposive evaluations and environmental roles and rules is the presentation of pictures to participants as the DOI: 10.4324/9781003313052-14

160

Methodology

basis for studying reactions to places. It is just not possible to reproduce the purposes, roles and rules that people have in a place by showing them pictures. However, although there have been an increasing number of studies of people in situ, the ease of showing photographs is still a dominant methodology. Results of a comparison of places people know and experience with those shown in photographs supported the hypothesis that these generate very diferent conceptualizations. These results, though, do not negate the value of studying conceptions of places but do put emphasis on ensuring the places being studied, even if represented by photographs and/or descriptive labels, are known to research participants. A further step towards honouring their experiences is to allow them to formulate the descriptors that they regard as relevant. Encouraging participants to assign places to categories of their own choosing is one way of doing this. The problem then arises of how to analyze those responses to reveal the underlying conceptual structures. The use of a rather unusual form of multivariate statistics solved this problem. Multidimensional scalogram analysis (MSA) represents places as points in space such that the more frequently places are assigned to the same category, the closer together they will be in the statistical space. This approach to respondents’ sortings, the multiple sorting task (MST) was derived from the more complex Kellyian ‘repertory grid’. It still kept the same ambition, though, of revealing people’s underlying conceptual systems. The MST with results generated by MSA has been widely used since it was first elaborated.

12 A non-reactive study of room usage in modern Japanese apartments David Canter and Kyung Hoi Lee

Introduction The selection and arrangement of domestic furniture are decisions which both enable people to cope more readily with their physical environment and do themselves contribute to, and to a marked degree modify, that environment. These decisions also, by their very nature, provide non-reactive psychological data. They also may be regarded as an indication of what people want, or expect, to happen in particular spaces. As such, they are presumably reflections of non-verbal, activity-oriented constructs of those spaces (‘this is a dining space’ or ‘a relaxing space’). They also (as Sommer’s 1969 work so clearly shows) provide constraints which contribute to behaviour-milieu synomorphy. That furniture has a significant efect upon the ‘meaning’ of rooms, especially how ‘friendly’ they are, which has been demonstrated a number of times (Canter and Wools, 1970; Wools and Canter, 1970). In the great majority of these studies, furniture arrangement has been shown to account for more of the variance than such architectural variables as wall colour, room shape, window size or lighting. Actual observations of furniture in houses have also been related to social psychological factors, such as the socio-economic status of the owners and their social history (Chapman, 1955; Laumann and House, 1970). Diferences between furniture arrangements in ofces have been shown to be related to diferences in roles of their occupants (Joiner, 1971). There is thus evidence to suggest that a study of domestic furniture arrangement may reveal patterns relevant to the understanding of room usage as conceived and classified by people and the efects of the physical environment on the structure of behaviour settings. Furniture in Japan In order to develop these ideas further, it was necessary to find a specific context in which the collection of suitable data was feasible. The study of Japanese domestic environments proved ideal in several respects. First and most obviously, the verbal constructs which are conventionally explored are not readily available to a Western investigator. The difculty the investigator always has in being sure of the meaning of his subjects’ words is greatly DOI: 10.4324/9781003313052-15

162

David Canter and Kyung Hoi Lee

increased in Japan. Even if he speaks Japanese, the roots of Western languages and Japanese are so totally diferent that a truly veridical translation is often impossible. For instance, a distinction between ‘parlour’ and ‘lounge’ cannot be found in Japanese, and the distinction the Japanese make between a guest room and a room for visiting relatives cannot be adequately expressed in English. This reflects a tradition as distinct from that of the West in its use of space as any of the more apparent diferences of, say, language or religion. Any Western explanation of the Japanese conceptualization of space gains much from the employment of a non-verbal technique. Furthermore, the sensitivity of the Japanese to the still relatively rare study of their way of life by Westerners requires that a non-reactive, preferably unobtrusive technique should be employed. The traditional approach to domestic interiors in Japan and their large diferences from any of those in the West, which is the second reason for the relevance of our formulations to Japan, is well illustrated by reference to Nishihara (1967): The West operates on the idea that each function has its own space. . . . The Japanese house, however, names its rooms by their location . . . without direct reference to function. . . . Japanese spaces, instead of having a fixed function, suit the function to the occasion and need. In other words, the use to which a space is put varies with the time of day. The size of a given space freely changes as we open or close the fusuma (sliding screen). The notion of continuous and uninterrupted spatial flow pervades even the storage spaces, which are treated much as the actual living spaces.  .  .  . Since these interiors conform to a number of needs, they actually serve no function at all. When a bed is needed, the Japanese bring it in, and when a table is required, they bring that in too. In other words, the Japanese house is functionally flexible. Because of this traditional, undiferentiated use of spaces, the question arises whether this is related to the number or average size of spaces available. It is possible that given a number of rooms, they would not be used in any consistent way from household to household. Over the past 100 years, but more especially since Second World War, this pattern of living has been changing, one of the most notable ways being the introduction of Western-style furniture. In recent years, this may have accelerated. In 1955, the ratio of space occupied by furniture to total floor space was 20% (Suzuki, 1971). Our data, collected 14 years later, shows a ratio of 30%. This increase is reflected in the advent of such large Western items as couches and dining tables, televisions and refrigerators. Thus, a third reason for looking at present-day Japanese furniture selection and location is to see the role of Western furniture within it. (Interestingly, there are many analogies in the West, such as the introduction and use of televisions – cf. Canter et al. this volume – movable cocktail cabinets or fireside chairs.) It is thus of interest to study room usage in Japan in an attempt to find underlying relationships to spatial forms. These relationships will, on the one hand, enable us to understand how behavioural settings relate to the traditions and forms available. On the other hand, because furniture reflects active decision making, its study should

