University of Southern California Dissertations and Theses

A capability-based approach to defining performance characteristics of the built environment

(USC Thesis Other)

A capability-based approach to defining performance characteristics of the built environment

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

A CAPABILITY-BASED APPROACH TO DEFINING

PERFORMANCE CHARACTERISTICS OF THE BUILT ENVIRONMENT

by

Ferdinand Lewis

A Dissertation Presented to the
FACULTY OF THE SCHOOL OF POLICY, PLANNING, AND
DEVELOPMENT
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(POLICY, PLANNING AND DEVELOPMENT)

August 2008

Copyright 2008 Ferdinand Lewis
i

Epigraph
I tried to listen to
the public voice of our time
tried to survey our public space
as best I could
––tried to remember and stay
faithful to details, note
precisely how the air moved
and where the clock’s hands stood
and who was in charge of definitions
and who stood by receiving them
when the name of compassion
was changed to the name of guilt
when to feel with a human stranger
was declared obsolete.
~ Adrienne Rich, from AND NOW (1995)

ii

Dedication
For Caleb, who gave me the most important three letters.

iii

Table of Contents

Epigraph.......................................................................................................... i

Dedication...................................................................................................... ii

List of Figures............................................................................................... vi

Abstract....................................................................................................... viii

Preface.............................................................................................................x
Capability, Affordance, and Built Environment Audits ......................... xiii
Three Exercises...................................................................................... xvii

Chapter 1: Towards a General Model of Built Environment Audits..............1
Overview.....................................................................................................1
Eight Dimensions of Built Environment Audits.......................................13
Audits are Evaluative............................................................................13
Audits Are Normative...........................................................................17
Audits Describe the Good as Equally Distributed ................................19
A Hierarchy of Descriptive Categories Separates the Good and
the Environment, in Audits............................................................22
Audits Contain Ethical Assumptions....................................................27
Audits Value Individuals in Absolute or Relative Terms.....................31
Audits Conceive of the Good Either as a Means or an End..................33
There are Three Types of Audit Information........................................38
A General Model of Built Environment Audits........................................42
Conclusion ................................................................................................46

Chapter 2: Built Environment Audits and the Ecological Model.................47
Introduction...............................................................................................47
Ecology, Correspondence, and the General Audit Model ........................52
History and Purpose of the Ecological Model ..........................................53
What is the Ecological Model Supposed to Replace? ..............................56
The Correspondence Model..................................................................56
Three Normative-Evaluative Approaches to Auditing
the Built Environment....................................................................64
Satisfactions: The Visual-Cognitive Approach ....................................64
Opportunity: The Rights-Based Approach............................................82
What is the Ecological Model Supposed to Do for AL Researchers? ......98
An Ecological Model Would Recognize Capability...............................112
Capability Best Revealed by Affordance................................................119
iv

Affordance Expresses a Proportional Relationship ................................124
Well-being as Purposive Activity...........................................................127
Capability as a Ratio...............................................................................128
Synomorphy Tradition........................................................................132
What Might an Affordance Audit Include? ........................................141
Conclusion ..................................................................................................144

Chapter 3: Evaluating the Audit’s Informational Basis:
A multiple case study...................................................................145
Analyzing Audits Using the General Audit Model.............................145
How Audits Gather Information.............................................................148
Case: Irvine-Minnesota Audit Instrument (IMI).....................................150
Description..........................................................................................150
IMI, Satisfaction, Opportunity, and Affordance.................................152
“Pleasurability” and Affordance.........................................................155
Case: The St Louis Audit Instrument......................................................156
Description..........................................................................................156
Satisfactions, Opportunity, Affordance, and the St. Louis Audit.......158
Information and Affordance ...............................................................161
Case: Systematic Audit of Green-Space Environments (SAGE)/
Green Vision Plan.......................................................................163
Description..........................................................................................163
Comparison: SAGE/GVP and the IMI ...............................................170
SAGE, Preference, and the Rawlsian Priority of Liberty .......................172
SAGE and the Ecological Model............................................................173
Case: Systematic Pedestrian and Cycling Environment Scan ................179
Description..........................................................................................179
SPACES, Satisfaction, Opportunity, Affordance ...............................183
SPACES and the Ecological Model....................................................185
Case: SOPLAY and SOPARC................................................................187
Description..........................................................................................187
SOPLAY/SOPARC, Satisfaction, Opportunity, Affordance..............189
SOPARC/SOPLAY and the IMI ........................................................191
SOPLAY/SOPARC and the Ecological Model ......................................192
Case: Measuring Urban Design Qualities...............................................195
Description..........................................................................................195
MUDQ, Satisfactions, and the Ecological Model...............................197
“Spot the Affordance”.............................................................................202
Is the Affordance Metric Feasible?.....................................................206

v

Disparities and Fairness ..........................................................................208
Conclusions.........................................................................................210

Bibliography ...............................................................................................218

vi

List of Figures
Figure 1. A Hierarchy of Descriptive Categories Between Determinants
and the Good 24

Figure 2. Individual’s Relationship to the Collective Good Along Three
‘Lines of Valuation’ 36

Figure 3. A General Model of Built Environment Audits 42

Figure 4. Example of “Correspondence” Model of Environmental
Perception 55

Figure 5. The HUD/CNU “Principles of Inner City Neighborhood
Design 87

Figure 6. Active Living Ecological Model 148

Figure 7. Barker’s “Eco-Behavioral Circuit” Model 157

Figure 8. Multiple Case Study after Yin 191

Figure 9. Green Vision Plan Study Area, Southern California Region 159

Figure 10. Average Index Values for Green Vision Plan Parks, Across
All the Categories of the Audit 160

Figure 11. Green Vision Plan Park Acres by Demographic 168

Figure 12. Systematic Pedestrian and Cycling Environment Scan 171

Figure 13. SPACES Audit, Path Material Assessment 174

Figure 14. SOPLAY Momentary Time Sampling Data Form 180

Figure 15. SOPLAY Mapping Variables Observation Sheet 193

Figure 16. Measuring Urban Design Qualities Conceptual Framework 200

vii

Figure 17. Tai Chi Class in the Picnic Area 203

Figure 18. Grassy Hills Afford “Running Down and Screaming” 205
viii

Abstract
Primary research, policy recommendations, and audit instruments are three
responses to the U.S. obesity epidemic offered by built environment-physical
activity researchers. This study is concerned with the theoretical foundations of
audits.
Primary research and audits play different roles: The former is
investigative, the latter is evaluative. Those roles are too often conflated, I argue,
altering the type and scope of information available to policymakers, and
obscuring an audit’s fundamental relationship to distributive justice. This study
proposes eight defining characteristics of built environment audits to illustrate the
distinctions between audits and primary research. Because audits are always
normative, and always distributive, they necessarily contain definitions of
distributive justice. Primary research, on the other hand, is not necessarily bound
by normativity or the definition of distributive justice.
Three types of audit information are identified––“satisfactions,”
“opportunities,” and “capability”––each corresponding to one of the three moral-
philosophical definitions by J.S. Mill, John Rawls, and Amartya Sen. A “general
model for built environment audits” is proposed here, based on the three types of
audits proceeding from each of the three definitions. The general audit model is
subsequently used to examine the “ecological models” currently guiding built
environment-physical activity research.
ix

I argue that ecological models are particularly suited to observing what
Sen calls capability, especially when they employ the units of observation called
“behavior settings.” However, the value of the behavior setting unit is
compromised, I assert, by the narrow definition it is often given. I propose an
alternative unit of observation, J.J. Gibson’s affordance, and undertake to
demonstrate how affordance equates with built environment capability.
In the final section, I use a case study protocol based on the general audit
model to examine six audit instruments. I attempt to determine the degree to
which these instruments gather affordance data within the general audit model. I
conclude that few of the audits could be called “affordance audits” outright,
however, some of their methodologies could be extended to gather affordance
data.
x

Preface
This is a theoretical study about built environment audit “instruments” or
“tools,” which are used to evaluate the relationship between the built environment
and human health. I did not plan to make a theoretical study, yet the need for
theory became clear once I got a sense of the implications of audits having no
theoretical model of their own. Audit developers have been relying upon either no
model at all, or else on whatever model happens to be guiding primary research,
which brings audit reliability in question. Chapter 1 takes up this issue,
concluding that primary research and audits are two quite different things, and
that although they are integrally interrelated, there are epistemological and ethical
distinctions that should be given serious attention by future researchers.
Primary research, audits, and urban design intervention are a sequence of
steps, and although those steps inform each other, they are sequential, and
distinct. Primary research informs audits, and audits subsequently inform built
urban design policy. Primary research and audit instruments are certainly focused
on the same world, but from different, though overlapping, perspectives. As I will
argue, the scope and usefulness of the evaluative information informing built
environment policy depends to some degree upon drawing clear distinctions
between primary research and audits.
The social justice implications of built environment audits has largely
gone unconsidered in urban planning research, and although I introduce that
xi

theme here, I do not develop it, due to time and space constraints. The synthesis
that I offer is meant to inform urban design research and policy-making regarding
the practical matter of how built environment audit tools are developed, and while
justice ought to be part of that process in general, my overarching work here has
been to determine whether audits are a subject worthy of serious scrutiny. I hope I
have laid a foundation for future research into audits, for myself and others.
The study references major ideas from three separate fields––moral
philosophy, environmental psychology, and urban design––but my narrow focus
on audits does not allow me to describe the fullness of those literatures, nor give
full credit where credit is due. For example, I will give considerable attention to
Amartya Sen’s innovative “capability approach” as a theoretical framework for
evaluating distributive justice, but I will make no reference to the practical
methodological challenges that Sen’s ideas have made to economists and others
who work on the causes, and relief, of dire poverty. The scope of Sen’s thinking is
truly staggering, and while it is possible to study development economics, aid
policy, and moral philosophy without agreeing with Sen, his contributions to
those fields cannot be ignored.
I draw on the “justice-as-fairness” concept in John Rawls’s Theory of
Justice, which I consider the first successful paradigmatic challenge to the
utilitarian conception of social good. A variety of philosophers and economists
have taken up the utilitarian metric as a methodological challenge, including J.S.
xii

Mill, whose influence can be felt in economics, philosophy, and in the psychology
of perception. Between Mill and Rawls there have been few who have thought so
completely about the public interest criterion, and none who have laid out
complete frameworks for re-thinking the very foundations of the idea of the
public interest. Rawls’s Theory of Justice argues for nothing less than a
reconceptualization of the social contract upon which civil society is based, and
this research leans heavily upon Rawls’s ideas.
At the publication of his theory of Good City Form in 1981, Kevin Lynch
was a prominent and learned voice in urban design theory, whose unflagging
passion for the human-centered view of urban development kept his work
evolving. Good City Form not only built on Lynch’s previous groundbreaking
accomplishments from the 1960s, but also superceded them. I maintain that up
until the end Lynch kept refining his ideas about how urban development can
contribute to human development. Good City Form represented another
remarkable leap in Lynch’s endless exploration beyond the boundaries of his own
accomplishments.
Roger Barker’s 25-year long behavior setting study produced an entirely
new way to observe, analyze, and predict human behavior in the built
environment. It described dynamics of the built environment-physical activity
relationship that had never before been articulated. The significance of Barker’s
xiii

work is evident in the re-emergence of his behavior setting concept in today’s
Active Living research, as I will discuss in the study.
J.J. Gibson was a colleague and contemporary of Barker’s, whose concept
of “affordance” offered not only an innovative framework for the empirical study
of human environments, but on a philosophical level it also represents an
alternative to the subject/object conceptual rift that has been fundamental to
science since Descartes. Gibson’s work takes center stage in this study, although I
draw on only a fraction of what he accomplished.
My findings are meant to be applicable to the practical work of policy
formulation for urban design. However, my argument is premised on the
theoretical idea that there is no valid model for built environment audits, and this
means that the core of the research is somewhat abstract. My teachers and
colleagues have suggested that I ground the study in concrete examples, and I
have attempted to do this whenever possible. In that spirit, I devote the remainder
of this preface to describing the story of this research project.
Capability, Affordance, and Built Environment Audits
The argument for pairing Amartya Sen’s capability and J.J. Gibson’s
affordance is unique to this study, although the idea for it began in Dr. Tridib
Banerjee’s first-year doctoral seminar at the University of Southern California, in
2004. Dr. Banerjee made the observation that certain theoretical arguments in the
urban design field parallel certain arguments amongst moral philosophers,
xiv

particularly regarding the definition of distributive justice and the evaluation of
collective well-being. Dr. Banerjee went on to propose that urban design theorist
Kevin Lynch’s 1981 Good City Form was a built environment corollary to
Amartya Sen’s capability approach. Further, Sen’s critiques of Millsean
utilitarianism and Rawlsian justice-as-fairness might be shown to have parallels in
Lynch. Dr. Banerjee suggested that could such a synthesis be made, it might be
used to normatively evaluate the built environment in terms of what Sen calls
capability.
I proposed investigating that idea further, and Dr. Banerjee encouraged me
to read in three areas: The background philosophical issues of distributive justice
that inform normative evaluations of the public interest; the history of normative
evaluations of the built environment; and environmental perception. In 2006, Dr.
Banerjee and my colleague JungA Uhm drew me into some research they were
doing on the childhood obesity epidemic, where the Robert Wood Johnson
Foundation’s Active Living Initiative (AL)
1
was funding dozens of major projects
that involved urban designers, public health researchers, and scholars from many
crossover fields. New research was needed on built environment “audits”––
decision-making tools for normatively evaluating how environments contribute to
human health. It seemed that Active Living researchers were particularly
struggling with the idea of how perception informs the decision to be physically

1
The “AL” of Active Living is not to be confused with the “A-L” or “Adaptation Level” Theory.
xv

active. In response to this, in the late 1990’s Active Living theorists had begun
proposing new “ecological” models for research. These often included references
to Barker (though not to Gibson).
By 2006, AL theorists had unveiled a new and very comprehensive
ecological model, to guide primary research and the development of audits.
Active Living ecological models seemed quite clear in their conception of
multi-level “ecological” determinants of physical activity. These “trans-
disciplinary” determinants called for new kinds of interdisciplinary research
between public health and a variety of other fields. Yet, I noticed, the AL
ecological models could be quite un-ecological when operationalized. Much
of the Active Living research on determinants of physical activity attempted
to predict the likelihood of physical activity purely in terms of how the
average person responds to direct environmental stimuli. I noted that AL audit
instruments tended to be lists of things, such as lighting and benches and
crosswalks, or else they were surveys of attitudes, usually through phone
interviews. None of the new AL audits––I review six of these in Chapter
Three––investigated attitudes and things together, and more distressing, few
of them had anything specifically to do with children.
The Active Living research on audits seemed useful for predicting how an
average person might use an average environment, but it did not seem
particularly suited to evaluating how a very particular group of individuals (obese,
xvi

inner city, low-income, minority children) could engage the environment, or why
they would want to. I began to see these two ideas as dueling conceptions, of the
good of an average person, and the good of the individual. In thinking about the
difference between the two, it became necessary to understand how our
perceptions inform our relationship to the good. More, since urban environments
are always public to some degree, the question of perception is not only an
epistemological matter but also a distributive one.
At the suggestion of Dr. Banerjee I had expanded my reading to earlier
studies of environmental perception, many of which are known in urban planning
to this day.
2
This included F. Stuart Chapin’s work on activity patterns; Sorkin
and Berger’s time-budgets; and Goffman’s theory of the public presentation of the
self. Parallel to those investigations, I was also studying systems of qualitative
inquiry with Dr. Polkinghorne.
3
Each meeting of Dr. Polkinghorne’s seminar
began with a review of the intellectual history of qualitative inquiry, starting at
least from Husserl, but sometimes as early as Mach and Brentano. It was here that
I first encountered Kurt Lewin’s innovative work on the “total environment” and
the “lifespace,” and Koffka’s principles of Gestalt psychology.
4
In that reading I

2
The study of environmental perception is by no means new to urban planning and design. Chapin
and the others mentioned above are known in the planning and design fields, but Gibson has
remained obscured to urban planning theorists.

3
Author of Methodology for the Human Sciences which, among other things, traces arguments for
qualitative knowing across the modern history of the human sciences, from Mill to Ricouer.

4
Barker also employs the concept of lifespace. See Barker 1968 p. 7. Like Gibson, Barker
xvii

came upon J.J. Gibson, who was strongly influenced by Lewin’s idea of the
lifespace, and he used it in a study in 1938.
5

Three Exercises
I undertook three exploratory “exercises,” described below. The exercises
were not meant to offer empirical conclusions about parks or anything else, but
rather they were an exploration, for my own edification, of how audits operate in
real contexts. The principal value of the exercises to me was that they underscored
the need for a theoretical model for audit development, and it was at that point
that I decided that I needed to undertake a theoretical study of audits, before I
could advance a new empirical one. The exercises also gave me the opportunity to
experience conducting Active Living audits first hand, and also the opportunity to
experience, in context, what AL determinants look like in the actual landscape. It
is one thing to read reports on the role of proximity-to-transportation in a person’s
decision to walk, for example, but it is quite another to actually walk those
distances and use that transportation.
The exercises were also a follow-up to a hunch. Audits are based on
primary research, and whatever credence is given to an audit by researchers is
usually based on the primary research, and not on the audit itself. Yet in

recognized the value of investigating the lifespace, but he also asserted that the lifespace is more of
a psychological phenomenon than an ecological one. Indeed, Barker says flatly of the lifespace that
“ it is not the ecological environment,” and that knowledge of the environment based solely on
“ psychological variables” misses the whole point of the human-environment ecology model.

5
According to Heft, Barker was also influenced by Lewin.
xviii

familiarizing myself with audits and their development processes, I got the sense
that there was a critical difference between audits and primary research that had
gone overlooked in the literature.
The first exercise I undertook was to conduct AL audits in the field, at two
separate Los Angeles parks. The second exercise gave me the opportunity to
consider how the AL conception of the park as an ecological environment
compared to Barker’s ecological conception of behavior settings. I found that in
some cases audits gathered data that could be called “ecological,” but with the
exception of one case, the data is used to understand the human-environment
relationship in linear, not ecological, terms. I observed that AL data could be put
to use as ecological data, if it were amended with other kinds of data, and given a
new kind of audit framework. In the third exercise I created a new “affordance
audit instrument” that used some items from affordance research, and also some
items from existing AL audits, and also some new items based on qualitative
observations of the parks.
The three exercises were not meant to produce conclusive outcomes, but
were explorations of theoretical matters in a practical context. The only real
conclusion that the exercises produced was a certainty that audit instruments are
fundamentally different from primary research on built environment determinants.
At the point that I reached this conclusion, my study had become a theoretical
one, and my search for a “general audit model” had begun in earnest.
xix

There were three things about audits that needed exploring in theoretical
terms:
1. The fundamental differences between audits and primary research. In
order for urban designers to put audits to the most practical use in
planning interventions, audits need a general model of their own. The
general audit model should be informed by AL primary research, and
should inform what sort of primary research is undertaken, which is not
the case at present. However, the general audit model has some concerns
that are distinct from those of primary research.
2. What it means in theoretical terms for audits to undertake ecological
observation of the built environment-physical activity relationship. How
does audit information differ in terms of distributive justice and
epistemology when it is viewed ecologically rather than in linear fashion?
3. If ecological audit information is critical in the ways that AL researchers
argue it is, then a way must be found to describe the built environment-
physical activity relationship in ecological terms, rather than in the
traditional terms from existing models. As I will argue further on, one
cannot describe circuit-like ecologies in terms of the traditional linear
models of perception. My hope is that the above narrative has set the stage
for the investigations and arguments to follow.

1
Chapter 1: Towards a General Model
of Built Environment Audits
Overview
The context for this study is the childhood obesity epidemic in the U.S.,
which poses a serious health threat to inner city children of low-income families
(AOA, 2007; Ebbeling, Pawlak et al., 2002, p. 480; Kumanyika, Jeffery et al.,
2002; Rocchini, 2002). In response to the acuteness of the problem, urban
planning and design researchers have joined with researchers in the public health
field to assess how qualities of the built environment impact children's physical
activity and “active living” (Active Living Research, 2006; Ainsworth, 2002;
Chawla & Heft, 2002; Ellen, Dillman et al., 2001; Flournoy & Yen, 2004; Miles,
2006; Moudon & Lee, 2003; Moudon, Lee et al., 2007; National Research
Council, 2005). Researchers from diverse fields such as health, urban planning,
transportation, recreation, sociology, and others have effectively created a
subfield that I will call built environment-physical activity research. In applied
terms, the goal of this field is to inform built environment policy about what sorts
of interventions are needed to address the U.S. obesity epidemic. There are two
principle forms of evidence used in building a case for policy interventions into
the built environment.
Physical activity determinants (“determinants”) are the products of
primary research. They are the factors shown to influence the decision and/or
2

ability to be physically active. For example, the lack of “attractiveness” of a
neighborhood may be given as a determinant of residents’ unwillingness to walk.
Major factors such as attractiveness, safety, and others may be aggregates of
multiple minor determinants. The external validity and construct validity of
determinants are generally derived from primary research on the built
environment-physical activity relationship.
Built environment audits are practical instruments or “tools” for evaluating
how an environment contributes to the well-being of its residents. Audits are not
primary research, but are subsequent to it. Audits are inventories, comprised of
sets of determinants that have been previously validated in primary research.
Audit development may also include conventional wisdom, collective intuition,
scholarly hypothesis, and expert judgments. The inventory allows auditors to
evaluate the environment against determinants. Audit findings may be used to
inform future primary research, but their traditional purpose is to inform
interventions into built environment policy and design.
In the context of the U.S. childhood obesity epidemic, primary research
has developed an impressive battery of determinants of physical activity
(Ainsworth, 2002; Ellen, Dillman et al., 2001; Humpel, Owen et al., 2002;
Moudon, Lee et al., 2007; Transportation Research Board, 2005). Determinants
may include not only objective measures such as counts of things in the
environment like crosswalks and street lighting, but may also contain more
3

subjective measures such as the perception of safety or proximity. The question of
children’s subjective perceptions of the environment is of particular concern in
the built environment-physical activity field because it introduces the sorts of
contextual variables with which researchers could explain why at-risk children
might avoid exercise even in optimal conditions, such as well-equipped play
areas, sufficient open space, and close proximities to parks (National Research
Council, 2005; Prezza, Alparone et al., 2005; Ritenbaugh, Kumanyika, et al.
1999; Zabinski, Saelens et al., 2003).
Literature on perceptions of the environment suggests that children bring
mediating competencies and needs to their experience of place (Derr, 2002; Hart,
1979; Korpela, Kytta et al., 2002; Piko & Keresztes, 2006). Children also have
mediating social constraints in the form of parental controls, to which adults are
not subject (Bauman, Silver et al., 2006; Bell, 2002; Borders, Rohrer et al., 2006;
Evans, Lercher et al., 2002). Also, children may simply have a sense of place
based on their experience, that is innately distinct from that of adults (Day,
Boarnet et al. 2006).
Built environment-physical activity research has generated a number of
built environment audit tools (“audits”) with which neighborhoods and parks and
city streets can be evaluated for the presence of goods such as “walkability” that
are shown to determine physical activity. The audits studied here are all within the
purview of (if not actually funded by) the Robert Wood Johnson Foundation’s
4

Active Living research program (AL). AL audit tools are the focus of this study.
The SAGE study, for instance, is not funded by Robert Wood Johnson. However,
one of the purposes of SAGE is to inform interventions in the childhood obesity
epidemic, which is a concern of Active Living research. Therefore, it falls under
the heading of what I am calling built environment-physical activity research.
A number of state-of-knowledge reviews of built environment-physical
activity research have been undertaken by researchers, including Owen, Humpel
et al.; Booth, Pinkston et al.; Ewing & Cervero; and the Transportation Review
Board (TRB), are just four examples. All of those reviews are concerned with
identifying valid determinants of physical activity related to the built
environment, meaning that they are surveys of primary research. Only a few
reviews or evaluations of actual audit instruments exist, notably Moudon and Lee
(2003), Emery, Crump et al. (2001), and Day. The audits being developed under
the aegis of the Robert Wood Johnson Foundation’s Active Living Research
initiative have intensified the interest in audits as tools for built environment
policy research.
This study is an inquiry into the theoretical foundations of audits. The
theories that guide audit development are rarely discussed in the literature, and in
fact it is most often the case that the audit model is simply assumed to be that
which guided primary research (Moudon & Lee 2003, p. 35, for instance; Pikora,
Bull et al., 2002, pp. 188-189). There is little question that valid determinants are
5

key to reliable audits, and primary research is the indispensable step in
establishing those determinants, yet I will argue that primary research and audits
are two distinct types of research, requiring two distinct modes of inquiry, and
two distinct conceptual models. Primary research on determinants is investigative,
while audits are normative-evaluative.
6
The world as modeled by primary
research and that which is modeled by audits are relative and overlapping, but not
synonymous.
The models used in primary research will be described in more detail in
Chapter 2, but here, in Chapter 1, the emphasis will be on problematizing audits
in a way that suggests their criticality, and how their underlying theoretical
structure is distinct from primary research. Of particular concern here are the
varieties of types of data that audits use to evaluate collective well-being, which
make up what Sen calls the “informational basis” of normative evaluations (Sen,
2000, pp. 54-60). The informational basis of audits is shaped by the determinants
employed in them, and also by audit data-gathering methodologies. I argue that an
audit’s informational basis is fundamentally defined by its ethical assumptions
about the meaning of collective good, and using this premise, I explore how those
assumptions might guide the development of a general conceptual model for the
development of built environment audits.