A non-reactive study of room usage

163

also reveal something of the conceptual structures involved. In both cases, we are using non-reactive data to clarify the links between environment and behaviour. Data 137 detailed plans of apartments in multistorey company housing estates in Tokyo, on which all furniture had been marked, were kindly made available by Prof. Suzuki (1971). They had been prepared by architecture students at Tokyo University. The apartments were made up of five standard layouts: 1DK, 2DK, 2D, 3DK and 3D. The number indicates the number of rooms beyond the DK (dining kitchen room) and D (dining room and separate kitchen). Figure 12.1 shows an example of the plans made available, and Figure 12.2 shows the five layouts. Table 12.1 gives details of the sample. Table 12.1 and Figure 12.2 show that an increase in the number and size of rooms relates to an increase in both the number of occupants and the space per person. Table 12.1 Summary of layout data Layout

Sample size

Average number of occupants

Average space per person

Number of rooms

1DK 2DK 2D 3DK 3D

28 25 28 27 29

2.9 3.4 3.1 3.7 4.0

23.7 26.7 33.0 35.5 37.0

2 3 3 4 4

Figure 12.1 An example plan.

164 David Canter and Kyung Hoi Lee Thus, the sequence, 1DK, 2DK, 2D, 3DK, 3D, is one of a steady increase in the spatial qualities of the houses. Because of the allocation system for these houses and the usual progress in Japanese companies, this sequence is also related to stage in family cycle and seniority of the household head, within the company. The sequence may, as a consequence, be regarded as one of ‘development’, in career, family growth and environmental quality. This ‘development’ sequence and its environmental concomitants, although not unknown in the West, is typical of the structured nature of Japanese society (cf. Halloran, 1970). Analysis The furniture indicated on the plans in each of the 16 rooms was first submitted to a content analysis. This gave rise to 43 categories of furniture. A basic matrix was then prepared showing the frequencies of each of the 43 furniture types in each of the 16 rooms. The problem with such a data matrix is that it requires some form of reduction to facilitate interpretation. However, no specific hypotheses existed to suggest the appropriate inferential statistics. Furthermore, conventional descriptive techniques, such as factor and principal components analysis, force a pattern onto the data. They make assumptions about its underlying dimensionality which we had no reason to accept. It was, therefore, decided to use a technique developed specifically to maintain the richness of the original data, making as few assumptions as possible. For this analysis, the matrix was enlarged by including the correlations between furniture-type frequencies across rooms and the correlations between furniture-inrooms frequencies across furniture types. The total matrix was then submitted to smallest space analysis (SSAP i: Lingoes, 1973). This places close together those types of furniture which occurred frequently in the same room, and those rooms which frequently contained similar pieces of furniture are also put close together. The degree of closeness represents the degree of similarity in the original frequency patterns. Mathematically, a space of many dimensions is feasible, but in fact, a two-dimensional space was found to be statistically acceptable and psychologically meaningful. The spatial configuration for furniture types is shown in Figure 12.3 and for rooms in Figure 12.4. These two configurations are synomorphic; thus, the presence of a furniture type in one location in Figure 12.3 and a room in the same location in Figure 12.4 indicates that the furniture has a high probability of occurring in that particular room. It must be emphasized that these figures are representations (or descriptions) of the observed frequencies and that inferences (or hypotheses) in relation to the original concepts must now be tested against these configurations. Results It follows from the mathematics of the spatial organization that the nearer a point is to a group of other points, the less distinct it is from those points. Thus, a point near to the centre, such as room number t, is generally similar to many other rooms. A

A non-reactive study of room usage

(A)

(B)

(D) Figure 12.2 The five layouts.

(C)

(E)

165

166 David Canter and Kyung Hoi Lee point near the edge is generally diferent from most other rooms. This interpretation of the configuration implies that the farther away points are from the centre (indicated by a black dot), the less varied is their function. The meaning of this degree of specificity may be seen most clearly with reference to the furniture configuration (Figure 12.3). One of the most central items is a Japanese cushion. This may be part of almost any setting (found in any room), whereas a refrigerator, which is very far from the centre, has a high degree of functional differentiation and thus is found in relatively few situations. Validation of this interpretation of the configuration was possible because Professor Suzuki had carried out interviews with the house occupiers and had collected details of the activities which took place in each room. From the published summaries of these activities (Suzuki, 1971), it was possible to give each room a score for the amount of activity which took place within it. The frequency of occurrence of four activities (household, reading, family entertaining, receiving a guest) were averaged for each room to give a general measure of ‘activity intensity’. Thus, the higher this score, the more multilpurpose is the room.

Figure 12.3 Configuration for furniture.

A non-reactive study of room usage

167

The activity intensity score was correlated with a measure of centrality (Lingoes, 1973) based on the standardized coordinates of the points in the two-dimensional space. The correlation was 0.69 which is significant above the 0.t% level, indicating that as centrality increases, so does the general activity level of the room. This correlation lends support, then, to the point that there are diferent degrees of functional diferentiation of rooms and furniture as well as the diferent kinds, obvious from the groupings in Figures 12.3 and 12.4. The next question to ask, then, is whether the degree of functional diferentiation relates to the major plan form diferences and their concomitant variables indicated in Table 12.1. The relationship between centrality score and layout is shown in Figure 12.5. From this, it is clear that the more rooms in the house plan, the more functionally diferentiated is each of those rooms. Earlier, it was suggested that the increase in family size, status, income, age and space available, which appear to vary with the number of rooms in the house, may all be conceptualized as aspects of ‘development’. The hypothesis thus emerges that ‘development’ is typically related to (or even produces) functional diferentiation. It is difcult to find any conclusive evidence for this hypothesis from only one set of data. However, we may clarify the subsidiary questions which future studies of the

Figure 12.4 Configuration for rooms.

168 David Canter and Kyung Hoi Lee growth of functional diferentiation in houses need to answer. First, an important empirical question is to identify the form of the growth in diferentiation through house types. For instance, which are the first room distinctions to emerge, and which the last? Examination of Figure 12.3 suggests that Japanese public spaces are first diferentiated from Japanese sleeping places and only thereafter do Westernstyle dining places emerge. A second point is to identify diferences in the rate of change of diferentiation from one form to another. Figure 12.5, for example, shows a flattening of the curve after the 2D house form. If further research lent support to the occurrence of a levelling of in room diferentiation, this could well provide an important criterion for design. A third question underlies these two: What aspect of the interaction between house form and development is it that is changing? The present data give us some clues to this. It did emerge as a possibility from our examination of Figure 12.3 that a later stage of development may relate to the introduction of Western-style furniture.

100

90

Centrality (distance from center)

80

70

60

50

40

30

20

10

0 1 DK

2 DK

2D Plan form

Figure 12.5 Average centrality for each layout.

3 DK

3D

A non-reactive study of room usage

169

Table 12.2 Degree of ‘Western-ness’ of Japanese furniture and centrality score Item no. Description Group 1 Japanese-style furniture