6
Urban design proposals are subsequent to audits, and are normative-instrumental, as will be
described in Chapter 2.
6

Audits use determinants for different ends than those of primary research.
Primary research informs audits about how the environment contributes to well-
being; audits inform policy about whose well-being is affected by the
environment, and what share of well-being they have a right to expect. As I will
argue, primary research has the option to take up the question of distributive
justice, while distributive justice is the formal basis of audits. Primary research is
a process for establishing the determinants of physical activity, while audits are a
decision-making tool.
I will argue there that there are basically three different kinds of audits,
and that there is no single audit framework applicable to all policy research. Each
kind of audit requires a different kind of determinant, and a different conception
of how the built environment contributes to physical health. Chapter 1 closes with
a proposal for a general audit model that describes each of those three approaches
to auditing the built environment.
The general audit model represents three different conceptions of what
determines physical well-being in the built environment, and three different
descriptions of how much of that well-being any individual has a right to hold.
Each conception will produce a different sort of audit instrument to observe
different types of phenomena, and each will employ a different data collection
method, and most importantly, each will define the environment’s contribution to
health in differing terms. One group of researchers will want to find out what
7

qualities of the neighborhood environment make people feel more like exercising.
A second team of researchers sets out to determine what resources neighborhoods
have to facilitate walking, and how the presence of those resources affects the
decision to exercise. A third team of researchers investigates why people who
have opportunities and resources for exercise are nonetheless sedentary.
Each of those projects frames the person’s relationship to the environment
in different terms, regarding who is affected by determinants, how they are
affected, and to what degree. In the primary research phase each produces a
different sort of determinant, and in the audit phase each gathers a different kind
of information. The first team would be interested in what makes any person want
to walk, and focuses on establishing the sorts of things a person might prefer to
find in the environment, such as crosswalks and shade trees and street furniture.
The second project is concerned with sufficient opportunities for the person to be
physically active, in terms of space, facilities, and the ability of the person to
argue for and choose such opportunities for herself. Like the second project, the
third is also concerned that the person has sufficient opportunities, but goes
further to emphasize the person’s ability to make use of opportunities given
certain constraints, such as a disability for example.
Because the first project would model the built environment-physical
activity relationship in terms of how the built environment satisfies the
preferences of most people, it would also observe those things that are assumed to
8

satisfy an average preference. Because the second project would model the built
environment-physical activity relationship in terms of equality of opportunities, it
would also observe the fair distribution of resources. The third project is
concerned with the aspects of the person’s own characteristics and circumstances
that make her more or less able to make use of fairly distributed opportunities.
The goal in Chapter 1 is to produce a general audit model that describes the
human-built environment relationship in terms of three distinct approaches
implied in the hypothetical projects above.
The issue of construct validity is one of the overwhelming concerns of
primary research in the built environment-physical activity field. Yet primary
researchers tend to assume that the validity of audit instruments is merely an
extension of the validity of determinants. Compared to the attention given to the
development of determinants, audit development is approached in an almost
offhand manner. Mostly, the reliability of the audit is all that it tested, to
determine replicability of audit findings. Reliability is most often statistically
established using test-retest and inter-rater reliability (IRR) measures, yet such
tests are meant to establish how replicable the audit might be, rather than how its
determinants represent the variety of built environment-physical activity
relationships contributing to the well-being of individuals. It is as if the
assumption amongst audit developers were that if the environment properly
inspires physical activity, then physical well-being will result.
9

A few researchers have taken up the issue of audit validity. Moudon and
Lee’s 2003 evaluation of extant audit instruments rigorously examined how audits
go about assembling their sets of determinants (Moudon & Lee, 2003). Their
findings included the statement that “most of these measures have not been
empirically validated” by the test-retest measures undertaken by audit developers
(p. 36).
To quote Moudon and Lee (2003):
[F]uture environmental audits might need to be tailored to the
characteristics of the physical environment itself (e.g., urban or suburban),
to the type of users (e.g. young or old), and to the different purposes of
physical activity (e.g., travel, recreation, or combined). [This review of
audit instruments] points to 'insufficient attention paid to the instrument
validation process.' (p. 36) [Emphasis added.]

The determinants produced in the travel demand model that has
traditionally guided built environment-physical activity research tends to
emphasize origin and destination points as travel, but Moudon and Lee (2003)
conclude in their evaluation of audits that ‘route quality’ and ‘activities in the
area’ may be more reliable determinants of people walking and biking. “[T]he
focus is not so much on reaching a certain destination as it is on being engaged in
the act of walking and bicycling” (p. 23). That challenge to traditional demand
models was also made by urban design theorist Kevin Lynch (1984), that travel
should not be seen as “pure cost,” and that moving from place to place has as
much potential to contribute to quality of life as any other activity in the daily life
of the city dweller (p. 194).
10

McCormack, Masse et al. (2006) recently underscored such challenges to
assumptions about the audit development process using a statistical method to
investigate the construct validity with which audit findings are turned into indices
of “walkability” and other such proxies for quality of life (McKenzie, 2006).
Using the Rasch statistical model to test the construct validity of the well-known
Systematic Pedestrian and Cycling Environmental Scan audit (SPACES), their
findings suggested that developing a reliable quality of life index from audit data
was more challenging that previously assumed. They concluded that alternative
construct validity tests were needed during the primary research process, so that
the products of the audit process were properly held in mind during primary
research. This challenges traditional assumptions about how sets of determinants
are currently bundled into audits instruments, and it also suggests that the use to
which an audit’s information will be put (indices of health, etc.) should inform
primary research at least as much as primary research is assumed to inform
audits. That, I argue, challenges the theoretical foundations of the audit
development process, sufficient to justify the theoretical investigation conducted
here.
Given the gravity of what is needed from built environment-physical
activity in the context of the obesity epidemic, I propose first to develop a general
model for audit development, second to provide clarification on existing primary
11

research models and how they relate to the general audit model, and finally I will
examine a set of audit case studies in light of the general audit model.
I will also show the parallels that exist between the three different types of
audit information and three arguments in an ongoing moral-philosophical debate
regarding the definition of distributive justice. From those three arguments, I
identify three definitions of distributive justice, each of which is applicable to
each of the three kinds of audit information. With these, I go on to formulate the
general audit model (GAM) at the end of Chapter 1.
In Chapter 2 I use the GAM to argue that the scope of the ecological
model currently proposed in AL research ought to be extended to observe what
J.J. Gibson calls affordances. In Chapter 3, I undertake a multiple case study of
AL audits, to determine the degree to which each one gathers affordance
information.
Unlike primary research, audits are always normative, I argue, and always
distributive, and therefore, always contain a justice dimension and a criterion
representing the collective good. However, I observe that the roles of audits and
primary research are often conflated, resulting in an obscuring of the fundamental
relationship between distributive justice and environmental audit data. I go on to
describe eight defining dimensions of audits.
I will begin by describing how audits are evaluative, in that they gather
information about empirical phenomena for comparison to some normative
12

standard of the good. Second, I will argue that audits are necessarily distributive,
and so contain implications for distributive justice in terms of some standard of
equality. I go one to argue that an audit’s conception of the good that is equally
distributed by the built environment is ontologically linked to the audit’s
conception of how that good is distributed. Chapter 1 ends with the presentation
of a general audit model, comprising three distinct kinds of audit information:
1. Utilitarian audits evaluate the built environment in the form of
preference fulfillment or satisfactions.
2. A general resources audit evaluates the environment in terms of a
fair distribution of those “things which it is supposed a rational man
wants whatever else he wants” (Rawls, 1993, p. 181). These take the
form of an equality of opportunity.
3. The capability approach describes what activities the environment
actually “affords” for particular individuals, and normatively, how it
supports the desired functionings of the individuals. This is Sen’s
capability approach, described by me in terms of an equality of
affordance, a term that I take from the work of environmental
psychologist J. J. Gibson. I argue that affordance effectively operates
in capability terms, representing an equality of opportunity plus
agency.
13

Eight Dimensions of Built Environment Audits
Audits are Evaluative
Audits are always normative, and in this they are distinguished from
primary research. Even when the work of primary researchers is guided by
normative values regarding children’s health, primary research is not normative
by definition, while audits must always be so. Primary research in the built
environment-physical activity field has as its aim the establishment of
determinants of physical well-being, while audits normatively compare an
environment to those determinants. Thus as noted previously, audits are
evaluative, while primary research is investigative.
Built environment audits are also distinct from program evaluations. The
term “evaluative” in the case of audits refers to normative assessment, which
describes things that ought to be, as opposed to descriptive or ex post facto
evaluations of programs and outcomes, which describe things as they are. While a
program evaluation is “the systematic investigation of the merit or worth of an
object for the purpose of reducing uncertainty in decision making about that
object” (McLaughlin & Jordan, 2004, p. 27), the work of normative audits is to
develop “the generalizable connections between human values and settlement
form” (Lynch, 1984, p. 37). So, primary research establishes knowledge claims
about determinants of physical activity and their connections to human values,
and audits use them to evaluate goodness of the environment in some specific
14

human-environment context. In this regard audit models resemble research
models, in that they are metaphors signifying and simplifying a description of
reality (Bradley & Schaefer, 1998, pp. 23-25). Yet the audit model is dissimilar to
the research model in that audits are not “reality testing” (Hoover & Donovan,
2003, pp. 35-36). While audits could be said to offer hypotheses, AL audit
developers have not explicitly done so in the audit development process. Nor have
they offered knowledge claims explicitly in terms of falsifiability, which is a
standard for validating social science research findings (Bradley & Schaefer,
1998, p. 54).
Like primary research, audits involve discrete quantification of variables
in the sense of “counting the presence or absence of a thing” or “differences of
quality” (Hoover & Donovan, 2003, p. 24). However, in an audit the relationships
between the variables that make up determinants are not hypothesized, but rather
are presumed. Audits are not meant to demonstrate the validity of primary
research findings, but to establish a relationship between the built environment as
an independent variable and physical well-being as dependent variable, with
behavioral outcomes such as walking and biking as proxies for well-being.
The audit is necessarily informed by valid primary research, but not
necessarily the other way around. For example, if primary research determines
that the “absence of crosswalks” and the “presence of graffiti” are determinants of
the sort of physically active or inactive lifestyle that is associated with children’s
15

avoidance of obesity and other illnesses, then audit information might evaluate the
incidences of crosswalks and graffiti in the neighborhood as representing the
likelihood of physical activity.
Part of the reason that the distinction between audits and primary research
is often overlooked may be because of the differences between how validity and
reliability are established between research and audit processes. Generally, audits
invoke the validity of some physical activity determinant from previous primary
research, and then set about establishing the reliability of the audit based on its
replicability through test-retest or kappa scores or other methods to establish the
audit’s reliability for observing the determinants (Lee, Booth et al., 2005; Saelens,
Sallis et al., 2003; Troped, Cromley et al., 2006, for instance). The assumption is
that the high replicability of audit findings re-validates the determinants. If valid
primary research has determined that “perceived safety from crime” is a physical
activity determinant, and that it can be operationalized as the absence of graffiti
and the presence of street lighting, then an audit developer might use test-retest or
kappa scores, inter-rater reliability tests, or other tests to show how an audit
reliably gathers such information (Day, Boarnet et al., 2005, pp. 146-147). In such
a case, valid findings of primary research establish independent variables
(graffiti/street lighting, and perceived safety) and the nature of their relationship
to physical activity (negative or positive), and then the audit gathers information
on incidences of those variables and relationships in the neighborhood.
16

The Irvine-Minnesota audit instrument (IMI), for example, is based on an
extensive review of previous research findings regarding the “characteristics of
the built environment” that “influence physical activity” (Day, Boarnet et al.,
2006, p. 146). Evidence of the IMI audit’s reliability is offered in terms of its
strong relationship to valid primary research, and internal statistical reliability.
While it is important to establish that an audit’s observations are
consistent with valid research, and that the audit can consistently observe the
same determinants in replications of the audit, audits are still fundamentally
different from primary research models in that primary research establishes the
existence of a generalizable relationship between dependent and independent
variables, while the audit observes these in a particular place.
So what would a general built environment audit model look like? It
would need to be a deductive model, abstract and idealized, meant to guide the
practical development of tools for use in the world (Bradley & Schaefer, 1998, p.
32). In fact, despite the differences between built environment audits and ex post
facto program evaluations, an audit model would be somewhat akin to the logic
models used in program evaluation (McLaughlin & Jordan, 2004). Evaluative
logic models are meant to conceptually organize the flows of activity and
information in order to guide decision-making (p. 27). Program evaluations are
quite different from audits in one way––program evaluations observe the expected
behavior of employees, types of service, and productivity in order to establish the
17

efficiency or effectiveness of programs––while built environment audits observe
determinants in order to estimate the intensity or quality of their distributed
effects amongst people living in an environment.
Audits Are Normative
Primary research and audits may both have some “good” as their subject,
such as the health of children, but audits are directly concerned with the
distribution of those goods, whereas primary research has the option to claim
value-neutrality, with goodness as a background concern.
I argue that the good that an audit evaluates––the health of a
neighborhood’s children, perhaps described in terms of a neighborhood’s
“walkability” for instance––is always socially distributed. Audits are
distinguished by their normativity in that while primary research identifies
determinants of physical activity in the built environment, audits subsequently
interpret these as goods. Audits are necessarily normative.
Illustrative in this regard are cases where the same data serves two distinct
investigative and evaluative purposes. An example of this is Lynch’s “cognitive
mapping” method, first described in his “Image of the City” study (Lynch, 1960).
7

Lynch’s primary research generated findings on how urban inhabitants cognize
their experience of the built environment. His now-famous taxonomy of built

7
Although the term “cognitive mapping” is generally associated with Lynch’s Image of the City
study, Lynch himself did not use the term. Nor is it known whether he was even aware of Edward
C. Tolman’s famous “Cognitive Maps in Rats and Men” which popularized the term. See The
Psychological Review 1948, 55(4), 189-208.
18

environment features––“paths,” “edges,” “districts,” “nodes,” “landmarks”, etc.––
describe those elements with which residents orient themselves and organize their
experiences of the city. Lynch subsequently employed the primary research
findings as a basis for evaluating the environment for its “imageability” (p. 9). He
described this as an evocative quality of the environment, associated with the
resident’s feeling and sense of place, orientation, and meaning in relation to the
built environment. In that case, Lynch’s primary research findings were used first
in an investigative manner, for arriving at knowledge claims regarding how
residents perceive and respond to the environment, and secondly as performance
standards for evaluating the environment in normative terms.
Audits are normative in the same sense that planning interventions are
always a public matter (Harper & Stein, 2006, p. xviii). However urban designers
hope the environment will contribute to well-being, that good is always collective
in some fashion, and so must be distributed according to some conception of
distributive justice. This means that audits must not only consider what goods
they distribute, but also the justice with which individuals hold those goods in
relation to the collective. It is necessary for audits to conceive of the public
interest in terms of a relationship between individuals and the collective good.
That is to say an accounting must be made, however implicitly, of the individual’s
holdings of some good, but also of the equality of the individual’s holdings in
relation to all other individuals in the community.
19

Audits Describe the Good as Equally Distributed
In his Tanner Lectures entitled “Equality of What?” Sen asserted that
evaluations of distributive justice always take some equality into account, and that
this provides a fundamental rubric for examining methods of evaluating social
systems (Sen, 1980). He referred to this “ubiquitous egalitarianism” as an
inescapable requirement of any normative evaluation of the good (Sen, 1992, p.
3). Any of the “social arrangements that have stood the test of time . . . ” he tells
us, “want equality of something...They are all ‘egalitarians’ in some essential
way” (p. ix). He offers as examples the equal welfare demanded by welfare
egalitarians, the equal utility demanded by classical utility egalitarians, and the
equality of liberties demanded by libertarians (p. ix). Sen and Rawls disagreed on
various points regarding the evaluation of distributive justice (Sen, 1990). Yet,
they concurred that a “basis of equality” is a fundamental component of
distributive theory, and that it is the “conception of justice” that offers a way of
working out “the respect in which human beings are to be counted equal” (Rawls,
1999, pp. 441-449). A debate continues amongst moral philosophers today about
whether Rawls’ is a theorist of equality or not (Wolff, 2007, pp. 131-132), but
clearly Rawls himself recognized the need for an equality criterion. Equality is the
essence of what Rawls was after with his “original position” (Rawls, 1999, p. 20).
Mill’s argument for the ubiquity of equality is foundational. In his words,
the subject of equality “often enters as a component part both into the conception
20

of justice and into the practice of it, and, in the eyes of many persons, constitutes
its essence” (Mill, 1993, pp. 190-191). Aristotle addressed the need to consider
equality as a matter of arithmetical logic, arguing that since justice and injustice
are always a matter of “more” or “less” of something, and because “a more and a
less also admits of an equal,” then distributive justice will be the median point
between more and less of some good (Aristotle, 1962, p. 118). Sen (1992)
suggests that egalitarianism is ubiquitous at least partly because of the “powerful
rhetoric of ‘equality of man’ (p. 1), but also due to a Kantian instinct that deems
credible for one person only what is credible for anyone (p. 17). He also explains
the ubiquity of the equality criterion as a way of establishing the “ethical
plausibility” of proposals (p. 3).
Pertinent in this regard is the idea that “[E]thical theories all presuppose
some form of equality” (Nagel, 1979). In the planning field, Lichfield (1971)
grappled for many years with the ubiquitous role of the equality criterion in the
evaluation of urban plans (p. 161). Moroni (2006) suggests that Lichfield
recognized that strict egalitarian accounting of the costs of development across a
community could not report fairly on how individuals bear such costs
heterogeneously, but because of the ubiquity of the question of equality, Lichfield
continued to struggle to conceptually extend the meaning of equality without
disposing of the criterion altogether (p. 29).