36 34 33 37

Sitting cushion Altar Brazier Mattress

15 34 39 48

9 19

Hanging shelves Low table Stool Wardrobe

53 74 74 75

5 7 23 32

Mirror Bookcase Cofee table Gas fire

79 82 86 88

17 II 27 20

Armchair Television Refrigerator Dining Suite

97 102 I13 113

Group 2 Western forms of Japanese furniture

Group 3 Japanese forms of Western furniture

Group 4 Western-style furniture

Centrality Average centrality

34

66

82

to6

This fits well with one of our original reasons for looking at furniture in Japan, the advent of Western style. This aspect of the furniture thus allows us to explore the ‘development’ hypothesis from a diferent viewpoint. Western-style furniture has been introduced slowly into Japan over the last two or so years; thus, it is possible to classify the degree of ‘Western-ness’ many pieces of furniture represent. Table 12.2 shows four groupings according to this dimension, for items on which the authors were in complete agreement, together with their centrality scores. Space does not permit the presentation of the reason for each item occurring in each group. But they may be summarized. Groups 1 and 4 are clearly diferent. Group 2 was made up of those items which existed in some form prior to Westernization but have been modernized in recent years. Group 3 consists of those items which, although clearly Western in origin, have been significantly modified by the Japanese. The notable thing about the average centrality scores is that they show a steady linear increase from Group 1 through Group 4. It would thus appear again that the influences over time represented by the penetration of Western style into the Japanese home relates to increasing functional specificity. Using the word ‘development’ for this introduces an unfortunate evaluative connotation, for it may be speculated that the increasing use of the Western style of furniture has slowly reduced the flexibility of the Japanese house. Indeed, putting this together with the house layout relationship would suggest that the larger the house, the more Western style it is likely to be. Is the introduction of Western-style furniture producing the need for

170 David Canter and Kyung Hoi Lee larger houses? Does this indicate that one way of reducing problems of crowding in the West is to introduce matting and cushions with sliding screens instead of doors? Another way of looking at these relationships is to speculate that the introduction of Western furniture has provided the opportunity for a form of environmental specification not possible with Japanese furniture alone. Is this furniture thus giving rise to, or being produced by, a restructuring of the pattern of behaviour settings? Some furniture is more likely to do one than the other, but it is highly probable that certain pieces reflect an important restructuring of the ways in which people interact with their environments. Discussion and summary One way of looking at the present study is the use of the smallest space analysis of furniture selection and location in Japanese apartments for exploring and clarifying the links between environmental constructs and behaviour settings. Forging these links is possible because the decisions which people make when choosing and arranging furniture may be considered to be both representative of the constructs they have about their rooms and indications of the behavioural settings for which those rooms cater. In doing this, we have been able to show that one of the crucial issues expressed in both these viewpoints, the need to examine and represent change in the systems involved, may be readily indicated from our analysis. Clearly, more detailed studies examining this are necessary, but there is every indication that the analysis of furniture in situ will reveal the forms and patterns of penetration into one or another of various conceptual or behavioural systems over time. An interesting further concept emerged – that of functional diferentiation. The questions it raises are whether the links between concepts, behaviour and environment are easier to understand if we look at the ways in which they become more diferentiated over time. Interesting theoretical questions emerge from examining the form of the relationships between diferentiation and development. For instance, in the Japanese situation, one may hypothesize a sequence of events in which Western furniture and its associated constructs and behaviour almost literally penetrate the existing Japanese domestic system. Thus, one hypothesis suggested is that systems interlock with one another by the newer system initially being absorbed at a specific level, only eventually being given a variety of uses. This, of course, makes psychological sense if one accepts the links between familiarity with an object and the range of ways it may be conceived. The development from an egocentric view of the world common in much developmental psychology theories (notably Piaget’s) also fits these findings. New entities in the environment are first distorted to fit the existing structure, only later themselves distorting the structure to be accommodated. It is, of course, important to examine parallel data in the West (possibly in a situation in which a distinctly foreign style of furniture is being introduced) if the previous generalizations are to be supported. Relating this data to other forms of psychological response is also necessary to see if people actually behave or conceptualize their rooms in the ways suggested by these formulations. Our central concern