21

Harper and Stein similarly recognize the ubiquity of the equality criterion,
but they go much further than Lichfield did in reconceiving equality in planning.
Taking their cue from Rawls, they argue that planning ought to value the
individual according to a “difference principle” that allows for an equality of
Rawlsian primary goods, which allows inequality of advantage so long as the
greater advantage is of maximum benefit to the least advantaged, thus
reconceiving equality in terms of difference (Stein & Harper, 2005, pp. 158, 165-
166).
Aristotle presents two useful ways to consider equality, either as an equal
portion of some good, or as an equal proportion of it. He reasons that there is an
inescapable need to focus on equality as regards distributions of the good, since
the just and the unjust are a matter of “more” or “less” of some distribution, and
because “a more and a less also admits of an equal” then distributional justice will
be the median point between more and less of the thing distributed: “Since the
equal is a median, the just, too, will be a median...Consequently, the just is
something proportionate” (Aristotle, 1962, p. 118). By proportionate he means an
“equality of ratios” between different persons and their variable amounts of the
same good, where A’s portion c is proportionately equal to B’s portion d, even
though their portions of the good are unequal.
8
So, when we are talking about two

8
Aristotle does not describe whether the relative relationship is regressive or progressive. However, in his
Inequality Re-examined, Amartya Sen takes up a closely related question in his arguments for the capability
approach, which has its roots in Aristotelian ideas. Sen argues that distributive justice in the face of relative
22

different holdings of the good, in Aristotle’s terms, we refer to a portion of a good
in some equally distributed amount, or else a proportion of the good, as an
“equality of ratios” (p. 119). In the proportional equality, the individual’s
relationship to her relative holding of the good makes it more or less valuable to
her depending on her situation.
A Hierarchy of Descriptive Categories Separates the Good
and the Built Environment, in Audits
It may seem tautological to argue that an audit’s conception of the good
and its built environment determinants are different, but evidence from existing
audits suggests that there can be confusion when attempting to distinguish
between the two. The distinction can be lost between a determinant’s being the
good, and its function as proxy for the good.
Within the evaluative space of an audit, the good and its proxies occupy
different positions. The individual holds the good, while its proxies are in the built
environment. To use a medical analogy, the presence of a cholera vaccine is not
itself physical well-being, but an indicator of it. The good itself is a dependent
variable or outcome such as health, and the independent variable is an input, in
this case the vaccine. Between the two is an evaluative space in which proxies for

proportions and competing demands could be evaluated according to an efficiency criterion. Sen does not
refer here to Pareto optimal efficiency, however, but to his own unique distinction between “shortfall
equality” and “attainment equality,” which proposes two kinds of equality. Shortfall equality evaluates the
degree to which A falls short of what he could maximally attain, and attainment equality evaluates the level
of equality that A has actually achieved (p. 90). Sen sees metrical advantages to both, depending upon the
disability causing the shortfall/lack of attainment (p. 144).
23

the independent variable may be found, such as “ability to digest food,” or “the
ability to work,” which are proxies of the good, but they are not the good in
themselves. To mistake the proxy for the good opens the door to the reification of
abstract concepts that can cause the evaluator to mistake the forest for the trees, so
to speak, potentially overlooking relationships between determinants, or
overlooking important mediating variables that influence the ability to make use
of what well-being one has. The good as something held in the individual is a
thing that affects the entire person on multiple levels. If we mistake “walkability”
for “physical well-being” we risk overlooking the other factors that play into the
person’s well-being and that influence physical activity. Complex, multi-level,
interactive ecological relationships are too easily overlooked.
As we will see in the case studies in Chapter 3, it is common for audits to
evaluate the proxy rather than the good. Most often, an inventory of determinants
is taken for indicating the good, with no consideration of what people are actually
able to do in the built environment. As an example, consider the Measuring Urban
Design Qualities audit (Ewing & Clemente, 2005) (MUDQ). In this case, the
ultimate good of the human being is described as “walkability,” which the audit
developers take for their dependent variable (p. 2). The independent variables
include determinants such as “human scale,” “transparency,” and “complexity,”
each of which has been developed through rigorous reliability testing. Field
researchers observe the presence of these variables in the form of data such as
24

“buildings...whose shape is not a simple rectangular box” and “where outdoor
dining is provided,” “the presence of pedestrians,” “noise levels,” and the
“presence of plazas and courtyards” (pp. 6-13). If there are enough of these
determinants in the environment, the street can be rated as “walkable” (p. 2).
A danger in mistaking the proxy for the good is in overlooking
relationships between determinants, in the sense that a deficiency of one
determinant might cause a decrease in some other. An inventory of outdoor cafes
and building shapes could conceivably overlook the thing that makes most streets
unwalkable, namely auto traffic. MUDQ only recognizes “noise,” while cars
could be associated with a variety of “setting aggravations”
9
such as pollutants,
crosswalk dangers, decrease in visibility, and others. Also, a designed street that
lies between points a and b may be the only efficient way to get from a to b, and
so counting the people who use it will only tell us volumes of pedestrian traffic,
and not the choice of that street over some another. Focusing on the proxies
fetishizes street qualities, to the detriment of the real issue at hand. If the
independent variable is “physical health,” and that is the end-state toward which
the street is oriented, the audit developers may considers options that they
otherwise would not have considered. (Actual examples of this are in Chapter 3.)
In what follows, I will offer a graphic representation of what I mean by the

9
Cf. Banerjee & Baer 1988.
25

evaluative space of a built environment audit (Figure 1). This is not the entire
general audit model mentioned earlier, but one half of it, as we will see.
The sideways pyramid in Figure 1 offers a way for the audit developer to
clarify the audit’s aims and values by clarifying the terms with which the good
and its determinants are described. The conceptual relationship between the good
and its environmental determinants is illustrated as a pyramid, with the good at
the smallest point on the left side, and built environment determinants in the
pyramid base along the right side. Between the largest layer and the smallest we
see another layer of description—proxies—for the independent variables. The
mistake is in thinking of the proxies as the good during the audit development
process, and thereby displacing the audit’s focus on values of wellbeing, off of
physical well-being for example, and onto things like “walkability.” Ultimately,
the proxies become the indicators of the good as a rating for policy purposes, but
during the audit development process, I argue, they should be seen as a way to
organize the conceptual space between independent and dependent variables. The
lowest layer includes crosswalks and graffiti, for instance, and the layers above
describe it in various categories including “safety” or “imageability.”
26

Suppose the dependent variable is y, the individual’s good, and y
represents physical well-being. Crosswalks and graffiti are the independent
variables x. They represent determinants of physical activity. Between x and y is
“walkability.”
Is walkability x or y? I argue that mid-level descriptions such as
walkability proxy for x, the independent variable, and not for y, the individual’s
well-being. The mid-level proxy x simply describes a condition that the built
environment ought to provide, namely that it be walkable.
The question of whether the mid-level proxy is x or y is important because
a proxy like walkability is too narrow a conception of well-being for the y
Th Environment The Individual

The good
(y)
Proxies
for the
good (x)
Built
environment
determinants of
physical
activity (x)
H
ealth
“Walkability”
“Access”
“Perceived
Safety”
(etc.)
Crosswalks,
Graffiti,
Distance,
Maintenance,
Bike paths,
Park acres per
1K,
etc.
Figure 1. A Hierarchy of descriptive categories between
determinants and the good
27

position, of the good. If we promote the mid-level proxy x to an artificially
elevated position of y, we have narrowed the definition of the good down to the
ability to walk. On the other hand, as a proxy for some quality of the built
environment x, walkability is one characteristic of the environment that
contributes to well-being, rather than being the whole of it. If the good is defined
in terms that encompass the entire life of the individual, then the mid-level of the
pyramid can contain as many proxies as required to define the contributions of the
built environment to well-being in appropriately complex and multi-dimensional
ways. Besides, walkability and bikeability, as y, narrow the definition of health to
those who can walk and bike, and exlcude those who cannot. As x, walkability
and bikeability are handy descriptors for families of determinants that contribute
to well-being, but are not the whole of well-being. Falsely elevating the proxy x to
the y position narrows the definition of well-being in ways that cannot help but
exclude individuals for whom things like walking and biking are not an option.
Audits Contain Ethical Assumptions
Audits are always concerned with collective good to some degree, since
their concerns are public policy oriented. Being concerned with collective good,
audits contain some conception of distributive justice, and in order to arrive at a
conception of distributive justice, it is necessary to define the value of the
individual in light of the collective. For primary research, determinants of
physical activity can be a value-free matter of scientifically establishing which
28

conditions predict the likelihood of physical activity. Because primary research
can focus entirely on likelihoods, AL primary researchers can produce
determinants that generalize across all people. It is after all one of the roles of
social science to produce generalizable knowledge. Audits, however, do not have
the luxury of being value-free, and they can generalize only up to a point.
Any evaluation of the built environment presupposes an ethical outlook
(Moroni, 2006, p. 21). This is the value in evaluation (Archibugi, 2006, pp. 73-
74). An audit tells the story of a good (health, e.g.) that affects either an
individual, or a group, or everyone. An audit of how the public space of a
neighborhood contributes to health must define, conceptually speaking, how that
good is distributed. It falls to the audit developer to determine what the good is,
who ought to have it, and how much they should have. These are the audit’s
values, and they are not simply an addendum to an audit, but are integral to it
because audits contain some criterion of collective well-being.
This link between values and evaluation is not a new idea among planning
and design theorists. The planner’s responsibility for explicating values has been
well articulated over the years. “When values go unexamined they are dangerous”
(Lynch, p. 34). In his “Ethics Behind Evaluation,” Moroni suggests that Lynch’s
maxim is the “central point” of considering evaluative ethics in planning. Moroni
cites a number of planning scholars who make the same argument as Lynch
regarding the need to explicate values in normative evaluation. Rauschmayer
29

argues that the choice of method used in environmental decision-making is
largely based on the “normative foundation of the analysis” (Rauschmayer, 2001).
House says that no matter what evaluation technique is chosen, it “presupposes an
ethical outlook” (House, 1980, p. 22). In his discussion of the values underlying
planning evaluations, Archibugi (2006) recognizes that any planning evaluation
has underlying values, and because these are most often left neutral or implicit,
evaluations tend to default either to an instrumental value-neutral position, or else
to make values explicit only regarding the ends of a plan, regardless of its means
(p. 75). He further argues that the “theory of value” and the concrete “context of a
decision” are integrally interwoven, and that the theory of value “depends
strongly on the process itself” (p. 77). In these calls for the explication of
underlying values in normative evaluation, planning and evaluation theorists are
in keeping with the more fundamental arguments or moral philosophers like
Rawls, who refers to a necessary “publicity” of values in his contract theory of
justice: “Society is not partitioned with respect to the mutual recognition of its
first principles” (Rawls, 1999, p. 510), and so publicity is necessary and
fundamental to justice, so that “citizens have a knowledge of the principle that
others follow” (p. 15). It is only with this publicity of values that parties can “rely
on one another to adhere to the principles adopted” (p. 153), which is fundamental
to decision-making because it allows us to better assess whether our undertakings
are in vain (p. 153). Rawlsian goods like civic participation and even self-respect
30

are based on reciprocities between persons, thus requiring public explication of
values such as planners might undertake. The assertion that “contemporary
planning theories [in western liberal democracies] rest on normative ethical
theories” (Stein & Harper, 2005, p. 148) has been developed at length in the
theory of planning offered by Harper and Stein, which draws on the Rawlsian
perspective (Harper & Stein, 2006).
Sen (2000), for whom “freedoms” are always the good in the form of what
he calls “capability,” similarly argues that the ability to participate in the selection
of values and priorities is “one of the strongest arguments in favor of political
freedom” (p. 30). When values underlying built environment-physical activity
research remain implicit, opportunities are lost for clarifying the goods upon
which communities––including their planners and designers––might develop
toward increased vitality. Explicit values are a necessary part of a conception of
the good. It is not the only part, however. I have said that to qualify as a good, a
thing must be equally distributable. Walkability was rejected as a good in the
Figure 1 example because it was not dividable equally amongst persons, due to
the heterogeneities of preference and ability amongst community members. This
sort of explication of underlying values is what I mean by developing a
conception of the good, although that was in general terms. The next two
dimensions of audits described below the process of explicating audit values in
more specific terms.
31

The value that the auditor assigns to the individual, then, is linked to the
nature of the good that is distributed. Because of the ubiquity of the equality
criterion, audits must assign values that address the heterogeneity of individuals,
whether to account for differences between individuals, or to dismiss those
differences. Indeed, the question of the heterogeneity of individuals is precisely as
ubiquitous as the equality criterion because equality cannot be defined in terms of
social goods without saying what any two persons in the society share equally, or
don’t. Below, I will describe how such values underlying evaluations may be
identified. The first has to do with the individual, namely whether the individual’s
equal apportionment is an absolute measure, or one that is relative to some
characteristic of the individual. The second has to do with the nature of the good
itself, whether it is deontological and an end in itself, or consequential and a
means to some end. In making decisions about these two theoretical notions,
namely the individual’s apportionment and the good’s consequentialism, the
auditor determines two key values that ultimately establish in very practical terms
what sorts of audit information will be available to planners and designers and
policy-makers.
Audits Value Individuals in Absolute or Relative Terms
In order to conceive of the good as equally distributable it is necessary to
address the difference between individuals, and this is done in audits in one of
two ways, according to how the auditor values the individuals in terms of an
32

apportionment of the good: In Aristotle’s conception, the good is held by the
individual either as an absolute portion (strict egalitarianism) or else accounting
for difference by apportioning the good as a relative proportion (equality of
ratios)
10
. The individual’s equal apportionment of a good may be an absolute
value, such as an equal portion held by each individual, or it may be a relative
value according to some characteristic such as ability, or deserving. For example,
if an equal distribution between two individuals––say, an elderly person 70+ and
a younger adult of 35––is based on their having an equality of holdings of some
good such as “safety from violent assault”––which is valuable and preferable for
everyone regardless of personal characteristics or circumstances––we might say
the good is distributed in terms of an absolutely equal portion to each resident. In
such a case the heterogeneities between persons are averaged, producing a metric
that reflects a homogenous value. On the other hand, if equality is based on
holdings that are relative to some individual determinant such as a physical
characteristic, age for example, then we could say that each person has a relative
value in distributive terms. We formulate the distinction between absolute and
relative values in terms of individuals having either an absolutely equal portion of
the good, or a relatively equal proportion of it. If the audit assumes that the
presence of qualities of the built environment such as “graffiti,” “bike paths,” and
“crosswalks” will increase the likelihood that residents of any age will be

10
What is equal in an ‘equality of ratios’ is the proportions that A and B hold relative to what they deserve by
grace of having some particular characteristic, such as a disability.
33

physically active, we can say that the audit assigns an absolute value to the
individual, in the sense that any resident is conceived as having an equal portion
of the good. If the audit determines that heterogeneous groups respond differently
to qualities of the built environment, producing different likelihoods of physical
activity, we can say the audit assigns a relative value to the individual, in the
sense that each group has an apportionment of the good that is relative to their
thresholds and abilities.
This latter criterion is, I suggest, a comparison of the differences between
persons such as has been called for in built environment research by Banerjee and
Baer, who at the conclusion of their study asserted that, "[T]he design paradigm
must provide for unequal design (inputs) that will result in more equal outcomes
for different groups” (Banerjee & Baer, 1984, p. 184).
Audits Conceive of the Good Either as a Means or an End
A vital aspect of an audit’s conception of the good is whether that good is
defined as an end in itself (e.g., having good health) or as a means to an end (e.g.,
the resources for pursuing good health). The difference between the two is
important in establishing an audit’s values, since an end in itself tells us what the
good is, while a means to an end tells us what the individual might do with the
good. The conception of the good as either a means or an end frames the audit’s
valuation of the individual in one of two ways, the choice of which is itself an
ethical question.
34

The issue of how to value the good would be simple if audit developers
could simply assign to variables the value of means or ends and be done with it.
The problem is that amongst goods that can be distributed equally, certain kinds
are only distributable as ends, and others are only distributable as means, and the
differences are not necessarily clear. Distinctions between means and ends are
further obscured when audit developers elevate the proxy description of x to the
level of the good y, as I will describe.
Consider the audit determinant “interestingness” in the Irvine-Minnesota
audit instrument (IMI) (Day, Boarnet et al., 2005). Is interestingness a means or
an end? The IMI assumes that the more interestingness there is in the environment
the more likely it is that people will be physically active. Interestingness is
assumed to provoke a feeling of interest in the resident, which stimulates exercise.
In that case, the environmental quality of interestingness is a means to physical
activity but is not the activity itself. Interestingness in that sense is equally
distributed as a resource, that is, a kind of emotional health promotion. Yet while
an environmental quality of interestingness may conceivably be distributed in
every neighborhood in the land, we cannot say the same about the feeling that it
produces in each individual. Nor can we say whether people have equal means to
respond to interestingness by becoming active. We must ask at some point,
“Interesting to whom?”
35

Conceivably, a person who never walks, and who only views the
neighborhood through a window, could still find the neighborhood interesting.
The feeling of interest may have value to the individual in ways other than
stimulating physical activity. Interestingness is a kind of resource for walkability
for some audit developers, but the resource is used by the active person as well as
the inactive. Referring to Figure 1, I would argue that although the IMI intends for
interestingness to be a means to health, it has instead elevated interestingness to
the level of y, the good, and defined the feeling as an end-in-itself. This limits the
scope of what the audit can tell us about the built environment’s contribution to
well-being.
Let us assume, like the IMI, that if a person has a feeling of being
interested in the environment that she will be physically active. Let us also
assume that an interesting environment can on average stimulate those feelings in
any resident, which means that the feeling has been “distributed” as a health
resource equally to all residents. Yet as I noted above there will always be
heterogeneous individuals who will not or cannot exercise, and so we cannot
assume that they will be as stimulated to physical activity as the majority of
people. Then, faced with the possibility of unfairly leaving certain heterogeneous
individuals out of the distribution, the IMI responds by describing interestingness
as an average feeling experience by all residents. The metrical average level of
interestingness reveals how the audit values heterogeneous individuals.
36

In terms of the Figure 1 pyramid, the IMI has dispensed with the middle
proxy level where “interestingness” was a means to the good, and proceeded to
ensconce the feeling of interestingness in the y position of the good. I do not deny
that the IMI definition of interestingness (landscape views, diversity of building
types, etc.) is a valid determinant of pedestrian activity. I am simply pointing out
that in distributing the feeling of interest in average terms it becomes metrically
necessary to define the feeling as an end rather than as a means. Such subtle
definitions and redefinitions are the unavoidable thickets of audit development. In
the case of the IMI’s interestingness, everyone can be said to have an average
amount of it, but what is averaged away is the instrumental value of being able to
say what a full, focused “dose” of interestingness might enable individuals to
actually do. I offer this as an illustration of my argument that an audit’s
distributive values will be defined and re-defined whether audit developers do it
with awareness or not.
Certain conceptions of the good cannot be distributed as an average. The
common planning measure of park acres-per-x persons, for example, is meant to
establish a distribution as a means to an end of recreation. By employing the
average, though, the acres-per-person metric loses the ability to say what
37

individuals are actually able to do with their acreage. Average distribution re-
defines acres-per-person as an end in itself.
11

In the first chapter, I began by describing what an audit is, and how it is
distinct from primary research, according to whether it is evaluative and
distributive. Inasmuch as they are distributive, audits must also conceive of the
good as equally distributed. These interconnections suggest that an audit’s
definition of distributive justice is not only connected to the good, but also to the
audit’s informational basis. What the audit can tell us about the environment is
integrally linked with its definition of distributive justice. I have also argued that
in order to determine what or how an audit conceives of the good, we must
explicate how it considers the individual. That consideration has to do with the
audit’s conception of the individual’s apportionment of the good, and also has to
do with the nature of the good itself as an end or as a means.
With these things in mind, the next section takes up three major
conceptions of the good, each of which represents a different distribution of the
good. These three conceptions are the choices that audit developers have in
drawing up their values, I argue. Choosing one or the other of them is unavoidable
in audit development, whether developers do it consciously or not.

11
This matter of equal green space is the subject of one of the audits discussed later (Sister, Wolch et al.,
2007).
38

There are Three Types of Audit Information
In Figure 2, I have attempted to graphically illustrate some of the points
made above. Figure 2 represents the individual in relation to the collective good,
in terms of a set of choices that the auditor makes.
There are three choices. I call them lines of valuation,and it is according to
these that an audit values the individual in relation to the collective good. The
individual and the equality of some good are at opposite ends of these lines of
valuation, and may be approached from either direction, as in a continuum.
Going from left to right, we first see the individual described in terms of
personal characteristics and/or social circumstances. Next the two lines of
valuation separate, offering choices of describing the individual either as holding
an absolute portion of the good, or a relative proportion of it. Following the upper
line, we can say that the metric for measuring the absolute portion is an average
across a population, and that this average functions to dismiss heterogeneity of
preference and ability in favor of some good that can be apportioned in absolute
equality.
Such a good is preference fulfillment, which can be operationalized as the
individual’s satisfaction. Satisfaction is a utilitarian end-state within the individual
39
Figure 2. Individual relationship to the collective good, along three lines of valuation.
Good as Means
40
such as a sense of safety (freedom from pain, in utilitarian terms). This line of
valuation is thus associated with a utilitarian conception of the good. Of the three
conceptions of the good discussed here, the most well-known may be the classic
Millsean definition of distributive justice as a Pareto-optimal average of
preference fulfillment, or equality of utility. Mill’s definition of distributive justice
is discussed, along with others, in detail in Chapter 2.
The lowest line of valuation would not be associated with utilitarianism
because the individual is viewed as having a relative proportion of the good,
rather than an absolutely equal portion. The lowest two lines would not be
associated with utilitarianism since they each allow for interpersonal comparisons
between heterogeneous individuals (children relative to the elderly, middle class
versus poor, etc.). Also, we see that the lower two lines of valuation frame the
good as a relative proportion, although in the middle line the good is a
consequentialist means to an end, whereas in the lowest line the good is a
deontological end in itself.
If the good is considered as a means, the individual’s proportion of the
good will be evaluated as some form of general resource, such as rights to
participation, or the access that safety provides. General resources are the things
that a person wants, ‘whatever else she wants’ (Rawls, 1999, p. 79). Equality of
general resources can be operationalized as a dependent variable of opportunity.
This “resourcist” conception of the good, as an equality of opportunity, can be
associated with a Rawlsian “justice as fairness” approach, or the Rawlsian theory
41

of justice (Rawls, 1999; Rawls & Kelly, 2001). On the other hand, if the good of
the heterogeneous individual is framed as both a relative proportion and an end,
then it is a good that is valued equally by individuals, but distributed in relative
terms, according to the characteristics and circumstances of the individual.
Relative proportions of the good are distributable as general resources, and the
equality of these resources can be operationalized as a dependent variable of
opportunity.
The final line of valuation is a very particular case. This line branches off
from the equality of opportunity only at the choice of means-versus-ends, which
makes a significant difference in how the good may be conceived. If the
individual’s good is valued as a relative proportion, but the good is also an end,
then the dependent variable is an opportunity. However, it is not only an
opportunity, since it is consequential. The good in this case is a general resource
that particular individuals may use heterogeneously in pursuit of some unique
end-state of well-being.
12
This good is a general resource to which all have
access, but individuals may use it only according to their particular characteristics
and circumstances, their agency. This can be associated with Sen’s justice-as-
capability.
The capability line of evaluation is also unique in terms of how “agent” is
defined, in Sen’s terms, as, “. . . someone who acts and brings about change, and

12
By “individual” I mean groups of individuals who share pertinent characteristics and/or circumstances.
42

whose achievements can be judged in terms of her own values and objectives . . .”
(Sen, 2000, p. 19). Sen tells us that evaluating a person’s agency to take
advantage of opportunities––their “capabilities” to pursuit a life they value––is a
two-way relationship.
These capabilities can be enhanced by public policy, but also, on the other
side, the direction of public policy can be influenced by the effective use
of participatory capabilities by the public (p. 18).