A non-reactive study of room usage

171

here is with the overlap and interaction of a variety of developing patterns, environmental, conceptual and behavioural. We have started this exploration at the scale of rooms, but it is meaningful at the larger scale of buildings or the smaller scale of parts of rooms. Presumably diferent patterns and linking concepts will be present at diferent levels of scale. Indeed, the relationships demonstrated here may well not exist at any other scale. All these developments and the others indicated in the paper require a massive amount of research. It would be deluding ourselves to believe that massive support will be forthcoming for this research. It is, therefore, necessary to develop simple techniques which can be cheaply and efectively used. It would appear that, although time consuming, the recording of furniture in houses has the attributes of economy and simplicity and thus is worthy of further exploration, particularly as it is a non-reactive, fairly unobtrusive measuring technique. One final note about the approach made here may be sounded. By taking the residue of human action on the environment as our central data and exploring aspects of its development and change, we find ourselves behaving much as if we were archeologists dealing with the remains of forgotten civilizations. It is thus no surprise that archeologists have been carrying out environmental psychology research using similar techniques to ours (cf. Hudson et al., i97 t). However, we have one significant advantage over the archeologists: we can talk to the people who use the ‘relics’ and to the people who have designed them and their surroundings. It is an advantage we should not throw away. References Canter, D., and Wools, R. (1970). A technique for the subjective appraisal of buildings. Building Science, 5, pp. 187–198. Chapman, D. (1955). The Home and Social Status. London: Routledge. Halloran, R. (1970). Japan: Images and Realities. Tokyo: Tuttle. Joiner, D. (1971). Social ritual and architectural space. Architectural Research and Teaching, 1 (3), pp. 11–22. Laumann, E. O., and House, J. S. (1970). Living room styles and social attributes. Journal of Social Research, 54, pp. 321–342. Lingoes, J. (1973). The Guttman-Lingoes Nonmetric Programme Series. Ann Arbor: Mathesis Press. Nishihara, K. (1967). Japanese Houses: Patterns for Living. Tokyo: Japan Publications. Sommer, R. (1969). Personal Space: The Behavioral Basis of Design. New York: Prentice Hall. Suzuki, I. (1971). Architectural Planning 6: Planning of Communal Houses. Tokyo: Tokyo University Press. Wools, R., and Canter, D. (1970). The efect of the meaning of rooms on behaviour. Applied Ergonomics, 1 (3), pp. 144–150.

13 Picture or place? A multiple sorting study of landscape M. J. Scott and David Canter

Introduction Landscape preference research examines the types of landscapes that people prefer and explores why they prefer them. Such information can be applied to conservation and landscape design, also towards a greater understanding of what is aesthetically pleasing across a wide range of places. For practical reasons, photographs are usually used as representatives of various landscapes, and it is often assumed that the photographs will be evaluated in the same way as the place depicted in the photograph. It is the aim of this paper to demonstrate theoretically and empirically that there is a diference between an evaluation of the content of a photograph and an evaluation of the experience of the place as if the person was actually there. Photographs as representatives

There has been considerable debate within the area of landscape research as to the suitability of using photographs as a substitute for direct experience of the landscape. Several authors (Zube, 1984; Uzzell, 1989; Carles et al., 1992) have stressed the importance of using senses other than just sight when researching landscape and that the information from other senses must be included, if possible, in the research. If this is the case, then information from all the senses will reduce the impact of one feature, such as the visual sensation of a lake, and combine to form experience. In examining the validity of photographs as representations, Kroh and Gimblett (1992) compared laboratory data with that gathered in the field for preference ratings of places. They discovered diferences in the preferred landscape compositions and the degree of preference stated. They went on to say that the discrepancies between the two sets of data indicate that people’s ability to rate a scene for preference and to reach a higher level of agreement is increased when multisensory stimuli are available (p. 67). Place experience includes information from all the senses, and although evaluation is based primarily on the visual aspect (Balling and Falk, 1982), the information from the other senses is important. Shuttleworth (1980), Zube et al. (1987) and Stewart et al. (1984) endorsed photographs as a valid simulation, and the latter suggested that if people were viewing a scene, then they would gather similar information to that supplied by a DOI: 10.4324/9781003313052-16

Picture or place? A multiple sorting study of landscape

173

photograph since they would be standing still and mainly use their sight to gather the information. However, even though photographs have been found to be a valid representation of a scene, there are still disadvantages in using them as substitutes for a field study. Brown and Daniel (1987) found that time of day and the seasons in which they are taken and viewed should be similar. Therefore, weather and number of people, cars and so on in the photographs should also be taken into account. Law and Zube (1983), however, found that there was no need to control for foreground detail and enframement in the photographs. Hodgson and Thayer (1980) discovered that using labels in the study could influence the preference rating for the scene depending on whether they inferred human influence on the land use. These results, however, assume that evaluations are intended to be made based on the information presented in the photograph rather than drawing on the experience of the place in the photograph. Another issue is the choice of scenes for the photographs. Danford and Willems (1975) point out that researchers select scenes which contain those properties of the environment which they believe will be salient in eliciting the response from the participants. Controlling events in this way again reduces the similarities between real life and the experiment because people experience many things and not necessarily those presumed to be salient. Coeterier (1983) compared on-site experience to photographs and found that none of the participants recognized where the photographs were taken despite the fact that they had visited the sites. This may be because what the participants thought was important at the site was not captured in the photograph. A way to solve the problem of what to take and how to improve recognition of the photographs is suggested by Hull and Revell (1989). They asked members of the participant pool to choose and photograph the scenes themselves. This meant that the photograph pool was generated by the participants and that, therefore, the photographs would be known to the participants. The literature, therefore, suggests that photographs can be considered suitable simulations of real landscapes, although there are reservations about their use and ways of improving the standard of the simulation. In fact, as Canter (1974) points out, if the simulations were not good substitutes, then they would not be recognized as representations of the objects they simulate. However, although photographs can be validly used as representations, their use and any conclusions which are drawn need to be done in the full knowledge that they are just representations and not the real experience. Preference and scenic beauty