This definition of the good that Sen (2000) is describing not only enlarges the
ability of people to help themselves, but also to affect matters in the world as a
way of pursuing the good of one’s self as well as the larger social good. At the
end of Chapter 2, I argue for an ecological model for primary research based on
the work of J.J. Gibson, which describes environmental perception in the same
terms that Sen evaluates quality of life: The agency to affect things in the very act
of being affected by them, to change things as one is changed by them.
This good that I am describing as the individual’s agency to make use of or
to “convert” general resources in the pursuit of an end-state of well-being, we can
describe as opportunity plus agency. In Chapter 2, I will adopt Gibson’s term for
the individual’s two-way ecological relationship to the environment, as
“affordance.”
A General Model of Built Environment Audits
In Figure 3 we see the general model, which is a re-organization of
Figures 1 and 2 into a single unit. It represents the general structure of the audit’s
43

three possible evaluative spaces stretched between the individual (far left), and the
environment (far right). The three equalities of good are at the center of the
general model, and are shared by the hierarchy of descriptive categories on the
right hand side of the model.
On the left, each line of valuation is operationalized in terms of one of
three dependent variables representing the good, namely an equality of utility, or
of general resources, or of capability. The upper line is operationalized as
preference fulfillment, which is the utilitarian approach. The lower line’s two
branches observe either opportunity, which is the general resource approach, or
opportunities-plus-agency, which is the capability approach.
In an earlier discussion, I described the generating of determinants of
physical activity as an activity that is integral to audits. In Figure 3 we can
imagine the operationalizing of health and the built environment, for example,
arranged to suggest how determinants from primary research are situated within
audits. In choosing a line of valuation, we determine how individuals will hold the
good (relative proportion or absolute portion), and also a fundamental quality of
the good (end or means). For the audit developer, the choice is always made
between the individual’s characteristics and the characteristics of the
44

T

The good
Proxies for
the good
Built
environment
determinants of
physical
activity
Hierarchy of descriptive categories
(y)

(x)
(y)
Individual’s relationship to the good

Figure 3. A general model of built environment audits
45

environment. The individual may be described as a homogeneous member of a
group, or as a heterogeneous member of a group, but in either case the question of
individuality must be addressed in the metric. The model must simultaneously
describe a relationship between the individual good and the collective good, as
well as a relationship between the collective good and the built environment.
These three occupy separate positions but on a continuum. Along that continuum,
the individual, the good and the environment are dynamically related through
interstitial steps that tell us on one hand how the individual relates to the
collective good, and on the other hand how the built environment is related to the
good.
As there are three main combinations of determinants and distributions,
there are also three kinds of audit information. Each kind of information delivers a
different sort of evaluation of the built environment-physical activity relationship,
and each serves a unique role in the development of audits as decision-making
tools for planning urban design interventions. The individual’s good is defined
along the three lines by one of three conceptions, as satisfaction, or opportunity,
or opportunity plus agency. The built environment’s contribution to these is
defined by the hierarchy of categories in the triangle. Those categories
conceptually connect the built environment to the good.
46

Conclusion
The general model in Figure 3 is offered as the start of a conversation
about the kinds of information that audits might need to gather to address
particular issues, such as the obesity epidemic. In Chapter 2 I will show how the
competing theories of distributive justice of Mill, Sen, and Rawls, described in
audit terms, can help us in making choices about what information audits should
gather.
47

Chapter 2: Built Environment Audits and the Ecological
Model
Introduction
In Chapter 1, I attempted to demonstrate that in the Active Living
Research literature, determinants of physical activity represent goods that are
expected to be distributed in some equal fashion. Accordingly, I argued that only
certain kinds of good–– satisfaction, opportunities, and affordances––could be
distributed in equal terms. While available primary research does not describe the
distributions per se, they are implicit. I bring them out in order to conceptualize
how different assumptions about distributive justice influence audit tools. Primary
research and normative-instrumental design intervention are separate processes,
and the audit process is a necessary intermediate step between them. I argue that
developing the theoretical and distributive basis of audits can strengthen their
practical effectiveness as decision-making tools.
Audits do not observe the good directly, in that they do not actually
measure the health of children with stethoscopes and lab results, but employ
proxy variables that stand in for the good. However, because audits assume an
equal distribution of the good, that good must be defined as something that can be
distributed in equal terms. I have said that what good is distributed is linked to
48

how it is distributed. I maintain that these ideas are fundamental to audits, and that
they can inform the work of primary research.
I have said that primary research has the option to be value neutral,
whereas audits do not. By this I mean that determinants of the built environment
may be researched purely as a value-free phenomenon, should we wish to do so.
Audit development, however, does not have that luxury. I assert that even the
most solidly objective audit tool (several cases are discussed in Chapter 3) implies
a definition of distributive justice. I have argued that determinants of physical
well-being can be conceptually linked with an audit’s distributions of that well-
being.
The left-hand side of the general audit model is my proposal for how
audits can respond to the question “Equality of What?” (Sen,1980) (Figure 3).
There are three answers to Sen’s question: satisfactions, opportunity, and
affordance. These three equally distributed goods represent three kinds of
dependent variable in any given audit.
Because the independent variables are determinants of physical activity,
we can say that primary research has a role to play in informing the general audit
model. The role of primary research is represented by the right-hand side of the
general audit model.
I will discuss how each kind of good––satisfaction, opportunity,
affordance––represents one of three evaluative approaches to urban design: the
49

visual-cognitive tradition, the rights-based tradition, and the synomorphic
tradition. On a theoretical level, each of those echoes one of three competing
theories of distributive justice––Mill’s utilitarianism, Sen’s capability approach,
and Rawls’s theory of justice. The debates between Mill, Sen and Rawls have
theoretical ramifications in the three urban design traditions, but also in primary
research, and in audits.
In this chapter, I will use the generalized audit model as a framework for
discussing the tension between the two dominant conceptual models guiding
research on the built environment-physical activity relationship: the
“correspondence” and “ecological” models. I argue that the correspondence
model observes determinants of physical activity in terms of satisfactions, while
the ecological model observes determinants of opportunity, and particularly of
affordance. I examine the Active Living ecological model in light of Sen’s
capability, and conclude that the ecological model is uniquely representative of
capability, although some modifications are required.
Using the concepts of environmental perception as a common
denominator between the ecological models of Active Living Research, Roger
Barker, and J.J. Gibson, I will argue that the Active Living model lacks a
dimension of synomorphic interactivity, and two-way influences between human
and environment, that a complete ecological vision demands. While Barker’s and
Gibson’s ecological visions are similar, I conclude that Gibson’s construct of
50

“affordance” is more likely to describe capability in the built environment, than
the Barkerian “behavior setting” concept currently employed in Active Living
research.
In this chapter I will first attempt to demonstrate how different types of
audit information are identified; next I will show that capability/affordance
information is not particularly emphasized in current ecological models for
primary research; finally I will argue that the Active Living audits––and by
implication, Active Living primary research––ought to be expanded to include
affordance information.
I take issue with the AL use of Barker’s “behavior settings” as a feature of
the AL ecological model, in that Barker’s concept is used in AL research only in a
limited way, that does not describe ecological human-environment relationships
as Barker intended. It is the reciprocal and interactive dimension that is missing
(cf. Lynch, 1960). It may be that the cumbersomeness of Barker’s method makes
it inefficient for AL Research, which may explain why the concept is used in a
limited way. I mean to argue that it is possible to embrace a full description of the
ecological environment while working more or less within the current AL
research and audit development process. However, I propose that Gibson’s
“affordances” could efficiently replace behavior settings in the AL model.
Affordances and behavior settings have many dissimilarities, but are also
akin in important ways. Heft describes affordances as, “ . . . a class of
51

environmental properties operating at the level of the individual” (Heft, 2001).
Meanwhile,
Behavior settings are a class of environmental properties that reside
among extra-individual relations, with the result that any individual in the
setting can be replaced by another individual…and the setting will
continue operating (ibid.).

An affordance is the relationship between a toddler who is learning to
walk and the furnishings in her home: Because of the height of the coffee table
and nearness of pieces of furniture to each other, the child will pull herself up, and
if she falls, the sponginess of the carpet will cushion her. The room “affords” her
learning to walk. It is easy to imagine a room with no such affordance, that is,
with no furniture to hold onto, and a concrete floor.
While the identification of affordance is utterly dependent upon the
characteristics of the individual, the behavior setting defines individuals only
inasmuch as they are part of the setting. “Standing in line for movie tickets” is a
behavior setting, as is “getting a ticket from the state trooper.” Any person in the
situation will likely behave in a predictable manner, given the physical and social
milieu. Behavior settings often contain a diversity of people who have the
behavior setting in common, while affordance involves a group of people who
share a common characteristic that constrains their ability to make full use of
some environmental resource.
52

Ecology, Correspondence, and the General Audit Model
It is important to keep in mind that this study in no way makes
recommendations for what “satisfying” or “afforded” environments ought to look
like. Because this is a theoretical study, its ultimate aim is to inform empirical
research, but it is not itself the sort of empirical research that produces
determinants of physical activity. This is a study of models, and the concepts with
which the human-built environment relationship is evaluated. I do not intend to
make recommendations for how environments might make children exercise
more, except to argue that Active Living Research is missing some fundamental
components to its work that might make such recommendations more effective in
design policy terms.
It is my contention that an affordance audit can be modeled on the lowest
line of valuation in Figure 3, the capability line. I will propose Gibson’s
affordance as the good, in terms of the built environment, as opportunity plus
agency, or capability. Before considering how audits could be developed,
however, we must consider more deeply how they currently are developed. This
will make it possible to contrast the proposed affordance model with the current
AL models.
It has been proposed by Active Living researchers that the ecological
model can offer important information about how the built environment
contributes to the health of residents (Humpel, Owen et al., 2002, for example;
53

Sallis, Cervero et al., 2006). I accept that the assertion is almost certainly correct.
However, the current AL ecological model does not fully frame the built
environment-physical activity ecology in AL audits. In this chapter, I will use the
categories in the general audit model as a rough rubric for examining how the two
major kinds of primary research models inform audits, with an eye to determining
what an “ecological determinant” for audits might be. The two kinds of models
examined are the ecological model, as it is variously described, and also what I
will call the “correspondence” model, citing Heft (p. 42). The ecological and
correspondence models are distinct enough in their conceptions of the human-
environment relationship that they gather very different types of primary research,
and thus inform audits in quite different ways.
History and Purpose of the Ecological Model
Urban design theory contains a variety of traditions according to which the
environment is investigated, audited, and instrumentally altered to improve
quality of life. Various design theories evaluate the concept of the good and its
appropriate distribution, in diverse ways. I will examine some of those traditions
here, within the framework of the general audit model, in order to understand
which of them can be most closely associated with the correspondence and
ecological models. On a practical level this should begin to suggest the
relationship between those traditions and satisfaction, opportunity, and capability
variables.
54

It is also my hope that describing urban design’s normative-evaluative
traditions in this way will bring the field further into the ongoing debate amongst
moral philosophers about the nature of collective wellbeing. Citing J.S. Mill, John
Rawls, and Amartya Sen, I draw conclusions not only about which kind of audit
information is associated with which of those thinkers, but also which kind of
information and urban design theory is most appropriate for framing built
environment-physical activity research in the context of the childhood obesity
crisis.
The history of model development for Active Living research begins with
attempts to reach across rigid disciplinary research boundaries. At the first Active
Living Research Conference, the history of built environment-physical activity
research was traced to epidemiological studies in the 1970s (Sallis, Linton et al.,
2005). Because the built environment-physical activity field was previously
divided between various disciplines such as transportation, health, and leisure,
natural disparities existed between the conceptual models each field employed,
representing the unallied purposes of each separate field. In 1998, in their review
of policy and environmental intervention, Sallis, Bauman et al. had concluded that
current models guiding primary research on the built environment-physical
activity relationship were lacking in a kind of ecological comprehensiveness, and
that a new theory was needed (Sallis, Bauman et al., 1998, pp. 379, 390). By
2005, the Transportation Research Board’s (TRB) report on the relationship
55

between the built environment and physical activity noted that the right model for
primary research remained an open question, and that, “[O]ne of the primary
limitations of research to date on the relationship between the built environment
and physical activity is the lack of an agreed-upon theoretical framework”
(Transportation Research Board, 2005, p. 126).
By 2005, models for built environment-physical activity research had been
proposed in a variety of fields such as transportation, urban planning, preventative
medicine, recreation, and leisure, many of which had begun joining together
across disciplines beneath the Active Living research banner (Bedimo-Rung,
Mowen et al. 2005; Godbey, Caldwell et al., 2004; Nasar, 2007; Owen, Humpel et
al., 2004; Sallis, Cervero et al., 2006). The TRB was able to survey pertinent
literature from a variety of fields including travel behavior, physical activity, land
use, design, infrastructure, as well as material on attitudes and motivations (pp.
153-159). A number of models are reviewed in that report, including psychosocial
ones emphasizing alterations in health behavior, models of community
development that emphasize programs and social networks, and transportation
models based in demand theory (Transportation Research Board, 2005, pp. 90, 95,
100). In some cases researchers have deemed the demand model insufficient
because it did a poor job of representing non-motorized travel, which was the case
with regional transport models (Transportation Research Board, 2005, p. 126).
Ecological models are particularly noted for their ability to inter-relate
56

characteristics of the built environment with individual behavior at multiple
levels, although they are criticized for being vague in their descriptions of the
environment (p. 128). The TRB study’s own organizing model is a hybrid
ecological-demand model. Another indication of the ongoing lack of agreement
about a primary research model is reflected in the fact that while instrumental
urban design policy is ultimately the point of Active Living research, some
models leave this as an implicit issue, while others take it up explicitly (Craig,
Brownson et al., 2002; Ewing, Schmid et al., 2008; Frank, Schmid et al., 2005;
Stokols, 1992).
What is the Ecological Model Supposed to Replace?
The Correspondence Model
I assert that the ecological model of Active Living research represents a
fundamental shift in how built environment-physical activity research views the
human-environment relationship. AL researchers who propose ecological models
do so as an improvement on existing models (such as the demand model, etc.), in
effect an expansion of those models to include an ecology of factors affecting
physical activity. The introduction of multi-dimensional determinants of physical
activity is significant in itself, but there is another rationale in place, that is
responsible for a more fundamental shift: A number of researchers have argued
for the development of an ecological model that can include perception as a
variable in built environment-physical activity research (Boslaugh, Luke et al.,
57

2004; Humpel, Owen et al., 2002; Kirtland, Porter et al., 2003; Miles, 2006;
Nasar, 2007; Owen, Humpel et al., 2004; Prezza, Alparone et al., 2005; Timperio,
Crawford et al., 2004). I argue that with the introduction of perception into the
ecological model, Active Living research did more than simply improve upon
existing models, rather, the human-environment relationship was reconceived in a
way that could accommodate new sorts of observations, and new kinds of
information with which to audit the environment and its distributions of goods.
Ignoring the full ramifications of the shift to the ecological model could cause
researchers and audit developers to overlook new kinds of information about built
environment-physical activity relationships, and new ways to evaluate their
distributions through environmental audits.
The significance of introducing perception into the AL research model
becomes clearer when perception is considered as a subject on its own, outside the
built environment-physical activity field. In her discussion of the introduction of
ecological models of perception in psychology, Fleishaker refers to the
“traditional” model of perception, which the ecological model is meant to replace
(Fleischaker, 1984). Ecological psychologists Roger Barker and J. J. Gibson,
whose ecological models are discussed at length below, similarly viewed the
ecological conception of the human-environment relationship as a fundamental
shift (Barker, 1968; Gibson, 1966/1979; Heft 2001). I believe that the theorists
who developed the AL ecological model have not yet recognized its fundamental
58

implications in part because the field is currently focused on bringing its myriad
and even disparate research disciplines under a common research model.
13

Drawing from Barker, Gibson and others I argue that the AL ecological
model represents a fundamental shift in our conception of the human-environment
relationship away from the time-honored correspondence model. The ecological
model holds the promise of new sets of determinants of physical activity, and new
kinds of information to inform audits and their distributions of goods.
In the traditional correspondence model the individual compares sensory
input from the environment with a store of past cognitive and affective
experiences of the same type, to establish how the new experiences “correspond”
to those from the mental storehouse. Key elements of perception, in
correspondence terms, are the inputs apprehended from outside the person by the
receptor senses; the individual’s past experiences; the ability to rationally
compare past and current experiences; and finally, the individual’s affective
reactions to those comparisons.
The concept of correspondence represents a model of verification and
meaning that is an old argument amongst the philosophers of perception (ibid.).
Modern ecological challenges to the correspondence model date to James, and
Dewey, and the pragmatist challenge to the “stimulus-response” model of human

13
Researchers have called the lack of a common conceptual model the principle reason for the delayed
progress of the field, as the TRB report notes (Transportation Research Board, 2005, p. 126).
59

response to the environment. Barker and Gibson were products of that intellectual
lineage.
Consider a parent who encounters graffiti and litter in the neighborhood
along his child’s walking route to school: We might imagine that the parent
compares the current observations of graffiti and litter experiences, and decisions
are made accordingly. (See Figure 4). My point is not that models such as Figure
4 are intentionally framed to describe correspondence. On the contrary, Figure 4
is associated with ecological AL research. My point is that the framework of
Figure 4. Example of "correspondence model" of environmental perception
60

correspondence is persistent and ubiquitous, and is common in this kind of
research. The description of the linear reaction from stimuli to correspondence to
response is not meant to suggest a conscious process on the part of the perceiver.
The most important point that I will attempt to make here is that correspondence
models for research––all of them in the built environment-physical activity field–
–view perception operating in a single direction, from the environment to the
perceiver. Most of the models discussed in Chapter 3 clearly conform to the
correspondence model’s singularity of direction. As I will describe at the end of
this chapter, Gibson argues for information that flows in two directions, at once,
For Gibson and for Barker, the correspondence model ascribes to the perceiver
complete responsibility for assigning meaning to the environment, while the
ecological model describes a dynamic relationship between person and
environment.
In the correspondence model, incoming stimuli are thin information,
which is given context and meaning by its correspondence with the perceiver’s
previous experience. Policy decision-making about the built environment based in
this model relies upon determinants that are presumed to cause parents’ minds to
change about allowing children to walk to school, for example.
Three separate authors, Heft, Fleishaker and Lombardo each trace the
historical development of the distinction between the correspondence and
ecological models to the distinction between Plato’s and Aristotle’s descriptions
61

of perception, respectively. Although many thinkers have argued this subject, the
Plato/Aristotle argument is fundamental, and therefore useful for providing major
categories for constructing models of built environment-physical activity
research.
Plato argued that timeless, unknowable universal forms lay behind the
phenomenal world, and it is in these that the ultimate order and meaning of
phenomena resides. Phenomena, on the other hand, are particularized and
perceivable matter. In Plato’s idealistic conception, we order and give meaning to
our perceptions of phenomena by generalizing from the perceivable particulars to
the universal unknowables. In Plato’s dualistic model of knowable particulars and
unknowable universals, the perceiver draws from a mental storehouse of
associations that are then used to generalize the meaning of the particulars.
Conversely, Aristotle’s unification of form and matter describes a world
that is empirically knowable within concepts of order and temporal flow
(Fleischaker, 1984, p. 42; Lombardo, 1987, pp. 362-363). The mind must order
the phenomena it perceives as in the Platonic model, but not entirely, since for
Aristotle phenomena contain some meaning of their own. In the Aristotelian view,
as Nussbaum and Putnam tell us, the structure of our understanding is something
we share with the world, and share in common with other perceivers (1995). If
phenomena have no immaterial unperceivable Platonic forms, they can be
62

empirically understood by separate individuals, so long as the sharing is “realized
in matter” (cited in Heft, 2001, p. 280).
Wohlwill’s early argument on this subject was titled simply “The
Environment is Not in the Head!” (p. 285). Heft summarizes that position:
…if environmental psychologists construe environmental experience as a
private, intrapersonal phenomenon, then an unbridgeable gap exists
between the common world that perceivers share and co-exist in, and the
idiosyncratic, intrapersonal “worlds” they inhabit (Heft, 2001, p. 280.).

In the Aristotelian view the structure of how a thing is perceived is in a sense
transferred to the perceiver from the thing itself (pp. 274-277). Also, we might
say that in the correspondence model the meaning of a perception is inferred from
correspondences between memories of experiences and new ones, a “matching
process between the characteristics of currently available particular instances and
a representation stored in memory of the gist of previous similar experiences”
(Purcell, 1986, p. 3). So, the individual receives sensory input, against which past
memories are compared, and then the individual’s affective response occurs when
there is a match, or mismatch, “between attributes of the current instance and the
attributes of the prototype” (ibid.). In such correspondence models, incoming
information is thin, and is enriched by mental and emotional material from the
‘storehouse of memories.’ With the correspondence model, primary research
might quite reasonably set out to determine which qualities of the environment are
strongly associated with walking, so audits could gather information on what
amount of those qualities the average resident has, and then urban design might
63

provision the environment with those qualities in places and amounts that can be
said to raise the average individual’s holdings of the good.
An alternative to the correspondence model, the ecological model
describes incoming information as thick, laden with meaning within which the
individual discerns. In the ecological model the perceiver is not a passive
recipient of behavior-causing stimuli, but a seeker of information about the
environment. The ecological model “proposes that the aspects of context that are
most influential as determinants of human behavior are those that are consciously
perceived and interpreted by individuals” (Clitheroe, Stokols et al., 1998 p. 104).
Lombardo cites Gibson’s decidedly Jamesian description of ecological
perception, saying “Perception is an accomplishment,” that “should not be seen as
a series of discrete acts with beginnings and endings” (Lombardo, 1987, p. 354).
The above discussion is meant to contrast, in theoretical terms, the
differences between the correspondence and ecological models. Below, I will
relate three urban design evaluative traditions and their underlying theories of
justice to the correspondence and ecological models. I describe the three
evaluative traditions of urban design in terms of three different theories of justice.
Typologizing evaluative traditions of design according to their respective theories
of justice will bring the evaluative traditions of design into the ongoing debate
amongst schools of moral-philosophical thought regarding distributive justice and
the evaluation of collective good. The aim here is to consider whether there may
64

be a kind of evaluative information that is best suited to Active Living research,
and how it might be best modeled for establishing the determinants that are
subsequently used in audit tools.
Three Normative-Evaluative Approaches to Auditing the Built
Environment
Satisfactions: The Visual-Cognitive Approach
Nasar’s Evaluative Image of the City is premised on the idea that a
“likeability” factor can be aggregated from the opinions and tastes of groups of
urban residents (Nasar, 1998, p. 59). Likeability describes a collective perception,
by groups of urban inhabitants, of certain qualities in the built environment that
produce perceptions of pleasure (p. 3). Alternately, likeability represents a
reduction in anxiety and fear (ibid.).
Nasar’s claim for the efficacy of his method is based in Lynch’s 1960
Image of the City study, the taproot of the visual-cognitive tradition of urban
design (Lynch 1960, pp. 10-13). Although Lynch never did employ the concept of
“likeability” as Nasar does, and avoided the term “aesthetics,” he was keenly
aware of the power of purely visual information in shaping our experience of the
city. Image brought rigorous analysis to the study of such information.
Amongst the visual-cognitive material that is of concern to Active Living
audits, one inescapable criterion is the environment’s “attractiveness” (Day,
Boarnet al., 2006; Saelens & Sallis, 2002). The condition of the environment is an
65

important contributor to attractiveness (Byrne, Wolch et al., 2005). The
“openness” of the space is taken seriously into account (Saelens, Frank et al.,
2006). “Transparency” and “Legibility” are also important in this audit tradition
(Ewing & Clemente, 2005). Most of the AL audits that I will study in Chapter
Three place great reliance upon the pedestrian’s visual-cognitive experience of the
built environment as a stimulant to physical activity. The visual-cognitive
experience, an important dimension of the rationale for the AL ecological model,
is described in great detail in Nasar’s study, making it a useful resource. The
Nasar study is also an exemplar of what was referred to earlier as utilitarian
satisfaction.
The emphasis on pleasure in Nasar’s study, or at least the absence of pain,
is reminiscent of Bentham’s definition of the good as either “pleasure itself” or at
least the “exemption from pain,” which was a fundamental conception upon
which Mill later built his theory of utilitarianism (Mill, 1993, p. 143-144). The
similarity between the defining principles of Nasar’s method and Mill’s is not
happenstance, but speaks to the persistence and popularity of Mill’s “greatest
happiness” principle as a metric for evaluating collective good.
Nasar advances and expands Lynch’ groundbreaking visual-cognitive
methods, bringing in new data collection methods and analysis, but the underlying
visual-cognitive evaluative focus remains the same as that in Lynch’s 1960 study.
Nasar quotes Lynch: “[T]he shaping and reshaping of the city ‘should be guided
66

by a visual plan: a set of recommendations and controls…concerned with visual
form on the urban scale’(Nasar, 1998, p. 2). To accomplish this, Nasar proposes
to study “how the public evaluates the cityscape and what meanings they see in it:
the evaluative image of the city” (ibid.). Nasar’s Evaluative Image framework,
also like Lynch, offers to determine “the probability that an environment will
evoke a strong and favorable evaluative response” (p. 3). Nasar categorizes
evaluative responses into a framework that includes categories of fear, pleasure,
aesthetic delight, meaning, and most importantly, “likeability” (p. 93). These are
all affective states, produced by the fulfillment of preferences, and the assumption
is that maximizing these in positive or negative directions (pleasure-positive, fear-
negative) is a good to be maximized. In terms of the general audit model
introduced in Chapter One, we can say that increasing likeability increases the
amount of the good.
Nasar’s likeability is teleological, an end in itself, essentially a feeling of
pleasure that residents share in more or less equal portion. While it would
certainly be the case that some individuals in Nasar’s study have stronger or
weaker reactions to a landscape, Nasar’s method averages those reactions to
identify the shared experiences of pleasure and meaning that are most common
across a population, as a consensus (p. 25). He develops design recommendations
based on those averaged qualities of environments, those that are “liked” by the
most people. Right action, for urban designers, is seen as maximizing the good
67

across all inhabitants in the study population (pp. 93-94). It is to the credit of
Nasar’s framework that it brings method and rigor to what could otherwise be a
highly subjective and purely intuitive design process. Here, the designer is not
simply guessing at what people want, but is rather establishing a normative stance
through rigorous and rational method.
Implicit in Nasar’s averaging of responses is the assertion of the equal
moral worth
14
of each member of the population, and a definition of distributive
justice in which each person receives an absolutely equal share of satisfaction.
Nasar’s clear emphasis on preference, and the role of public resources in
satisfying preferences, especially the emphasis on human affective states like fear
and pleasure, points to the underlying utilitarian principles implicit in his work.
The Evaluative Image assumes an underlying utilitarian correspondence model of
the human-environment relationship, in which urban inhabitants receive visual
information about the built environment, and add to it their affective responses in
the form of feelings, based on previous experiences (see Figure 4). There are a
number of advantages to this approach. First, it should be noted that in his Good
City Form Lynch turned a critical eye on the common habit amongst urban