Much of the landscape preference literature has focused on identifying features or qualities included in the photograph which seem to correlate to either preference decisions or to high scenic beauty scores (e.g., Herzog, 1985, 1989; Gimblett, 1990; Patsfall et al., 1984; Bernaldez et al., 1987; Ruddell et al., 1989). From this type of work, some consistent results have been found. For example, when people categorize various landscapes, the division between what is rural and what is urban is strong (Ulrich, 1977; Uzzell and Lewand, 1990), with rural places being much

174

M. J. Scott and David Canter

preferred to urban places (Zube, 1973; Hodgson and Thayer, 1980; Zube et al., 1983; Hull and Revell, 1989). The presence or absence of water is also a clear division (Zube, 1973; Zube et al., 1983; Blankson and Green, 1991). Its presence seems to increase scenic value (Zube, 1973; Blankson and Green, 1991). Whilst these factors may indeed be relevant and consistent when people are initially asked to consider photographs of places, they are not necessarily salient in judgements and evaluations about places or in preference decisions. ‘Preferences exist within a context of values, beliefs and experiences’ (Kroh and Gimblett, 1992: 58), and thus, as Uzzell and Lewand (1990) and Grahn (1991) suggest, there is a need to look at the models people hold rather than just regarding preference as a function of physical aspects. Attempts have been made to move landscape preference research towards a more structured and coherent model of preference for places (Hull and Revell, 1989; Schroeder, 1991; Kroh and Gimblett, 1992). A useful guideline for such research is Canter’s (1977) model of place, which suggests that a place is a combination of the physical attributes, the behaviours which occur there and people’s concepts of that place. In much work on landscape preference, expected behaviours have been ignored as an influence on preference decisions. However, Genereux et al. (1983) found that people have particular ideas about the types of behaviours which are expected and suitable in places and thus highlights the importance of including such factors in any preference evaluations. In researching the impact of the senses other than visual input, Kroh and Gimblett (1992) noted that ‘tactile, dynamic factors significantly contribute to preference’ (p. 67), and Carles et al. (1992) found that most of the environmental preference was dominated by the auditory factors. These factors would be experienced in the field but would not be included in research using photographs unless referred to, or inferred, specifically. Visual vs place experience

The main problem for researching landscape evaluation is that humans experience landscapes when they are actually in a place. Paintings and photographs of landscapes require an interpretation of the scene by the perceiver (Hodgson and Thayer, 1980) and particular aspects of the scene are given importance by the artist of the painting or photograph (Davis, 1989). Ittleson et al. (1974) write that ‘all environments carry a set of meanings . . . [which] are recognized by the individual in terms of his [sic] own perceptions and values’. (p. 17). Humans do not just see something and understand it. They use experience to interpret what they see (Hume, 1777; 1986 edition). Thus, various influences on people can play a large role in determining the interpretations people give to landscape and nature. In addition, not all of the representation of the place in a person’s thoughts is gained through direct experience. Other knowledge can be gathered through books or contact with other people. (Appleyard, 1970). Experience is generally understood to mean the knowledge gained from participation in an event through which a person has lived (Flew, 1979; Reber, 1985). It is

Picture or place? A multiple sorting study of landscape

175

the ‘sum total of what one has observed, learnt or undergone’ (Penguin English Dictionary, 1965: 257). Thus, experience is a sum total of knowledge which is gained from what people see, hear, feel, smell and taste. Furthermore, the reasons a person has for being in a place and what they have done or can do there also shape their experience of that place. It includes the understanding of the sum total of experience of the place. Evolutionary theories (e.g., Kaplan, 1987) derived from studies of photographs may, therefore, be misleading because they ignore the meanings and associations which people have for places. Although there are strong theoretical reasons for distinguishing between evaluations of photograph content and the evaluations of the place which is represented by that photograph, the techniques which have frequently been employed to research landscape preference, such as semantic diferentials and bipolar adjective scales, are not necessarily able to show this distinction since the participants are not able to use their own concepts when rating the photographs. By using researcher-imposed concepts, information about how the participants conceptualize and what they consider to be important is lost. If the aim is to understand the evaluations or preference decisions, it is important to explore a person’s own understanding. A method which can explore this and show any diferences between these two types of evaluation is the multiple sorting task. The multiple sorting task