14
For clarification’s sake, let it be stated that no one has proposed an alternative to utilitarianism’s equal
valuation of individuals as “unequal valuation of individuals.” The arguments against utilitarianism described
above challenge the idea of valuation of individuals. This is not because Rawls and Sen for instance do not
value the individual, indeed they consider individual liberty and rights as being of the highest priority in any
just society. Rather, it could be said that they are each arguing that a way must be found to express the well-
being of the population of residents in light of the obviously unequal effect of distribution––some have better
access, some have worst aggravations to contend with, some are subject to certain societal forces, some are
not, etc. (Moroni, p. 29, f 33)––but without resorting to a “single quantifiable universal denominator”
(Moroni, p. 26).
68

theorists of using the utilitarian metric, of pleasure or not-pain. Lynch used the
metric himself, but argued that pleasure/not pain is not all that life is made up of
(p. 369).
One of the advantages of utilitarian satisfaction as an evaluative focus is
that it can be framed in terms of everyone having an equal share of it. This is
particularly important in evaluations of public investment. Another advantage to
evaluating with a utilitarian metric is that the individuals who comprise any urban
population will have heterogeneous preferences, which complicates the evaluation
of distributions of goods like walkability and bikeability, since some individuals
will prefer those modes and some will not.
15
Further, the utilitarian metric offers a
very clear method for aggregating diverse preferences.
However, by focusing on a utilitarian average of satisfaction, preference is
aggregated or “sum-ranked” into what Bentham called “The sum of the interests
of the several members who compose it” (cited in Campbell & Marshall, 2002, p.
167). This approach to distributive justice as the greatest good for the greatest
number is as clear to policymakers today as when Mill first described it, and it can
still be found at the heart of contemporary evaluations of social good, and
utilitarianism has justifiably been called “the most influential theory of justice”
(Sen, 2000, p. 60) The utilitarian metric’s averaging of preference fulfillment,

15
Yet, it is possible to evaluate the averaged preferences of diverse groups within a single population, and
Lynch was aware of this, and I do not dismiss the idea. I simply point out that Nasar does not do this, and
neither do any of the AL audits that use cognitive-visual-type conceptions of perception.
69

along with the assumption of Pareto optimality as a social welfare function, has
“reigned in discussions [of the collective good] over the past two centuries”
(Mueller 2003 p. 597). We should not be surprised that the utilitarian metric
guides urban design evaluative traditions, and indeed we should recognize its
value. However, we must also recognize its limitations.
The utilitarian metric is criticized for putting the evaluator in the position
of an imaginary average individual, who can “…discern the ‘real’ interests of
individuals better than they could themselves…” (Rawls 1999 p. 163-164). That
sort of instrumentalism was one of the critiques leveled at utilitarianism by moral
theorists including Rawls. In applied planning research, theorists since the latter
half of the 20th century have been critical of the utilitarian metric for a variety of
reasons (Campbell & Marshall, 2002). The value of the utilitarian metric persists
amongst planners and designers, yet its flaws have not gone un-noted, for instance
in Klosterman’s early critique of the utilitarian position, or Stein and Harper’s
Rawlsian critique of utilitarianism in their “dialogic planning” (Klosterman, 1978;
Stein & Harper, 2005, p. 157). Sen challenged the utilitarian averaging of
preference from a number of positions, including the unique argument that the
poor tend to adjust their preferences downward to suit reduced circumstances,
giving them an unequal (and unfair) weighting in such evaluations (Campbell &
Marshall, 2002).
70

No challenges to utilitarian thinking in urban planning to date focus
specifically on the built environment audit. The closest to this would be the 1984
study Beyond the Neighborhood Unit, in which Banerjee and Baer showed how
an environment’s fulfillment of the average resident’s preferences can reveal what
residents may or may not do (Banerjee & Baer, 1984). Banerjee and Baer were
essentially relying upon a fundamental correspondence/utilitarian-type model,
although in their conclusions they noted the need to expand research to somehow
include “what determines preferences” in terms of income, social inequalities, a
resident’s knowledge of the neighborhood, possessions, “holdings,” and other
mediating variables that determine preference must be part of the equation, and
cannot be described in terms of averages (p.154). Their conclusions suggest the
need for multiple-dimension determinants that consider individuals as holding
relative proportions of the good. This set the stage for current calls for critiques of
the utilitarian metric in planning, I believe, and for a new way to think about
distributive justice in terms of heterogeneous groups of individuals rather than in
terms of averages. "[T]he design paradigm must provide for unequal design
(inputs) that will result in more equal outcomes for different groups” (p. 184). I
suggest that Banerjee and Baer’s book was an early call for an ecological model
to guide urban design research (p. 193).
Theirs was not the only call, however. It is well-known that planning
theorists have long been frustrated with the use of the average as a metric for
71

collective good, at least since the 1960’s. We will briefly consider how planning
evaluation theorist Nathaniel Lichfield, for example, struggled with traditional
aggregative metrics, which he argued were an oversimplification of planning
impacts on human communities. Lichfield’s work was not concerned with
perception, but was certainly concerned with the unequal costs of planning
practice among heterogeneous community members, and it frames the problems
with the correspondence/utilitarian model in useful way. For all the differences in
Lichfield’s and Nasar’s methods, their approaches have the averaging of
preference as a fundamental concern. However, where Nasar accepts such
averaging unquestioningly, Lichfield is more critical.
Lichfield theorized about evaluating plans, and over the years he sustained
and deepened his critique of utility-based cost-benefit analysis and impact
analysis approaches, both of which he had championed earlier in his career. He
came to believe that the two methods limited planners to a utilitarian efficiency
criterion that overlooked the competing interests of individuals:
[P]eople are not homogeneous but must be seen as actors with conflicting
interests in any project proposal or plan; the sectors cannot all be
beneficiaries, since some must lose. (Moroni, 2006a, p. 65).

For Lichfield, the planning field’s long-held utilitarian emphasis on
optimal distributions of satisfaction across a population obscured difference (p.
32). His argument was that the aggregation of costs couldn’t represent
intrapersonal comparison: “The impact on the community is the impact on the
72

people in that community,” he declared in his criticism of the utilitarian metric’s
“aggregative conception of the public interest” (p. 64).
Ultimately Lichfield’s alternative was to abandon utilitarian sum-ranking
while still holding on to the concept of preference fulfillment, although he
expanded the definition of “fulfillment” to include residents’ equal access to
rights and positions of authority in the planning process (p. 30). His distributive
justice was no longer an equality of strict averages, but more of a ‘satisficing
minimum’ of costs distributed among residents, plus the ability to participate in
decision-making (pp. 30-31). To Lichfield’s credit he recognized and argued for
the value of public discourse and recognized the need to make account of
pluralities.
Critiques of Lichfield’s conclusions point out that while he was able to
“revise” the utilitarian focus, his continued reliance upon a consequentialist
definition of well-being as preference fulfillment defied the possibility of a real
alternative, such as a deontological definition of well-being as an equality of
“social primary goods” (Angrosino & Mays de Perez, 2000; Moroni, 2006b, p.
20).
In terms of the general audit model, we could say that Lichfield’s critics
are pointing out that he was attempting to operate on two lines of valuation at
once. Lichfield found a way to consider the unequal costs of planning, but
evaluated their distribution in terms of a utilitarian average of fulfillment.
73

Unfortunately, such a revised utilitarianism only partly answers Banerjee and
Baer’s call for “new contextual values” (Banerjee and Baer,1984, p. 196). Yet,
Lichfield’s efforts prefigured recent work by AL researchers Sallis, Cervero et al.,
who argue that, “As is true in other social sciences, researchers are recognizing
that if only averages are considered, important aspects of the experience are
missed” (p. 310). Sallis, Cervero et al. have pointed out that the averaging of data,
a foundation of survey research analytical method, is over-emphasized in built
environment-physical activity research. Similarly, the 2005 TRB report on the
built environment-physical activity relationship framed this same issue in land use
terms, challenging the methodological shibboleth of averaging, for its inability to
observe the actual effects of built environment interventions on travel within very
local areas:
Statistical averages are normally used to represent intrazonal travel. This
can understate the impact of mixing land uses or improving pedestrian
ways within a city block in promoting walking because averages ignore
any variation around the mean (Transportation Research Board, 2005, p.
110).

Also, we should consider that in some cases the concept of averaging is at
issue in research when the actual term “averaging” is not explicitly used. When
the TRB refers to a “gap in current knowledge” about how context, culture and
identity operate as mediating factors in the decision to be active, they are referring
to a gap that must be filled by information that is indescribable by metrical
averages. Like Sallis et al., and Banerjee and Baer, the TRB report calls to extend
74

the scope of research to include an accounting of how determinants of physical
activity are mediated by interpersonal comparisons regarding “subpopulation,
urban setting, climate, and other contextual factors” as well as “differences in
various sub-groupings” and “people’s lifestyle preferences and attitudes” (p. 173).
We see the utilitarian model at work in AL primary research models that
inform the development of audit tools, and its frustrations are noted by AL audit
developers Day and Boarnet, who call for researchers to consider the specific
characteristics and identities of groups of individuals, which their IMI audit could
not (Day, Boarnet et al., 2006, p. 150). Their audit “might have generated more
novel responses if participants had been questioned about how their own identities
(as teens, low-income individuals) influenced their opportunities for active living”
(p. 150). In general audit model terms, Day and Boarnet are operating from one
line of valuation, but recognizing the need to be in another. They recognize that
the correspondence research model sum-ranks such identities into an average that
limits the kinds of determinants that audit developers can use.
Sallis, Cervero et al. emphasize the need to take into account the sorts of
subjective information and complex interactions that occur between humans and
their environment, and that these relationships can define individual identity. This
transcends the averaging required by the correspondence model:
Though survey research dominates the leisure and recreation literature,
there is a recent trend to include interpretive, qualitative, and case
studies…Therefore, in attempting to understand not only the “what”
75

related to physical activity, leisure researchers are finding the questions of
“how” and “why” to be of importance. (p. 310)

Particularly applicable to audit tools, it should be noted that the limits that
the correspondence model places on what the designer can know about the
human-environment relationship also limits what recommendations might be
made in design interventions.
[E]qual design can have unequal effects with respect to, say, the elderly,
and how any given design has a cost that can be afforded better by some
groups than by others. Are equal standards for all really equitable when
incomes are not? (Banerjee & Baer, 1984, p. 184)

We must ask if an audit’s assessment of the average person’s satisfaction
with a neighborhood park can fairly represent those children, for example, who
are not allowed to use the parks because of parents’ gender biases about safety, or
those children who are raised by elderly grandparents who are unable to bring
children to the park. An audit of satisfactions in the correspondence model does
not reveal how particular characteristics of the environment support the physical
activity of individuals. Such audits are unable to evaluate, inter alia, how the
experience and perception of the environment are shaped by the characteristics
and circumstances constraining the individual's ability to make use of some
resource. After all, we can only assume that crosswalks increase the safety of
everyone in the community if we can also assume that everyone in the community
is capable of crossing the street and is desirous of doing so, which may not
necessarily be the case with obese and diabetic children, for example.
76

Despite the criticisms of utilitarianism, its “bad breath” is not easily
avoided (Campbell & Marshall, 2002). One reason for the persistence of the
utilitarian metric in urban design is that by focusing on the satisficing minimum
or average, it offers planners and designers a way to describe policy objectives in
terms of something that everyone can conceivably receive equally, namely
preference fulfillment. In order for normative evaluations of the environment to
recognize difference, their definitions of collective well-being must be couched in
terms of individual development, in which heterogeneous persons hold portions of
goods that are relative to their characteristics and circumstances.
With respect to the question of values, we are assuming that the new
contextual values will be largely individual...the new approach is: How do
we design the residential area so as to achieve well-being in each resident?
(Banerjee & Baer, 1984, p. 196)

The audits that inform such designs may include some information based
on an average of preference, but they would also need to include information
about the characteristics and circumstances that make individuals more or less
able to engage the built environment in physical activity. In general audit model
terms, the development of an ecological model by AL researchers is an attempt to
move off of the utilitarian line of valuation, in order to take complex constraints
into account as mediating variables. As Lichfield’s critics pointed out, however,
to consider the individual requires leaving the utilitarian metric behind altogether.
The most outstanding critique of utilitarianism in the 20
th
century was
John Rawls in his Theory of Justice (Rawls, 1999; Rawls & Kelly, 2001). Rawls
77

described a fundamental error in the Millsean Greatest Happiness Principle, that it
requires the evaluator to extrapolate the assumed choices of an imaginary
individual into a population-wide social choice (pp. 23-24). Rawls tells us that in
order to establish a normative utilitarian position, one must usher inan “impartial
spectator” in whose mind an equal distribution of utility is accorded each
individual in equal measure (p. 24). We can say that Nasar’s urban designer is
such an impartial spectator, in the Evaluative Image of the City. The Rawlsian
challenge to the impartial spectator is that no such average person exists, unless
we intentionally overlook differences between people. Nasar’s designer sees a
kind of ‘preferenscape’ in which everyone equally shares in the same level of
fulfillment, like water finding its level in a variety of lakes. Rawls pointed out that
the idealized observer would be unable to take seriously the distinction between
individuals. If a built environment audit reveals that every home in a
neighborhood has equal proximity to a jogging trail, it can only evaluate this as an
equality of preference fulfillment if everyone in the neighborhood prefers jogging
to other forms of exercise.
Utilitarians traditionally start with a clearly individualistic claim. In fact,
they usually start by saying that individual preferences or interests are
what really matters in the public realm. The problem is that the utilitarian
aggregative criterion of maximization of total utility (as a simple sum of
all the individual utilities and disutilities) disavows the original
individualistic claim: the single individual in fact gets completely lost in
an aggregative supra-individual calculus (Moroni, 2006a, p. 60).

78

The correspondence model such as Nasar’s in Figure 4 might seem to
ideally represent the individual since it describes an individual’s personal,
psychological experience of the environment, but Rawls rejected the idea that any
such individualistic claims could be made when an average is employed (Rawls,
1999, p. 26). The Evaluative Image preferencescape averages the satisfaction that
residents receive upon viewing the urban landscape, and in so doing lets
individuality go. In this approach, an audit’s informational basis will be informed
by visual-cognitive “attractiveness” and “transparency” and other such
determinants; their impact on individuals will be assessed according to the
satisfaction that they bring, on average, to neighborhood residents. Satisfaction
will be taken as the good, and walkability as a proxy for the necessary
combination of determinants. If individuals are satisfied with the landscape, it will
assume, they are more likely to be physically active.
It should be apparent that there is a strong relationship between research
models, audits, and their underlying ethical assumptions about distributive justice.
Our next step is to consider alternatives to the correspondence model, by
considering alternatives to utilitarianism. I have described a variety of opinions
about the correspondence model, and its metrical cousin, utilitarian preference
fulfillment, and now we will consider alternatives to these, one of the most
forceful of which was Rawls’s Theory of Justice and the planning frameworks
that it inspired. Rawlsian “general resources,” which I refer to as “opportunities,”
79

offer a complete alternative to utilitarian satisfaction. They also suggest distinct
differences between the correspondence and ecological models. The innovation
proposed by the Rawlsian Theory of Justice is nothing less than an alternative
conception of distributive justice. I will argue that in the same way that the
utilitarian theory of justice underlies the correspondence research model and the
visual-cognitive design evaluative tradition of design, a similar relationship exists
between the general resources theory of justice, the AL ecological research
model, and the rights-based tradition of urban design evaluation. The two
approaches are also similar in that they define the types of data that would be
observed by an audit based in that model, tradition, and theory of justice.
It would be difficult to deny that the pleasure that children take in the
physical and aesthetic qualities of parks is an important factor in evaluating the
park’s success. They run screaming and laughing down hills, splash in fountains,
climb trees, play games, hide under picnic tables and watch the world go by. It is
easy to watch a single child at play, or even a dozen, and roughly evaluate the
success of the park. It is quite another thing to evaluate the park in terms of all of
the people who will visit it over a five-year period. According to what criteria do
we evaluate goodness? In utilitarian terms, children’s preferences for the park
may be averaged by undertaking a visual-cognitive study such as Lynch’s or
Nasar’s in order to determine what makes parks “legible,” and what orientation
they provide, or what meaning. Or, still in utilitarian terms, we could count
80

children’s visits to the park and assume those who come are there because they
“prefer” the facilities, landscaping, and amenities offered by the park. Or, a
utilitarian survey of park visitors could also determine what sorts of things park
visitors prefer on average. The problem with the utilitarian metric is that it limits
what we can know of the individual’s relationship to the park to how an
individual feels about it. The evaluative emphasis is on the inner affective state of
the individual park visitor, so that even if we make counts of the numbers of
children actually using the park, in utilitarian terms those numbers do not speak of
experience, but they are considered a proxy for children’s satisfaction with what
the park provides. A utilitarian metric is focused on the inner affective state of the
child, and can tell us about the physical environment and how it is used only
indirectly. Further, because the park designer must come at preference indirectly,
she must put herself in the position of knowing what park visitors want better than
they know themselves.
The above examples express the Rawlsian critique of the utilitarian
evaluative approach. They also illustrate the limitations of utilitarian audit
information, which is limited to averages and estimates of preference. It is here
that Rawls proposes his alternative to the utilitarian evaluation of the collective
good, and steers our attention to another urban design evaluative tradition.
Rawls’s alternative to the utilitarian approach is a matter of shifting the evaluative
focus to an entirely different emphasis, as I will explain.
81

Rawls proposed that the utilitarian must average satisfaction first because
utilitarianism is consequential, in that the good being measured is an end-state of
being (satisfaction). Another reason is because the utilitarian evaluator, in an
attempt at being fair, assigns an absolute portion of the good to each individual.
This overlooks two key points of justice: First, the evaluation overlooks those
people who do not prefer what everyone else prefers. Second, the evaluation does
not take into account the different abilities and situations of individuals.
Rawls proposed instead that the emphasis should be placed not on an
equality of end-states like satisfaction, but rather on an equality of things people
prefer ‘regardless of whatever else they prefer.’ Rawls called those resources
“primary goods,” which are “things which it is supposed a rational man wants
whatever else he wants,” (Rawls, 1993, p. 181). A general resources audit does
not take inventory of things like “pleasant views” and “adequate sidewalks,”
however, because everyone does not use these equally, they are not “general”
resources. Rawls proposed things like “basic rights and liberties,” “freedom of
movement” “free choice of occupation and diverse opportunities,” and other
general means, as primary goods (Mueller, 2003, p. 609). These are the place-
based general resources that Kevin Lynch describes in his Good City Form, as the
goods “which seem to be the keys to obtaining other goods” (Lynch, 1984, p.
227). Rawls argued that if the focus were taken off of an equality of utilitarian
ends and instead placed on an equality of means, two things were automatically
82

achieved: (a) the utilitarian averaging of preferences was rendered unnecessary
because the primary goods are things that everyone wants, and so preference is
not an issue; and (b) the heterogeneities of preference became moot, since diverse
individuals could take their equal share of primary goods and do with it what they
liked. We can say that while traditional utilitarianism maximizes the mean, the
primary goods approach “maximizes the floor” (Sen, 2000, p. 74). The
distribution of the primary goods can be described in terms of ‘what is best for
everyone’ also being ‘best for those with the least resources to work with.’ The
Americans with Disabilities Act––the actual federal law–– is an example of a
general resource, since it is meant to maximize the floor for the disabled, so that
everyone––able and disabled alike––has equal opportunity to pursue whatever
diverse ends they choose. The general resources approach is not concerned with
how people use the resources, as that is a matter of heterogeneous preference.
Rather, it is concerned with what I will refer to as general resources or
opportunity.
Opportunity
The Rights-Based Approach
In their “dialogic planning” approach, Harper and Stein argue that rather
than focusing on the preference fulfillment of all members of a population,
planners should instead focus on guaranteeing that all persons have an equitable
share of the primary goods that are obtainable in the form of place (Stein &
83