A method which reveals people’s own concepts is the multiple sorting task (Canter et al., 1985; Canter and Monteiro, 1993). The sorting task allows this to happen through providing material to be sorted, for example, photographs, but then allowing the participant to categorize according to their own concepts and constructs. It enables the participants to respond to the provided material, unhindered by prespecified concepts from the researcher (Groat, 1982). In this way, the participants provide their individual conceptualizations, rather than the researcher imposing his or hers on the participants. By providing the participant with information, the sorting task seeks to uncover the conceptual structure through which the participants understand and arrange the information and thus the nature of the interpretation of the data upon which they make their evaluations. This draws heavily on the work of Kelly (1955) and personal construct theory. Kelly maintained that people have theories about why things happen, derive hypotheses and then test the hypotheses to see whether the they are validated or invalidated (Fransella, 1984). He suggests that psychological understanding of ourselves and others comes through studying ‘the personal constructs we have each evolved in order to help us predict events in our personal world’. (Fransella, 1984:  132). Clearly, from this position, preferences, evaluations and decisions stem from the constructs and models which we hold about the world. By looking at the models which people hold and their goals or purposes, a greater understanding of people’s evaluations and preference decisions can be made. Taking this further, Canter (1984) suggests that it is the meaning which a place holds for a person which may in some way be entwined with the pleasure of

176 M. J. Scott and David Canter FACETS

ELEMENTS

Physical

Action s

Cognitive

Urban

Walking

Like

Rural

Playing

Dislike

Local villages

Viewing

Not unanimous

Seaside

Tourist

Specific things

Sport No action

Figure 13.1 The three facets and the elements within them for photograph selection process.

a place. To address preference, therefore, people need to be able to express their conceptualizations about a place. As a result, the multiple sorting task is useful for uncovering the models which people hold when they make their evaluations. The study, therefore, used the multiple sorting task to investigate whether diferent evaluations are made if the participants are asked to evaluate the photograph, compared to evaluations of the place in the photograph. Method Participants and items

A secondary school in North Yorkshire, England, was chosen as the source of participants for the study because its catchment area covers all the school-age pupils (11–18) from the local towns and villages, and the socio-economic status of the students at the school, therefore, reflects that of the immediate local area. Many of the students are resident in the area for the whole of their academic career. This means that students at the school are likely to be familiar with the surrounding area. The participants, students in the lower sixth form (aged 16–17), were chosen as a homogeneous group because they were of similar age and educational standard. Most will have turned 17 years old during the educational year and so may be learning to drive and gaining independence. This increases opportunities to explore the locality for themselves, something which may be denied to younger students. Another important factor was that these students were not involved in external examinations during the time of the data collection and so were available in school during June and July when the data was collected. The subjects of the photographs used in the research were proposed by a sample of the participant group in a group discussion. They were asked to suggest landscapes and places which they liked and disliked within the local area. The local area was regarded by the participants as a stretch of North Yorkshire reaching from Whitby across the North York Moors to Teesside. The school is situated at the foot of the Cleveland Hills, 10 miles from Middlesbrough and 20 miles from Whitby. Thirty-seven photographs were taken of all the places identified in the group discussion, over four days in May 1993, under similar weather and time-of-day

Picture or place? A multiple sorting study of landscape

177

conditions. The selection process for the printed photographs used the three aspects or facets of place as defined by Canter (1977). A unique number was assigned to each photograph and then each one was classified in terms of one of the elements within each facet. The three facets were physical attributes, actions which could be carried out in the places captured in the photograph and cognitions of the places. These photographs are shown in the ‘Appendix’. For the physical attribute facet, the elements were urban, rural, local villages, seaside and specific things or areas. The elements for the actions facet were walking, playing, viewing, ‘tourist’ activity, sport and no action. The third facet of cognitions arose from the original information which the students gave the researcher about whether they liked or disliked the landscapes and places. (See Figure 13.1.) By categorizing each photograph in terms of the previous facets, items which always appeared together could be assumed to be showing similar landscapes or places and, therefore, not a great deal of variety. By discarding similar surplus photographs, the item pool was reduced to 20 photographs which best reflected the wide variety of locations. The process of reducing the photograph pool was intended to be exhaustive but with no redundancy, as this would bias the weighting in the sorting. Labels were attached to the front of the photographs to indicate the focus and location of the photograph, and they were re-numbered 1–20 on the back. The research was carried out during May, June and July 1993 at Stokesley School. The tasks were completed by 41 students working alone. There were 23 males and 18 females. Each participant used a set of the 20 colour photographs, and response sheets were provided for each stage of the process. Instructions were given verbally plus some written instructions for the sorting tasks provided on the response sheets. The multiple sorting task Photograph sort

In this sort, the participants were presented with data in a similar way to many landscape preference studies (i.e., they were simply presented with the photographs and asked to respond). In this case, the participants were asked to look at the photographs and then to sort them into groups in such a way that the photographs in any group were similar to each other in some way and diferent from those in other groups. Each photograph could be in one group and one group only. There could be as many groups as they liked and as many photographs in each group as they liked. It was emphasized that there were no right or wrong answers and that it was their opinions which counted. The common aspects identified by each participant for the photographs in each group were then written down by the participant, together with the numbers of the photographs which were in each group. Place experience sort

When the participants had completed sorting the photographs, they were asked to spend a few moments thinking about the places in the photographs. They were asked

178 M. J. Scott and David Canter to imagine being there and to think about how they would feel and what they would expect to see, hear, smell, taste and touch and to focus on the activities there and what their moods, feelings and emotions would be like in this place. They were then asked to repeat the sorting procedure, this time focusing on the places in the photographs according to their feelings about the places, rather than in terms of the photographs themselves. This moved the focus of the research to the places in the photographs. The order of the sorts could not be alternated since the first sort was intended to allow participants to respond based on the presented information without any direction. The second sort, however, deliberately focused on the place represented in the photograph. Analysis Multidimensional scalogram analysis (MSA)

The data produced by the sorting tasks was analyzed using multidimensional scalogram analysis (MSA, Wilson, 1995). In this type of analysis, the groupings imposed by the participant on the photographs can be represented by a row of data or ‘profile’, which can be compared with the profiles for all the other photographs. Comparisons of the profiles show how often photographs have been grouped together. The photographs can then be plotted as points in space. The more frequently photographs have been grouped together, the more similar their profiles and, therefore, the closer points are to each other. Points which are distant from each other indicate that the photographs have generally been considered to be conceptually dissimilar in some way. The plots are divided into areas by reference to the original data supplied by the participants and lines drawn to best represent the categorizations of the participants. A line dividing two adjacent points indicates that although the points are physically close together, the overall structure of the participants’ categorizations shows that they belong in separate sections of the plot. It indicates a construct on which these items are similar but which is not as strong as the constructs which have created the divisions. The descriptions of the areas in the plots come from the names given to the groups by the respondents in the sorting. These are the aspects of the photographs which those in a group have in common. Content analysis was carried out on these descriptions for each sort. The data was divided into four groups to check for reliability across the sample. The sample was divided according to the gender of the participant and the science/ art bias of their chosen A level subjects. (A levels are exams which are accepted as entrance qualifications to universities.) The data is presented to show four MSA plots for each sort. By dividing the data in this way, consistent themes of evaluation can be demonstrated. Random division of participants could have been applied; however, this division also tested whether there were any gender or educational diferences. Content analysis of the two sorts

Table 13.1 shows a content analysis of the descriptions used in the two sorts. In the first sort, 98 diferent descriptors were used, compared to 194 in sort 2. In sort 1, 12

Picture or place? A multiple sorting study of landscape

179

Table 13.1 Table of key headings from the sort and frequency  

1st Sort

2nd Sort

No. of diferent descriptions given Descriptors used five or more times

98 Agriculture Buildings Farming Hill landscapes/views Industry Man-made Moors Open spaces Seaside Tourist Town Water

194 Boring Busy Cars Children playing Familiar Free Fresh Happy Lots of people Noisy Peaceful Pollution Quiet Relaxing Summer Tourism Windy

descriptors were used five or more times, compared to 17 descriptors which were used five or more times in sort 2. In addition, a content analysis of the type of descriptors used in the two sorts was carried out, with each descriptor being coded by two researchers according to combinations of the aspects of place-physical features, behaviours or conceptualizations or some combination of these (see Canter, 1977). A chi square test was carried out to investigate whether there was an association between the type of the descriptors used and the focus of the two types of sorts. Table 13.2 showed that there was a highly significant diference between the photographs and the place sort (χ2 = 57.3 df. = 6 p