Harper, 2005, p. 165). These place-based primary goods are what I am calling the
opportunities that people need in order to pursue whatever place-based ends they
wish. Here we will consider Harper and Stein’s ideas, to determine whether the
arguments they would produce align more with the correspondence or ecological
models, and how that would inform a built environment audit. We will then ask
the same questions of a park equity study conducted in Los Angeles by Sister,
Wolch et al. That study very clearly defines justice as a set of general resource
opportunities, I argue, and it may serve as an excellent example of how primary
research on available opportunities can inform built environment audits.
There is no exact laundry list of opportunities in the general resources
conception, and so we must carefully consider how justice-as-equality-of-general-
resources would play out in urban design research. “[I]nstitutional forms are
embedded within the conception of justice,” John Rawls tells us in his Theory of
Justice, but we must keep in mind that the Theory is not a model of practical
organizational theory or even contracts and agreements (p. 231). It is not to be
literally applied to shaping institutions but is a “procedural rendering” of the
background principles of a well-ordered society (p. 233).
[O]ur aim is not to describe and explain how people actually behave in
certain situations, or how institutions actually work. Our aim is to uncover
a public basis for a political conception of justice… [Emphasis added]
(Rawls & Kelly, 2001, p. 81)

What role would such a conception of distributive justice have to play in a
built environment audit tool? Rawls tells us that because his Theory is a “social
84

contract” theory, it is required to “move toward just institutions as speedily as the
circumstances permit”(Rawls, 1999, p. 396). We must consider what a general
resources conception of distributive justice looks like, played out in the
institutions and circumstances that are the context of built environment audits and
also the models for built environment-physical activity research.
Rawls tells us that that in his approach we must conceive of rights,
liberties and opportunities as having priority among a person’s concerns (Rawls,
1999, p. 28). This idea is the key. It is not the role of a just social system to fulfill
an individual’s aims, but to “define the scope” within which the individual
develops them, and pursues their ends, and we must imagine that the individual is
conscious of the role that rights, liberties and opportunities play in this. So, in
thinking about rights-based evaluative traditions in planning and design, the goal
is to identify the public basis for an equality of opportunity, in the form of the
opportunities that people want in order to achieve whatever ends of well-being
they have reason to pursue (Rawls, 1993, p. 181). In general audit model terms,
the Rawlsian approach departs from the consequentialist line of valuation and its
equality of satisfactions, to focus instead on the deontological means with which
people may pursue what ends they wish.
If opportunities are perceivable in the built environment, then one of the
two AL research models, correspondence or ecological, will describe how they
are observed. Because the general resource approach is procedural, it is unlikely
85

that an audit of opportunities in the built environment would be a simple
inventory of environmental characteristics, although the physical environment
would likely be a part of the audit.
Is it possible to audit a children’s environment, say a public park, in
Rawlsian terms? Rawls tells us that we must give priority to the opportunities that
we have in a particular institution or circumstance, and I suggest that this would
be part of such an audit. In built environment-physical activity research, the
opportunity for physical activity is typically seen through a rights-based lens
which influences perception of the built environment, and that perception informs
decisions we will make about our choices and behaviors. In Rawlsian terms a
public park is first of all a palpable bundle of rights to which citizens have access,
for instance in the ability to temporarily claim parts of it for private use by laying
down a picnic blanket between certain hours of the day, or the ability to alter its
ecosystem by driving sports vehicles on its trails, or holding a birthday party
there, and other such rights to achieve individual functioning needs. In this
conception, the actual physical park, its lawns and fountains, its plantings and
restrooms, and its play structures or spaces, are sensory data that the individual
associates with access to certain desirable opportunities to picnic or ride dirt bikes
and so forth. In Chapter 3 we will examine what I would call a Rawlsian or
general resources ‘opportunities’ audit, the SAGE/Green Vision Plan. Sage is
86

discussed in this section, below, as a broad evaluative framework for urban
design.
Which of the two research models, correspondence or ecological, best
describes the Rawlsian general resources environment? We must consider how
opportunity is perceived by the individual. In the correspondence model, the
senses receive thin information that the individual then enriches from the mental
storehouse of “rights, liberties and opportunities.” In the correspondence model,
Rawlsian determinants occur in the mind of the individual park user, and data on
the particulars of the park––swing sets and fountains and lawns––are mere
referents that activate the individual’s mental associations to rights-based
opportunities. In the ecological model, the individual is conceived as
investigating the environment, and the particulars of the park are sensory data that
arrives already laden with meaning about “rights, liberties and opportunities” for
park use. In the ecological model, Rawlsian determinants are a kind of
confirmation of what the individual is already seeking. In Rawlsian terms, we
might ask of a child who sees a park through the window of a passing bus, does
she see a set of opportunities for play, or a set of familiar objects that signal the
possibility of play opportunities?
The “Green Visions Plan for 21
st
Century Southern California” is a multi-
phase primary research project that offers an example of primary research that
would inform a general resources audit of opportunities. It is also an example of
87

an evaluative approach to built environment intervention that is rights-based. The
Green Visions Plan is not funded by the Robert Wood Johnson Foundation’s
Active Living Research project, although there are obvious affiliations: The
Green Vision Plan employs an audit instrument called SAGE that is adapted from
an AL audit called SPACES (Pikora, Bull et al., 2002). Also, the rationale for the
Green Visions Plan includes the effects of the childhood obesity epidemic in inner
city Los Angeles.
Released in 2007 as one of the Green Vision Plan studies was a report
called “Park and Open Space Resources in the Green Visions Plan Area” (Sister,
Wolch et al., 2007) (POSR). The POSR authors undertook an extremely detailed
“inventory and assessment of…recreational open space assets” to determine
“[T]he environmental justice dimensions of park supply” (p. 115). The POSR
studied over 1,800 recreational parks and open spaces in the Southern California
region (p. 7). The study developed an audit instrument with which researchers
inventoried parks for their facilities, equipment, and conditions (Byrne, Wolch et
al., 2005). Spatial analysts mapped data on the distribution of park space relative
to Southern California neighborhoods, and also on park infrastructure, facilities,
amenities and maintenance (p. 7). The study ultimately revealed gross inequalities
in the spatial distribution of recreational spaces, as well as discrepancies in access
and maintenance, controlling for “specific race/ethnic populations…as well as
socioeconomic class” (p. 115).
88

An earlier study by Wolch, Wilson et al. used an “equity mapping
analysis” method to determine the equality with which access to park space was
distributed amongst children and residents according to their race/ethnicity and
socioeconomic status. That study also mapped the distribution of municipal park
improvement and expansion grants. In the study findings, the authors concluded
that “Neighborhoods with the largest shares of young people received half as
much [park improvement and expansion] funding on a per youth basis than areas
with the least concentration of youth” (Wolch, Wilson et al., 2005, p. 3).
We must note that the POSR’s approach is distinct from a study like the
Evaluative Image in the same way that the Rawlsian Theory is different from
Mill’s utilitarianism: The POSR’s focus is ostensibly on distributions of amounts
of park lands, but its underlying assumption is that the unjust distribution of park
land reduces the child’s equal opportunity to play. More, while lack of park
resources is the POSR’s ostensible focus, it implicitly assumes that if park
resources were increased children would get more physical exercise. This
assumes that children will recognize that certain resources are available to them.
Rawls describes this recognition of opportunity as procedural or “background”
justice, which is the set of procedural principles that define an outcome as “fair”
(Rawls, 1999, p. 74-75). I point this out to distinguish the justice of equal park
space per child, and the justice of the child’s recognition that resources are
available to them. As Rawls explains it, humans can intuitively recognize the
89

conditions of fairness so long as they have a stake in it, and he offers the
metaphor of a group of people cutting a cake fairly: The procedure for fair
distribution is what he calls the background principle that makes the outcome fair,
and all of the cake eaters are aware of that procedural principle. The satisfaction
that the people take in eating the cake is not part of the Rawlsian evaluation, but
their desire for the opportunity to eat cake is. The difference between this
approach and a utilitarian one is that this is focused on procedures-as-input, and
fairness-as-output, whereas in a utilitarian scheme the input is a mathematically
equal distribution of cake, and pleasure of eating is the output.
General resources are opportunities to do something, and opportunities to
do things are precisely the sorts of things that we might reasonably expect
children to pursue, which suggests that the child’s environmental perception is
one of recognizing the presence or absence of opportunities in a variety of forms
which give her more or less access, such as play facilities, even the heights of
drinking fountains, or the “defensibility” of the space (Newman, 1972/1976). The
POSR notes a variety of perceptions that would affect the child’s opportunity to
play, beginning with the sense of safety, since those who perceive a lack of safety
are less likely to use the park (Sister, Wolch et al., 2007, pp. 19, 21). The POSR
observations also consider access in terms of availability of information (p. 25),
participation in park planning for the disadvantaged (p. 32), and technical
assistance to help the same disadvantaged neighborhood residents in preparing
90

park funding proposals (p. 25, 32). I would argue that in this we see the POSR
approach as distinct from the approaches that are more squarely situated in the
correspondence model, such as Nasar’s Evaluative Image. In the POSR, the good
is not evaluated as an equality of pleasure or satisfaction, but rather as an equality
of opportunities for pursuing one’s ends. The child on the passing bus sees the
park much as she would a birthday cake that is about to be fairly divided, trying to
determine if she has an equal opportunity to participate. Primary research that is
modeled on the perception of opportunities would produce metrical determinants
of physical activity representing opportunities that the child recognizes in the
environment, rather than utilitarian determinants that predict the child’s feeling of
satisfaction in response to the environment.
16

For the evaluator to imagine the child recognizing her opportunities means
imagining the child perceiving multiple levels of variables, including not only
swing sets and water fountains but also the effects of park policies, transportation
access, and other such complex constraints. That begins to suggest what AL
researchers refer to as the “trans-disciplinary” dimension of the ecological model.
In this conception of the human-environment relationship, the environment does
not provide satisfaction, but opportunities for satisfaction, and it is framed as

16
This is Gibson’s “education of attention,” a description of the distinction between perceiving an ecology of
“affordances” in which the perceiver compares her needs to what the environment affords, versus a
perception of correspondences with past experiences: “If what things afford is specified in the light, sound,
and odor around them, and does not consist of the subjective memories of what they have afforded in the past,
then the learning of new meanings is an education of attention rather than an accrual of associations” (Gibson,
1966, p. 320).
91

though these are perceivable by the child, and will motivate her to engage the
environment in physical activity.
Yet, I do not want to rush to the conclusion that the POSR is based in the
ecological model until we have further clarified our understanding of both the
opportunity-based approach and what AL researchers hope to gain from using an
ecological model. We have not fully clarified how the Active Living definition of
“ecological” differs from that of ecological psychologists like Barker and Gibson,
for example. For the moment, we will consider one more example of a rights-
based normative-evaluative tradition of built environment intervention, the New
Urbanism-inspired “Principles for Inner City Neighborhood Design” guidelines
for HOPE VI inner city housing redevelopment projects (“Principles”).
A range of criticism and praise has been leveled at the New Urbanism
paradigm, from excoriation to exhortation, but I do not take a position here on
relative merits of the New Urbanism. Rather, I am concerned with the Principles
as a very special case of collaboration between the Congress for New Urbanism
(CNU) and the federal Department of Housing and Urban Development (HUD),
one that uniquely emphasizes what I am calling a general resources approach, or
an equality of opportunity. We are concerned with the Principles as a unique
document from the rights-based normative-evaluative tradition of urban design.
Very little of the broader New Urbanism literature is as explicitly rights-based as
92

the Principles, although the movement is certainly underscored by a social
agenda.
17
Brain summarizes the social agenda of the New Urbanism as:
…a practice for constructing a comprehensible and pragmatically feasible
ecology of choice, engaging individual action in a collaborative project
that balances freedom and responsibility for collective outcomes at
multiple levels…

The New Urbanism program is a social one, Brain assures us.
The agenda of the new urbanism implies…a politics and a sociology that
has to do with the substantive empowerment of communities that might be
realized in and through the self-conscious production of humane and
livable places (Brain, 2002, p. 234).

I am not arguing that projects based on the Principles always achieved the New
Urbanism ideals, but merely that the Principles are an example of a set of rights-
based normative-evaluative design codes that focus on Rawlsian procedural
justice as general rights, in the sense of what the Principles call “communities of
opportunity” (HUD/CNU, 2000, p. 2). Like the Principles, HUD’s “Habitat II”
report––which served as a foundational document for the Principles––is emphatic
in its calls for “Educating all segments of the population to gain the skills to
participate fully in the economic, social, and civic life of the community in a way
that promotes sustainability” (HUD, 1996, p. 48). Each of the Principles describes
some dimension of the redeveloped built environment providing the poorest
residents with access to what Rawls calls “rights, liberties, and opportunities, and
income and wealth” and even “a sense of one’s own worth…” (Rawls, 1999, p.

17
Not even the famous Charter of the New Urbanism is as detailed and explicit as the Principles regarding
urban redevelopment as a set of rights and opportunities (CNU, 2001).
93

79). While the Principles do not specifically cite Rawls or any other theory of
distributive justice, they are clearly aimed at procedural and institutional fairness
in their concern, for instance, with citizen and community involvement in the
design process; the redistribution of economic opportunity; strategic infill to
“conserve economic investment and social fabric”; citizen control in future
development through design standards and codes; fair redistribution of economic
and political opportunity to underrepresented groups; and the role of the built
environment in self esteem (pp. 4, 15, 32, 9, 27 respectively) (see Figure 5). These
are all the means to pursue ends, rather than utilitarian ends of satisfaction and
pleasure, or environmental determinism. The contrast between the Principles and
Nasar's Evaluative Image is clear, as the former emphasizes the means to pursue
one's ends, and the latter is the end in itself. In keeping with a general resources
approach, the Principles take difference into account by evaluating the
environment in terms of the equal opportunities with which individuals may
equally pursue whatever heterogeneous ends they please. The Principles
recognize the presence of inequalities of wealth and privilege, and the distorting
effect of gentrification on inner city families as well as on the investments of
developers, and argue for development procedures that increase the fairness with
which residents hold opportunities. The Principles accept that some benefits of
development will accrue to the most advantaged, but they attempt to arrange
94

Citizen and Community
Involvement
Engage residents, neighbors, civic leaders, politicians, bureaucrats,
developers, and local institutions throughout the process of designing
change for neighborhoods.

Economic Opportunity
The design of neighborhood development should accommodate
management techniques and scales of construction
that can be contracted to local and minority businesses.
Diversity
Provide a broad range of housing types and price levels to bring people
of diverse ages, races, and incomes into daily interaction—
strengthening the personal and civic bonds essential to an authentic
community.
Neighborhoods
Neighborhoods are compact, pedestrian-friendly, and mixed use with
many activities of daily life available within walking distance. New
development should help repair existing neighborhoods or
create new ones and should not take the form of an isolated “project.”
Infill Development
Reclaim and repair blighted and abandoned areas within existing
neighborhoods by using infill development strategically to
conserve economic investment and social fabric.
Mixed Use
Promote the creation of mixed use neighborhoods that support the
functions of daily life: employment, recreation, retail, and civic and
educational institutions.
City-wide and Regional
Connections
Neighborhoods should be connected
to regional patterns of transportation and land use, to open space, and to
natural systems.
Streets
The primary task of all urban architecture and landscape design is the
physical definition of streets and public spaces as places of shared use.
Neighborhoods should have an interconnected network of streets and
public open space.
Public Open Space
The interconnected network of streets and public open space should
provide opportunities for recreation and appropriate settings for
civic buildings.
Safety and Civic Engagement
The relationship of buildings and streets should enable neighbors to
create a safe and stable neighborhood by providing “eyes on the street”
and should encourage interaction and community identity.
Provide a clear definition of public and private realm through block and
street design that responds to local traditions.
Dwelling as Mirror of Self
Recognize the dwelling as the basic element of a neighborhood and as
the key to self-esteem and community pride. This includes the clear
definition of outdoor space for each dwelling.
Accessibility
Buildings should be designed to be accessible and visitable while
respecting the traditional urban fabric.
Local Architectural Character
The image and character of new development should respond to the best
traditions of residential and mixed use architecture in the area.
Design Codes
The economic health and harmonious evolution of neighborhoods can
be improved through graphic urban design codes that serve as
predictable guides for change.

Figure 5. The HUD/CNU Principles of Inner City Neighborhood Design

things so that those advantages will also compensate the least advantaged, as in
Rawls's "difference principle" (Rawls, 1999, pp. 65-68). Wealth generated by the
95

development must include local businesses, and increased property values are
meant to benefit those whose homes are near the project but not part of it.
Residents, we may presume, perceive all of this, since the Principles are
essentially a set of performance characteristics for urban design.
My purpose here is not to champion or dismiss either utilitarian or general
resources frameworks, or correspondence and ecological research models. The
Rawlsian approach remains quite open to critique. Pogge (2002) sees "resourcist"
approaches like Rawls’s as unable to consider important diversities, since they
distribute resources in absolutely equal terms. Pogge points out that the resource-
based approach strongly emphasizes freedom of choice, and implicitly assumes
that those who live in polluted air, for example, are no worse off than those in
clean air, since those in polluted air have the ability and right to move. By
bringing the environment into his argument in this descriptive way, Pogge’s
argument suggests that the individual’s perception of opportunities might be more
a matter of discernment and implicit trade-offs between resources, to discern what
the environment will support.
Variations in environmental conditions...can influence what a person gets
out of a given level of income…It must be said, however, that actual
resourcists have paid insufficient attention to variations in environmental
conditions. They are not discussed by Rawls, for instance, and play no role
in his criterion of social justice (Pogge, 2002, pp. 22-23).

Similar to Pogge’s critique of resource-based approaches, Amartya Sen’s
critique of the Theory says that an equality of general resources alone cannot tell
96

evaluators what people can actually do with those resources. This is a charge to
which Rawls responded, pointing out that in the Theory the principles of justice
are those that all people would accept if they were all unaware of how they would
personally benefit from the enactment of the principles. Given that condition,
Rawls says, individuals who could not make use of a proposed resource would not
accept it if their individual limitations prevented them from using it (Rawls &
Kelly, 2001, p. 170). Rawls makes a good point, yet Sen counters that the causes
of poor health, and the ability to maintain good health, are complex and may
transcend even the high priority that Rawls places on rights and liberties, which
would obviously affect how the individual perceives the environment (Sen, 2000,
p. 72). The POSR authors seem to agree with these critiques of the rights-based
approach. They conclude their study saying that the scope of future research
should be extended to consider more multidimensional interactions between
characteristics of a target population (the lack of physical fitness among children
of low-income immigrant neighborhoods, for example) and how such
characteristics interact with specific circumstances (crime and gang activity, for
instance) to determine how these limit the child’s use of park resources (Sister,
Wolch et al., 2007, p. 115). The POSR recommends that future research should
employ a more nuanced definition of “access,” for example, beyond simply the
presence of certain resources and facilities, or even a child’s proximity to them.
Access must also refer to how an individual’s ability to make use of certain
97

resources is constrained by individual characteristics and social circumstances, as
well as how our perception of those abilities constrains us. The Green Vision Plan
authors recognize that their current SAGE park audit cannot reveal the individual
characteristics and circumstances that prevent individuals from making use of
park resources, regardless of the equality with which those resources are
distributed.
I have attempted to draw a connection between the general resources
rights-based approach and the idea of environmental perception, to raise the
question of whether urban design audits ought to model place-based opportunities
on the correspondence framework, or a more ecological one. I see evidence that
the rights-based approach operates in both the correspondence and the ecological
model. A correspondence model is suggested by the POSR’s assumption that the
child’s recognition of opportunities is based on the child’s previous knowledge of
what is available to them. Yet, the child is not modeled as experiencing
satisfaction as a motivation to physical activity, but more in terms of recognition
of opportunities to pursue one’s desires. I suggest that for a child to recognize
what opportunities are available to her would require a somewhat more ecological
set of determinant variables that include not only the various qualities of the park,
but also what the child knows of them, and even how she came to know them. So,
there appear to be aspects of both the correspondence and ecological models at
work in the general resources approach. We must hold off on further conclusions
98

in that regard until we can get a more specific picture of just exactly how the
ecological model works, and what AL researchers believe it can do for them.
What is the Ecological Model Supposed to Do for AL Researchers?
Active Living researchers offer the “ecological model” of the built
environment-physical activity relationship as an improvement upon previous
Active Living primary research frameworks (Humpel, Owen et al., 2002; Sallis,
Cervero et al., 2006; Spence & Lee, 2003). The value of the ecological model is
increasingly recognized in Active Living research, for allowing researchers to
bring the question of perception into the built environment-physical activity
relationship, in the sense that decisions and ability to exercise are constrained not
only by the objectively measurable environment, but also by subjective
perceptions of it (Clitheroe, Stokols et al., 1998; Hoehner, Brennan, Ramirez et
al., 2005; Nasar, 2007; Sallis, Cervero et al., 2006, for instance). Many AL
researchers have weighed the merits and weaknesses of the ecological model
(Humpel, Owen et al., 2002; Nasar, 2007; Ramirez, Hoehner et al., 2006; Sallis,
Cervero et al., 2006; Spence & Lee, 2003). Spence and Lee point out an important
trap in using the ecological model, the “ecological fallacy” of “inferring
individual-level relationships from relationships observed at the aggregate level”
(Spence & Lee, 2003, p. 17). One value of the ecological model that AL
researchers offer is that it is “transdisciplinary,” in that it describes the
determinants of physical activity in multiple levels at once, from the individual to
99

the institutional (Godbey, Caldwell et al., 2004; Humpel, Owen et al., 2002;
Sallis, Cervero et al., 2006). Also, it is pointed out that ecological models
incorporate intra- and extra-individual influences (Spence & Lee, 2003).
The most recent, and most ambitious, ecological model of the “Four
Domains of Active Living” is that offered by Sallis and Cervero et al. (see Figure
6). As in any case of the ecological model, the Four Domains model is an attempt
to describe human activity and the environment as a multi-layered phenomenon of
complex interdependencies, particularly regarding the question of the role of
perception in determinants of physical activity. The precise meaning of
“ecological” remains fluid amongst AL researchers, and this may be cause for
some concern. This fluidity is evidenced in the various meanings attached by AL
researchers to Barker’s “behavior settings,” which is employed variously by a
number of separate authors (Humpel, Owen et al. 2002; Sallis, Bauman et al.
1998). The term “behavior setting” is used variously in Active Living research to
mean the “physical and social contexts in which behavior occurs” (Sallis, Bauman
et al., 1998, p. 380), the “places where physical activity may occur” (Sallis,
Cervero et al., 2006, p. 302), and, “[P]laces associated with particular behavioral
and organizational programs that are directly (objectively) observable and recur at
regular, specified intervals” (Clitheroe, Stokols et al., 1998, p. 104).

100

Figure 6. Active Living Eecological Model (Sallis, Cervero et al., 2006)
101

I suggest that the diversity of meanings assigned to behavior settings in
AL research diffuses the concept in a way that reduces its usefulness. As an
example, note that the latter definition by Clitheroe, Stokols et al. includes the
element of time, whereas in the former definition offered by Sallis, Bauman et al.
and Sallis, Cervero et al. the behavior setting is described in static terms.
Active Living researchers tend to rely upon the ecological model for its
ability to describe “trans-disciplinary” environmental influences (King, Stokols et
al., 2002; King, Bauman et al., 2002; Sallis, Linton et al., 2005). Also, the
behavior setting is relied upon to describe a “standing pattern of behavior…with
specific temporal-spatial coordinates” (Schoggen, Barker et al., 1989, p. 31).
While the behavior setting does both of those things, it is also an ecological model
of the human-environment relationship, not only trans-disciplinary but also
reciprocal, a concept largely overlooked in AL Research. This refers to the
ecological model’s closing of the traditional gap between subject and object,
viewing person and environment as a single reciprocal unit. That is a far cry from
AL descriptions of behavior settings as, for example, “relationships between
particular physical environment attributes and physical activity behaviors”
(Humpel, Owen et al., 2002, p. 189).
The concept of the behavior setting was a product of some 25 years of
research by the prodigious ecological psychology researcher Roger Barker.
Barker defined the concept in a very specific way: “[A] standing pattern of
102

behavior and a part of the milieu which are synomorphic and in which the milieu
is circumjacent to the behavior” (Barker & Wright, 1971, p. 45). Synomorphy
here refers to the similarity in structure between a place and some human activity
that takes place there (p. 46). Milieu refers to the community, including not only
people but also buildings and streets, schedules, and agreements between persons
(ibid.). The circumjacency of milieu and behavior refers to standing patterns of
behavior (p. 47). Because the patterns are ongoing they are predictable, which
was one of the great strengths of Barker’s research. Note also that in the Active
Living definitions of behavior setting, it is the environment that is the subject of
the definition; in Barker’s definition it is a pattern of behavior and the
environment that are the subject, that is, a reciprocal human-environment
relationship.
The AL ecological research model is quite different from the traditional
correspondence model in many ways, mainly for its ability to consider variables
at multiple levels of interaction (Figure 6). This is the “Four Domains” ecological
model of Active Living, in which the individual is conceived as interacting with
the environment at multiple levels. Behavior settings are one of those levels. The
different levels and their interactions “illustrate the roles numerous disciplines can
play in research on active living,” and suggest various combinations of variables
that primary researchers might use for investigating determinants of physical
activity (Sallis, Cervero et al., 2006, p. 300). Perception is in the domain
103

immediately surrounding the individual, and through perception the three outer
domains are known, including the behavior setting one. For example, “household
activities” might be viewed not only as a sociological behavior influenced by
variables such as the presence of “labor saving devices,” but also in relation to
policy research-oriented zoning codes, for example.
18

The model offers “community organizations” as an example of a behavior
setting. Here the behavior setting is surrounded by an outer layer of influences at
the policy level and contains an inner layer at the “perceived environment” level.
In this model, a community fishing club could be conceived as a locus of a variety
of levels, for instance perceptions of safety, attractiveness, and access to local
fishing facilities, as well as the larger park policy that stocks the ponds with fish
and provides upkeep and security. Thinking in this transdisciplinary way,
combinations of variables could be chosen that reflect all of the appropriate levels
necessary to provide a rich determinant of physical activity behavior. The
community fishing club is not merely an instance of a club, but a multi-leveled
place “in which certain kinds of activities may occur” (p. 302).
In Barker’s model of behavior settings, there are no separate layers of
human activity and environment, but rather they are part of all the domains in
which a person in involved. The person and the environment are conterminous.
Barker employed two criteria for describing behavior settings that fall outside the

18
Cf. M. Powell Lawton’s concept of “environmental press.”
104

Active Living description of ecology. First, there is what Heft, Lombardo and
others call a reciprocal relationship between individual and environment (Heft,
2001, p. 7; Lombardo, 1987). Second, individuals within a behavior setting are
transposable, in the sense that if other persons replace any or all members of the
setting, the activity in the setting will continue unchanged. With this, Barker
implies that behavior settings are extra-individual (Schoggen, Barker et al., 1989,
pp. 13-14, 31, 202, 254, 310). There are other criteria, but for the moment we will
focus on these two.
19

Consider a child entering an outdoor basketball court, hoping to join a
game. He enters the space where a game is being played, scans the layout and
activity and the other people there, and either takes a seat on a bench or leans
against the wall. Players and people on the sidelines cast a glance at him. His
taking up a waiting posture signals that he is interested and available to play, and
everyone there knows it. The boy is not only aware of the physical structure of the
court when he enters, but also of the patterns of behavior in it, as well as of his
effect on those patterns of behavior, in the sense that his being there means that
there is one additional player on hand, and also one less opportunity for another
person to enter. The meaning of the place and its events are reciprocal in that the
players and the physical and social setting are interacting to create the meaning. In

19
Barker’s behavior settings represent “bounded” patterns of behavior, where there is often an
“interdependence” between settings, and the settings are “extra-individual,” to name but three more
(Schoggen, Barker et al., 1989, pp. 30-31).
105

this instance, we would not describe the behavior setting as “A Basketball Court,”
but rather “Joining the Basketball Game,” that is, an activity occurring according
to a synomorphic relationship with the physical place, according to certain
patterns of behavior, and according to an exchange of influences between persons
and between persons and the environment. The behavior setting “Joining the
Basketball Game” describes the perspective of a transposable position or role in
the scene, into which many sorts of people could step. The behavior setting
remains the same regardless of who is taking up its roles.
Notice also that in this case the boy’s actions are in response to his
discriminating amongst information about the environment that is already laden
with meaning when it appears to the boy’s senses. The boy is not merely reacting
to stimuli, but discriminates between what he is allowed to do as well as what he
allows himself to do, as well as what anyone in the situation can do.
Barker argues that anyone in the situation is highly likely to feel the same
pressures, and to behave within the patterns and constraints set down by the
dynamics of the behavior setting. This means that participants in the setting are
transposable in the sense that anyone in the setting is likely to behave in a
predictable manner (Schoggen, Barker et al., 1989, p. 91).
Yet this is not meant to suggest that environments make people into
automatons. On the contrary, the transposability of individuals is meant to throw
focus onto the constraints and possibilities within which the individual might
106

maneuver. The boy’s entrance into the setting is itself a factor in his eventual
decision, and his decision affects the decisions of the other players, and his arrival
affects how anyone else there will participate that day. Barker takes all these to be
perceivable by the individual. In the sense that the boy’s presence in the behavior
setting affects the behavior setting itself, we can say that the behavior is
‘reciprocal’ (Lombardo 1987).
One value of transposability it that it allows for stable observations of
what happens in a particular human-environment interaction. It allows for a study
of necessary interactions within an environment, as it is experience by anyone
entering it, rather than a one-way relationship in which the environment
stimulates a psychological response. Barker’s behavior setting conceives of
setting characteristics that “…persist even when current inhabitants of the setting
are replaced by others” (Schoggen, Barker et al., 1989, p. 31). Also, the extra-
individual focus allows for an overall view of the environment’s structure as a
social process as well as a physical one, because it allows for observation of “how
structure is shared” between individuals. When two individuals share the
environment, their actions affect each other reciprocally. Perhaps most
importantly, Barker discovered that the transposability of individuals within the
behavior setting provides a stable framework for predicting behaviors at certain
times and places, and with rather surprising accuracy (Schoggen, Barker et al.,
1989, p. 202).
107

I suggest two reasons why the predictability of behavior settings has not
been exploited further in AL research, one technical, and the other theoretical.
First, the amount of data and analysis required to fully describe a behavior setting
is extensive. Even those few researchers who have designed an abbreviated
behavior setting survey acknowledge that to remain true to Barker’s method
requires time- and data-intensive research.
20
The second reason why behavior
settings surveys are not exploited for their predictive ability is that they are
theoretically misunderstood, I suggest, as making predictions based on numerical
frequency counts of behaviors by people in particular places, a more or less
standard quantitative approach. In such cases the unit of observation is the person,
and the behavior setting is a context, as opposed to the behavior setting being a
unit of observation in itself, including persons, behavior, and context. Schoggen
found that the predictive power of the behavior setting is in the assumptions and
informal ‘rules’ that persons bring with them from behavior setting to behavior
setting. Standard quantitative approaches based on correspondence models as
described above make predictions based on “…empirical correlations between
counts of the number of inhabitants within behavior settings and measures of their
behavior” (p. 202), or else based on “psychological characteristics of the
inhabitants” of the behavior setting (ibid.). Rather, Barker argues, the person is
only a part of what makes up a behavior setting, and so he proposes that the

20
Bechtel and Wicker offer the most notable examples of abbreviated behavior setting surveys that remain
faithful to Barker’s original method (Bechtel, 1977; Wicker 1979).
108

behavior setting is itself the unit of observation. What occurs in the behavior
setting is highly predictable (Joining the Basketball Game will always occur more
or less in the same way), and inasmuch as the behavior setting is stable one can
predict when people will enter it and what they will do there, where they will sit
or lean, even where they will set their gear down, or how they will leave the court,
but what is predicted is not an alteration in the individuals’ psychology, since the
behavior setting will operate irregardless of the cognition and affect of the persons

Figure 7. Barker’s “eco-behavioral circuit” model (Barker 1968).
109

who are part of it. For example, the boys on the basketball court may have high
self-esteem or haphazardly. The Barkerian ecological model documents
environmental perception as a human-environment “circuit,” able to observe
larger social and even institutional determinants of individual behavior, and how
the power that individuals have in a setting is in large part measured by their
ability to influence the rules governing the setting (pp. 202, 367-68) (Figure 7).
Barker observed behavior settings in this way throughout his monumental 25-year
“Midwest” study, in part because he was able to document the “programs” or
“operating instructions” that are present in every setting (Barker & Wright, 1971,
p. 31). Again, though, that is not to suggest that people cannot be autonomous
from those instructions, but rather that they depart from standing patterns for
individual reasons.
Many papers cited above speculate on what sorts of combinations of
variables an ecological determinant might contain, but none have taken up the
question of what an ecological audit would look like. That is to say, ecological
determinants have yet to be theoretically associated with the distributions of the
good that are a necessary part of audits. As I mentioned, in Active Living research
the conception of the ecological human-environment relationship mostly
emphasizes its trans-disciplinary dimension, and this is clear in the ecological
“Four Domains” model of environmental perception used in Active Living
research (Sallis, Cervero et al., 2006). Given the limitations of the correspondence
110

model, I would say that the ecological model is a great advance for the field, if
only because of its trans-disciplinary view of the determinants of physical
activity. However, it must also be pointed out that in overlooking transposability
and reciprocity of human and environment, the AL ecological model retains
something of the correspondence model’s conception of the individual as a vessel
for equally distributed feelings. Fleishaker’s description is useful here, of how the
properties of some external physical object are assumed to generate an internal,
causal representation in the mind of the perceiver (Fleischaker, 1984, p. 41).
21
The
Four Domains model uses the concept of behavior settings to describe influences
on physical activity as moving in one direction only, either from the outside layers
toward the individual at the center, or in the other direction, but not in both at
once, reciprocally. AL research tends to conceive of the environment as acting on
the individual, stimulating physical activity, with the individual as a kind of
passenger in the center of the model.
An important point about Barker’s “standing patterns” of behavior is
striking when thinking about AL research designs: Such patterns are not
generalizable among behavior settings, Barker discovered. The ability to predict
behavior within one behavior setting does not necessarily allow for generalizing
that prediction to all similar settings. Making such generalizable observations
would require imposing an artificial set of analytic units onto settings, which is

21
Fleischaker referred to the subjectivist model as “traditional,” which is the term that I will employ
henceforward.
111

quite a different thing from the “self-generated” units that define behavior
settings, which Schoggen argues are “natural units in no way imposed or created
by an investigator” (Schoggen, Barker et al., 1989, pp. 14-15). Here once again
Barker’s behavior settings depart from the current AL ecological model, which
imposes ‘artificial’ units of observation.
In the most general terms, the point of the ecological model as Barker
describes it, and as Gibson will, further on, is that,
It is clear that structure cannot be discovered by observing a single part,
such as the point of intersection of the environment with a particular
person, or by considering the parts separately, one by one…By these
methods, the structure of the context is dismantled and rearranged; the
structure is destroyed. (Schoggen, Barker et al., 1989, p. 12)

None of this is to dismiss the Four Domains model, but merely to suggest
that the scope of data gathered by the model may not be as rich as it could be if
reciprocality were introduced.
How to introduce the reciprocity of perceiver and environment into AL
research is a challenging question. As Barker created it, the behavior setting
method is quite unwieldy, and would provide cumbersome service in this regard.
Researchers such as Bechtel have abbreviated the behavior setting survey so that
it requires less labor-intensive fieldwork and analysis, and I suggest that these
methods could supply data for built environment-physical activity research that
cannot be observed otherwise. However, there is a more critical issue with the
behavior setting as a metric for built environment-physical activity research: The
112

behavior setting survey is a positive description only. While design researchers
such as Bechtel have successfully used Barker’s methods for planning the layouts
and patterns of environments, the behavior setting method does not produce the
sort of precise normative targeting of groups of individuals that is required in the
childhood obesity epidemic. Further on, I will argue for a different, though
related, unit of observation, Gibson’s “affordance.” I suggest that affordance may
be uniquely suited to built environment audits for Active Living research, because
it describes a ratio of individual and environment, and so is uniquely conceived to
extend the opportunity-based approach to include the individual’s agency, or
actual ability to make use of resources. I will argue that affordance is a metrical
embodiment of the third kind of audit information described in the general audit
model, opportunity-plus-agency, or what Sen calls “capability.”
An Ecological Model Would Recognize Capability
We established earlier that the most common evaluation of the collective
good is the utilitarian one, which appears in AL research and audits as
determinants of “satisfaction.” The Rawlsian critique of utilitarianism
subsequently established an alternative kind of determinant, based on Rawlsian
general resources, which I have termed “opportunity.” It has also been argued that
AL researchers frame the human-environment relationship with one of two
conceptual models, either the correspondence model, or the ecological model.
Satisfactions are ideally observed with the correspondence model, while
113

opportunities are somewhat more observable with the popular AL ecological
model. Here I will consider how capability information fits within the ecological
model of the human-environment relationship, starting with a brief description of
what capability is, and how it contrasts with Rawlsian general resources. I will
consider a third normative-evaluative tradition of urban design, the
“synomorphic” tradition, and argue that because capability data is a ratio of the
individual and the environment it is uniquely appropriate to the AL ecological
model. I ultimately argue that Gibson’s “affordances” provide a way to observe
capability in the built environment.
There is no question that the childhood obesity epidemic in the US poses a
serious health threat to inner city children of low-income families (AOA 2007;
Ebbeling, Pawlak et al., 2002, p. 480; Kumanyika, Jeffery et al., 2002; Rocchini,
2002). Nor is there doubt that researchers have now joined across disciplines to
assess how qualities of the built environment impact children's physical activity
(Active Living Research, 2006; Ainsworth, 2002; Chawla & Heft, 2002; Ellen,
Dillman et al., 2001; Flournoy & Yen, 2004; Miles, 2006; Transportation
Research Board, 2005). Research clearly shows that children's perceptions of the
environment are a factor in establishing built environment determinants of
physical activity (Prezza, Alparone et al. 2005; Ritenbaugh, Kumanyika et al.,
1999; Zabinski, Saelens et al., 2003). Literature on functional perceptions of the
environment suggest that children bring particular competencies to the
114

understanding of place that point to descriptions of the child-environment
relationship that are as complex as an adult’s (Derr, 2002; Hart, 1979; Korpela,
Kytta et al., 2002; Piko & Keresztes, 2006). Children also have social constraints
on decision-making about exercise that suggest that they must be understood
ecologically (Bauman, Silver et al., 2006; Bell, 2002; Borders, Rohrer et al.,
2006; Evans, Lercher et al., 2002). The new audit instruments for the built
environment that have been brought to bear in these assessments are all
influenced to some degree by an ecological understanding of the human-
environment relationship. However, if too narrow a view of the ecological model
is taken, research runs the risk of excluding important constraints and mediating
variables from physical activity determinants. The ecological model shows
promise in this regard, although a richer kind of information is required than the
general resources approach alone can deliver. Now we will consider what the
capability approach might have to offer.
As Amartya Sen describes it, the capability approach has its roots in both
the ancient Sanskrit Upanishads and also in Aristotle (Sen, 2000, pp. 13-14). Sen
draws on these two sources for the fundamental idea that goods are not
accumulated for themselves, but rather as means to an end. In the 1970s Sen
began to challenge the traditional measurements of social welfare, inasmuch as
they were based on a “behavioral theory of economics” that views increased
income as the aim of development (Sen, 1980). This was the utilitarian approach
115

that equates increased income with preference fulfillment. The utilitarian
approach, Sen argued, is unable to describe the poor’s unequal ability to convert
income into well-being, according to their heterogeneous abilities and
circumstances (Sen, 2000, pp. 54-86). A person may be judged to have an
equality of utility in terms of income, but remain impoverished in other categories
of living, such as political liberties or education, Sen argued. His innovation was
to argue that the good is neither an equality of utility nor an equality of
opportunity, but instead an equality of what he calls “capability,” which is
opportunity plus the agency to convert opportunities to well-being. Capability
represents the sum of a person’s potential states of well-being, including physical
well-being, mental well-being, etc. These potential states are metrically
recognized in terms of a person’s ability to make use of some resource in the
pursuit of a desired end-state such as physical well-being, given their personal and
situational constraints. Sen sometimes refers to capabilities as “freedoms” to
emphasize that they mean an increase of choices and the abilities to pursue them
(pp. xii-xiii).
The capability approach recognizes how an impoverishment in one
category of living can increase impoverishment in another, but conversely, also
how increased well-being in one category can increase well-being in another, in
the sense that freedom from physical assault may increase the ability to participate
in the political decisions that affect one’s life, or in the sense that the inability to
116

travel safely to community meetings may reduce the public participation of the
elderly. Freedoms are “interconnected,” and thus an evaluation of what a person
can do or be requires a complex understanding of those interconnections (Sen,
2000, p. 4). These interconnections between freedoms introduce the idea that the
built environment’s role in well-being is multidimensional and interconnected
with states of being like health in ways that require an ecological understanding of
how a person’s characteristics and circumstances constrain her ability to make use
of the opportunities available to her. In terms of modeling the human-environment
relationship in capability terms, we must see it as not only a set of the resources
available to a person, but also how that person’s particular identity constrains––or
enables––the ability to make use of those resources.
Sen argues that even the equal distribution of Rawlsian general resources
cannot address varying abilities and disabilities of diverse persons to convert
goods to well-being (Sen, 2000, p. 72-74). In any heterogeneous community there
will be individuals who will convert resources at different "rates." Even a perfect
equality of primary goods could not compensate for variations in social climate,
personal psychological or physical characteristics, how goods are culturally
distributed within a given family, or diversities of ‘environmental circumstance’
(Dixey, 1998).
Sen took fundamental exception to Rawls’s Theory of Justice in its
emphasis on the good as an equality of means. Sen critiques Rawls’s assumption
117

that an equal provision of resources necessarily means that each individual can
equally convert those means to well-being (Sen, 1980). He argued that a new
evaluative framework for equality was needed, that could measure the ability to
convert resources into desired states of well-being. It is in the spirit of this
argument that Sen asked his question, “Equality of What?” for which the general
audit model in Chapter One offers three built environment answers: satisfactions,
opportunities, and affordances (Sen, 1977).
Challenging Rawls’s argument for an equality of primary goods, or
“commodities,” Sen argued for an equality of the “freedoms generated by
commodities, rather than the commodities seen on their own” (Sen, 2000, p.
74).
22,23
Sen argues that the Rawlsian Theory of Justice cannot account for how
heterogeneous individuals convert their share of primary goods resources into
desired end-states––in the sense that a person who cannot ride a bike cannot
convert the gift of a free bike into mobility (Sen, 2000, pp. 62, 67). While Sen
recognized the logic of Rawls’ switching the evaluative emphasis from equality of
utility to equality of resources, he went on to point out that individuals convert
such resources at different rates, which means that while measuring the provision
of Rawls’ equality of resources is an essential aspect of evaluating development,

22
Rawls refers to “primary goods” and Dworkin to “equal resources,” both of which Sen labels as
“commodities.” We will refer to them all as “resources” to indicate their function in a capability-based
evaluation of the built environment.

23
The remainder of this chapter will be devoted to considering how AL research and built environment audits
might go about observing determinants of capability/affordance.
118

it stops short of considering the rates at which individuals convert resources to
well-being. Rather than focusing on either an average utility or an equality of
resources, Sen says instead that conversion to well-being ought to be the focus,
measured in the form of actual opportunities, or “what people are actually able to
do and to be” (Nussbaum, 2000, p. 5). A general resource of, say, the opportunity
for property ownership, could be distributed equally across a community, but if
the potential property owner is constrained from getting a job or a loan, then one
cannot say that the person has been able to convert the right of ownership into
material well-being, irregardless of preference. Sen would say that the person had
a choice, but not a real choice.
Sen’s capability approach has come to “play a major role in political
philosophy and normative economics” (Pogge, 2002, p. 1). Yet one critique of the
approach is that Sen has actually offered no real alternative to the dominant forms
of evaluation of collective good, but has simply switched a utilitarian metric from
constraints on income to constraints on other sorts of goods (p. 3 f). Rawlsians
have not yet fully defended their position, but capability theorists have not either
(p. 27). Considering the capability approach in terms of the built environment, as I
do here, cannot resolve those fundamental arguments. Rather the hope here is to
bring the contributions of the built environment to well-being into those
fundamental arguments, to frame the built environment as more than mere context
for human well-being, but rather as an integral part of it.
119

Capability Best Revealed by Affordance
Capability is innately ecological: It describes the interconnections between
variables in a trans-disciplinary way, and it requires a reciprocal understanding of
the built environment-physical activity relationship. However, in order to
operationalize capability in the built environment, it must be framed as a
synomorphy that includes the environment, the individual, and the constraints on
the individual’s physical activity. I propose that the concept of “affordance” be
used to describe capability in the built environment. Gibson’s affordance and
Barker’s behavior settings are distinct concepts in terms of how they are
operationalized, but they also share an ecological conception of the world, and a
pragmatist lineage.
Ecological psychologists J.J. Gibson and Roger Barker were
contemporaries, both American, and shared colleagues and intellectual influences,
but ironically they never met. Although Gibson's and Barker's theories are very
similar, and the sources of their inspiration both traceable to William James's
work on perception,
24
they never worked together, and referenced each other only
a little (Heft, 2001, pp. 273-325). Their students have noted the uncanny parallels
in the research of these two perception theorists. Heft argues that Gibson’s
affordance describes what Barker's unit of analysis, the behavior setting, “affords”
the individual, and that is the view taken here. Heft argues that behavior settings

24
James more directly influenced Gibson, but as Heft argues, Gibson and Barker share the influence of Fritz
Heider, who is a strong link to James.
120

may contain affordances, but affordances do not have to be in behavior settings
(p. 296).
Heft speaks of the affordance metric as generally more deft and precise in
its descriptions than the behavior setting:
The description of behavior-milieu synomorphs can be made more precise
in light of affordances…In order to provide something other than a very
global description of supportive milieu features, shared characteristics of
the participants in the setting could be considered, and the affordance
approach can help in this regard. (Heft, 2001, p. 294)

Like Barker’s behavior settings, Gibson’s affordances are perceptions of
what the environment will support, but as perceived "outside the skin" of the
individual, rather than in the mind alone (Schoggen, Barker et al., 1989, p. 8).
Like Barker, Gibson argued that the ecological model of perception described
human discernment of the relationship between themselves and their proximal
surroundings (Gibson, 1966 pp. 31-46) An affordance is a confirmation of an
investigation of the environment by individuals, and individuals of the same type
are likely to identify the same affordance. Gibson and Barker both conceived of
the ecological model in terms of “how structure is shared” between individuals in
the environment (p. 280).
Like the behavior setting, affordance is extra-individual, in that
individuals are transposable within it, and the synomorphy between activities and
the physical space are essential to both measures.
121

Gibson’s ecological model coincides with Barker’s in that they both stand
in contrast to the correspondence model. In Gibson’s and Barker’s ecological
models individuals do not respond “to the things,” as Gibson says, until they are
in a relationship with the already existing meanings of those things (p. 282).
Those things include situations, objects, and the behavior of others.
Gibsonian affordances have been studied mainly in classical laboratory
controlled experiments to date, and that research has amply demonstrated the
observability of affordances (Warren, 1984, for instance, or Zhang, 2000) and
also the falsifiability of the overall concept of affordance (Stoffregen, Gorday et
al., 1999). In Gibson’s ecological model the meaning of a perception is
discovered, in the sense that it is assumed that the individual is in some sense
active in the environment, and not merely a passive receiver of information.
Incoming information is “not only [passive] stimulation from the environment but
also stimulation from the attentive actions and reactions of the observer…”
(Gibson, 1966, p. 33). This means that the individual gathers information not only
about things and other people in the environment, but also information about what
happens because the individual is it.
Animal and environment are a reciprocity, but they are reciprocal because
they are dynamic in form. Perceiving is an ecological event that does not
take place [only] within the animal––it is an ecological process involving
both perceiver and environment. (Lombardo, 1987, p. 18)

122

In the Gibson model the pedestrian, for instance, would be conceived as
essentially exploring the environment, and “walkability” would be a confirmation
or refutation of what is sought. In the ecological model “information become[s] a
matter for investigation” rather than one of passive reception (Gibson, 1966, p.
266). More, the “learning of new meaning is an education of attention rather than
an accrual of association” (p. 320).
Correspondence-based perception, by contrast, proceeds in a linear
process of input-evaluation-response in which the person is struck by stimuli to
which she must react. In contrast, Gibson’s ecological perception is a circuit,
where the perception of walkability is as much a confirmation of the individual’s
investigation of the environment for what it has to offer in the way of movement,
transport, diversion, exercise, as it is a set of stimuli for movement.
25
The
correspondence model of perception, Gibson argued, "…has been conceived as a
process of 'enrichment,' whereas it might be better conceived as one of
'differentiation'" (p. 269). Rather than being built up as an accretion of stimuli, the
environment's meaning in relation to the individual’s previous experience is an
assessment of constancies, or invariants. Objects, time, and space, are all
apprehended through the person’s investigation of the environment, in a process
of discerning what the environment affords the perceiver in terms of use.

25
All the Gibson scholars cited in this study trace Gibson’s debt to Dewey’s 1896 “Reflex Arc” argument.
Dewey critiqued the then-traditional “reflex arc” stimulus-response model of experience as “disjointed” (p. 4)
in that it “leaves us nothing but a series of jerks” (p. 3). He proposed a circuit model in which perception of
the world is viewed as a “continuously ordered sequence of acts” in which contact with the phenomena are
confirmations or refutations of something that is under investigation by the mind (p. 4).
123

Affordances are "meaning... revealed in the environment" rather than in the mind
of the perceiver (Lombardo, 1987, p. 308).
The notion of discerning invariants may echo with those readers who are
familiar with the Wertheimer branch of the Gestalt movement, of which Kurt
Lewin was a part. Lewin trained Barker, and was an influence on Gibson’s
thinking.
There is no evidence that Kevin Lynch was specifically aware of Lewin or
the principles of Gestalt, yet the influence seems apparent in his work as early as
1968s “Where Learning Happens,” co-written with Stephen Carr (Lynch & Carr,
2002). The individual’s total engagement with the environment is made normative
in that paper, in which we hear early statements of Lynch’s theme of urban
development as a support for individual development. That idea would go on to
define Lynch’s Good City Form in 1981. In “Where Learning Happens,”
continual growth and exploration are integral to the individual’s pursuit of an
“integrated life,” and Lynch and Carr argue that the city can normatively support
that sort of ongoing person-environment engagement of the person. Lynch and
Carr argued that the person ought to be conceived as integrated with the
environment, and that because that inter-relationship is overlooked in urban
design policy, the individual is only partly engaged with the city, and with
society, and with himself. They argued that policies could be made to develop the
city as “a purposefully designed ‘school,’ a place for learning and growing
124

throughout life” (p. 429). This is a vision of the city as affording individual
development.
More than an exclusive product of the mind, affordances represent the
individual and the environment as a single, dependent, and mutually defining
relationship, expressible as a ratio (Warren, 1984). Affordance cannot be
perceived only as a quality of things, but always in relation to the perceiver, in the
sense that, “The glass is only fragile if there are possible circumstances in which
it might shatter” (Chemero, Klein et al., 2003, p. 4). Chemero suggests that we
might think of affordances as “dispositional properties of the environment” that
are, “…only conceivable when paired with actualizing circumstances…in which
the disposition becomes manifest” (ibid.).
Affordance Expresses a Proportional Relationship
Warren found in the laboratory that the choices individuals make are
imminently predictable when viewed in terms of affordance (Warren, 1984, p.
686). He published findings from an affordance study in 1984 suggesting that
individuals bring a “preferred behavior” to their engagement with the
environment, and that they commit to that behavior at the point when they
perceive that the environment affords their purposive activity. Much AL research
to date is in contrast tp this view of Warren’s, in that the AL models have largely
assumed that the individual must be somehow provoked into physical activity,
prodded by a mental trigger in the environment. The new AL ecological audit
125

model may promote that same kind of conception if an expanded view of the
ecological relationship cannot be found. I argue that the individual’s pursuit of
well-being in the built environment may be seen as a purposive activity, and that
given a “critical point” of affordance, individuals will make certain predictable
choices.
Like affordances, capabilities offer a ratio-based unit of observation for
the built environment-human relationship. As in Warren’s affordance findings,
individual capability means that there is an optimal point in the relationship
between person and environment at which resources are best converted to well-
being, a “critical point” at which people will make certain choices. In terms of the
obesity epidemic and active living, I am arguing that children do not need to be
provoked into active living, but rather need only be given the real opportunity to
pursue it, and they will likely do so.
Warren points out that it was Gibson who first suggested that to formally
express an affordance, environmental properties “have to be measured relative to
the animal” (Warren, 1984, p. 686). It was Warren’s “leg-length study” that first
tested the reliability of the affordance ratio in a laboratory setting. An affordance
as a unit of observation represents neither the individual nor the environment, but
the relationship between them: Warren showed how the numerator side of the
ratio can represent some standard of the physical environment––the height of a
stair riser, for example––and the denominator side of the ratio is some property of
126

the individual––leg length, eye height, or age, for example. The property of the
person (Warren names it A) “may be taken as a standard for measuring an
environmental property E” (ibid.). Thus affordance is expressed as a ratio:
π = E/A
26

In laboratory studies Warren found that given a variety of stair heights to
which their leg-lengths are variously proportionate (too short, too long, just right,
etc.), study participants assessed particular ratio “critical points” at which the
energy expended and the riser/leg ratio were optimized, and at that moment they
climbed the stair.
27
Warren’s study showed that such critical points are quite real,
and suggests that the mind is searching for them all the time, in the sort of
reciprocal process that I argue ought to be considered in AL ecological models.
To say that a purposive activity such as climbing stairs is “afforded” by the
environment is to say that a would-be stair climber is looking for the critical
point, and if she finds it she is likely to make use of it.
Inasmuch as Warren has shown that critical point ratios of affordance
between stair and climber can be established in advance, he argues that it is
possible to reliably predict what choice the climber will make.
28
This harks back

26
Note that the converse, π = A/E, can describe the personal characteristics and social circumstances that
constrain the individual’s engagement with the environment. This will be useful as a guide in developing the
human factors of the affordance measure, and could be used to inform a public participation process.

27
Warren’s test participants made such decisions more or less subconsciously. They were not discussed
between participants and researchers, and the participants reacted to critical points seemingly without
realizing it.

127

to Barker’s description of the predictability that behavior settings lend, although
in the case of affordance the identity of the individual is the point of the measure,
whereas in Barker identity is somewhat subsumed in the activity of the group. In
any case, I suggest that the kinds of predictability that Barker and Warren
describe is an entirely different kind of predictability than that employed in most
AL research today.
Well-being as Purposive Activity
Defining purposive activity as well-being makes it possible to use
affordance measures to describe how well individuals are able to convert certain
resources to states of well-being. In an affordance-oriented critical moment, the
individual is attempting to engage in some purposive activity that she is more or
less able to achieve according to what is afforded. “Purposive” does not
necessarily imply a conscious purpose on the part of the individual, as in a
scheme or plan, but is more to indicate the consequentialist nature of affordance.
An affordance teleologically enables some desired state of well-being. In the
same fashion, capability is consequentialist in that ‘having capability’ means that
an individual has the real opportunity to be or do something that they desire to be
or do.
29
By defining the individual’s purposive action as well-being, then, we can

28
Warren’s “properties” of people are what I have been referring to throughout this study as “characteristics
and circumstances.”

29
The consequentialism of the capability approach differs significantly from the consequentialism of the
utilitarian approach, however, in part because capability represents a state of being or doing, whereas
utilitarianism is concerned with a feeling of satisfaction.
128

frame affordance in normative terms, opening up the human development
dimension of affordance.
Capability as a Ratio
Sen’s capability is expressible as a ratio in the same terms as affordance.
On the numerator side is some resource of the built environment that the
individual must use to attain well-being, for example proximities to grocery
stores, or the distribution of green space, or lack of exposure to water pollutants.
On the denominator side is some identifying characteristic or circumstance of the
individual––poverty or health, for instance, or the state of being homeless after a
natural disaster. Just as in affordance, capability describes some standard
characteristic (and/or circumstance) of the individual in proportion to something
in the environment that must be used to fulfill a purposive activity.
In the same manner that affordance can be used as a measure of capability
in the built environment, Sen’s concept of “conversion” can describe the human
development measure of affordance. Sen tells us that we cannot evaluate a
person’s well-being solely according to whether they have enough resources,
without also taking into consideration whether the person can also convert––make
actual use of––those resources in the pursuit of well-being. Sen gives a
hypothetical example of the inability-to-convert, describing a person with
dysentery who receives food aid in the form of rice: Because the person’s
gastrointestinal illness prevents her from absorbing the nutrients in rice––
129

converting them to health—the distribution of rice alone cannot be the measure of
her well-being.
In terms of a child’s built environment, the presence of an empty parking
lot might be a determinant of "skateboarding" but to certain children it might not
be useful in that way at all, perhaps because it is dominated by social groups that
exclude certain children because of age or other factors (Woolley & Johns, 2001).
In order for an auditor to say that the neighborhood “affords” skating by an
individual child, the audit must observe not only the child’s characteristics, but
also the relationship to the physical parking lot in social terms, how the lot is used
by others, what patterns of activity make it more of less useable to the 11-year-
old. An assessment that does not take the individual into account in this way
would have to either assess ‘skateability’ as an average across the population of
skaters, or else as a certain level of skateable surface resources that have been
deemed a fair amount per skater. Affordance is always of something, for someone.
For instance, we might describe an environmental affordance as “Affording after-
school park use by children ages 6-11 who are raised by their grandparents.”
30
To
develop the audit in this case, primary research would need to establish who is
being afforded (“children 8-11”), as well as the constraints on their park use
caused by having a grandparent as a guardian (“inability to negotiate public
transportation,” “fragility of grandparent,” etc.), as well as the relevant qualities

30
A growing demographic in poor neighborhoods, research shows (Hayslip & Kaminski, 2005; Musil,
Warner et al., 2006).
130

of the environment that support the child’s park play (“elderly-assist bus stops at
park,” “handrails at park playground,” “sufficient benches in eyesight of
playground,” etc.) In Chapter 3 I will offer examples of how affordance is
observed in more specific terms.
I suggest that affordance is a more suitable measure for Active Living
audits than behavior settings, for three reasons. The first is that affordances,
unlike behavior settings, are about individuals. In the affordance metric the
particulars of the individual’s characteristics are integral, but in behavior settings
they are optional. Behavior settings may have diverse kinds of persons in them, or
not. Affordances are about groups of people whose identity makes them relative
to the environment. In the behavior setting, people are part of a behavior-milieu
synomorph and therefore have certain behaviors that identify them. Behavior
settings tell us that a 10-year-old child is likely not to press the button to activate
the blinking crosswalk, and will simply cross over. Affordance will tell us that the
button is too high for the 10-year-old to reach. The behavior setting is based on
long observation of the behavior-milieu synomorph. Affordance has as its premise
the child’s particular relationship to the setting. Again, the lines between the two
are fuzzy, and indeed there is crossover since behavior settings contain
affordances of something. But with behavior settings it is more difficult for the
auditor to specify what she is looking for; the relevant characteristics of the
people involved cannot be known until after long observation.
131

In behavior settings the description of the individual and the milieu are not
hypothesized, but discovered. Barker is quite clear about this point throughout his
writings. On the other hand, affordance has a kind of normative instrumentality
that supports hypothesizing and experiment, which suggests that it could serve in
built environment-physical activity research. Environmental psychologists have
shown that in the moment of deciding to take some purposive action, the person
knows what an affordance needs to be in order to support some activity, say stair
climbing. Laboratory tests of affordance show that the person is not only
receiving inputs from the environment, but is actively gathering information on
the environment and his relationship to it (Chemero, Klein et al., 2003;
Stoffregen, Gorday et al., 1999; Warren, 1984). There is a useful normative
dimension in affordance that I suggest makes it useful in normative-evaluative
evaluations of the built environment. That is the second reason that I recommend
affordance to describe capability in the built environment.
The third reason to recommend affordance is mentioned in the previous
paragraph, regarding the laboratory experiments: Gibson’s ecological model
views perception of the environment as a two-way, reciprocal event. Gibson
argues for two kinds of information about the environment that the individual
receives, which he dubs “imposed” and “obtained” information. Imposed
information comes toward the individual as input, and grabs the attention, like the
sound of a car wreck out in the street, or a forgotten kettle suddenly whistling in
132

the other room. Obtained information is that which the person seeks, gathers, and
investigates.
Neurophysiologists have known for decades that the idea of humans
having five separate senses is a myth. Indeed, it is difficult to say that the senses
are distinct, since hearing is involved in maintaining equilibrium, the eye
coordinates where to focus the hearing, and touch and smell are almost
indistinguishable. Yet, the old models persist, and it was these that Gibson was
challenging with his affordance concept. The process of perception, Gibson
argued, "…has been conceived as a process of 'enrichment,' whereas it might be
better conceived as one of 'differentiation'" (Gibson, 1966, p. 269). Rather than
being built up as an accretion of stimuli, the object's meaning in relation to the
individual’s previous experience is a matter of the successive order of
constancies. Affordances are "meaning... revealed in the environment"
(Lombardo, 1987, p. 308). Affordances offer a metric for interpreting those
meanings in a precise way that could make a unique contribution to built
environment-physical activity research.
Synomorphy Tradition
The third and final normative-evaluative urban design tradition that we
will consider is what I am calling the synomorphic tradition. I propose that if it is
possible to evaluate the built environment in affordance terms, synomorphy will
be a key concept, since synomorphic determinants of physical activity require
133

understanding human engagement with the environment as a relationship, in ratio
terms. The synomorphic tradition is concerned with human-environment
relationships as necessary and integrated, in the sense that there are particular
human activities that are strongly associated and even inseparable from particular
built forms. In synomorphic design terms, the environment and human use of it
are a single unit, much as Barker describes his behavior settings.
31
Naturally,
various design theorists apply the concept of synomorphy is varying terms. Below
I will briefly outline how the synomorphic idea has been employed in some of
urban design’s more well-known evaluative traditions, to get at how it might
bridge evaluative thinking about audits, affordance, and human physical well-
being.
One of the most well known urban design studies of the interdependent
relationship of human activity and the physical environment was William H.
Whyte’s Social Life of Small Urban Spaces (Whyte, 1980). Whyte’s detailed
ethnographic observations of street life in New York broke new ground in terms
of how design viewed the human-environment relationship, as a sort of
improvised ballet, in which the individual’s body interacts with street furniture,
space, and other people in ways that define not only how space is used, but also
how the body is used. Whyte went into his study assuming that various external

31
Barker’s research revealed that behavior settings shape the “goals and cognitive possibilities” of their
members in a variety of ways, including a person’s possible choices, which are to a large degree limited to
what settings are available, and so people fit themselves to the settings. “[P]eople select behavior settings that
satisfy their motives and harmonize with their cognitive styles; they inhabit behavior settings (except those
they are required to inhabit…) only so long as they gain some personal satisfaction” (Barker, 1968, p. 196).
134

stimuli (the presence of sunlight, for instance) caused behaviors in public places,
in the sense that people would come out to sit in a plaza because the sun was
shining there, or that children played in the street because they had no
playgrounds. What his research uncovered was a far more complex and
interactive relationship in which individuals assess what the environment will
support in terms of their characteristics and purposes, and act in a kind of
negotiation with the environment, improvising a bench on the corner of a
building, dragging street furniture to a better location, and responding to patterns
of crowd movement in various ways. This synomorphic view revealed a new
perspective on human purposes and abilities in relation to the environment––
Whyte found that many children play in the street not because they have to, but
simply because they want to. Whyte took into account what the actual
environment affords in relation to the particular person––the human body’s ability
to use a fountain as a bench, in light of what the person was trying to accomplish,
such as eating lunch outdoors. Whyte’s synomorphic observations suggested that
public spaces had a “life” of their own that was neither solely the life of the space
or of the individual, but a necessary relationship between them.
32
Physical activity

32
This can be considered in light of Lynch and Carr’s vital, continually changing educative city in “Where
Learning Happens.” The idea of a dynamic and continually changing city is clearly in conflict with the
permanence and persistence of the physical form of cities, unless the concept is opened up to include multiple
uses and adaptation. Whyte gets at some of that with his descriptions of the multiple-use possibilities of urban
set-pieces such as fountains and stairs. Walzer argues for a more institutionally “open-minded space” that
could be conceived to address concerns such as Lynch and Carr’s, at the level of policy and zoning as well as
in terms of the multi-functionality of street furniture and the like (Walzer, 1995).
135

is determined neither solely by human purpose nor solely by the built
environment, but by a complex and dynamic relationship between them.
It must be noted that Whyte’s synomorphic relationship between people
and public environments was in the end an attempt to generalize public behavior
in relationship to things like street furniture and fountains and crowds. Its method
was purely ethnographic and surveillance-oriented, essentially counting behaviors
and interpreting them in terms of broad categories of human purposes (“sitting in
the sun,” “socializing,” etc.). While the Social Life is a landmark study, it was
only a start on the synomorphic research road, a first step in exploring what
ecological determinants could tell us.
Another step further down that path is John Ziesel’s Inquiry by Design
method. Zeisel is as methodologically rigorous as Whyte, but is more inductive
and qualitative in his emphasis on how particular kinds of persons with specific
intentions, in specific contexts, interact in a synomorphic way, with the
environment and with each other. Apropos of capability, Zeisel is concerned with
how these interactions limit, or support, particular activities and intentions. Zeisel
breaks the synomorphic relationship between persona and place down into a set of
ecological categories of actors, acts, significant others, relationships, contexts, and
settings (Zeisel, 1981, p. 124). Into this information he introduces qualitative
interview data of a reflective nature, to characterize the individual’s particular
experiences of the environment, such as why children of a certain age prefer to
136

play basketball on a smaller court, away from older kids (p. 150). A variety of
techniques are finally introduced that establish an average response, including the
“cognitive maps” that figure so prominently in Nasar’s Evaluative Image method.
However, Zeisel is careful to point out that such techniques are only useful “if
you know what you want to find out from people, if you want to discover
regularities among groups of people with particular characteristics, and if you
want to quantify your data” (p. 176). Zeisel is not averse to deductive approaches
in his research, but he is clear in his ultimate aim of arriving at more inductive
conclusions that include a range of human-environment interactions that include
not only “perception” and “meaning” but also the formal and informal
institutional rules of the environment, exchanges of information, and relationships
between people (pp. 180-181). This kind of specificity and complexity allows
Zeisel to generalize within particular cases of synomorphic relationships ––as
opposed to Whyte’s generalizing behavior in all plazas everywhere––in categories
of “degree of privacy,” “territoriality,” and “wayfinding” (p. 217). Significant to
the earlier discussion of behavior settings, Zeisel is also able to consider, in
reciprocal fashion, how “other people’s perspectives” are part of an individual’s
relationship to surroundings (ibid.) Zeisel’s method attempts to be sensitive not
only to the environment’s affect on people, but also to people’s effect on the
environment, for instance in observing how a worn footpath connects two
concrete sidewalks as users adapt the environment to their needs. That sort of
137

“erosion” is a key element in Zeisel’s approach to auditing environments (Zeisel,
1981, pp. 100-102).
I suggest that because it offers a reciprocal view of the human-
environment ecology, Zeisel’s Inquiry by Design takes us a long way toward an
affordance-based description of the determinants of physical activity. However,
we must keep in mind that in Sen’s terms, capability-based determinants of
physical activity would tell us how “setting aggravations and deprivations” in one
category affect them in another category, both in the present and over time
(Banerjee & Baer, 1984, p. 14). Whyte’s method is synchronic in that it offers a
snapshot of an eternal present, which limits the evaluator’s ability to describe how
urban space is produced in a historical context. On the other hand, an historic
perspective must be ecologically associated with other dimensions of human
activity in order to understand the interconnections between aggravations and
deprivations. Alan Jacobs’ Looking at Cities gathers data on the environment’s
past to explain the present and future trajectories of produced space, but like
Nasar’s Evaluative Image focuses on the visual dimension of human activity to
the exclusion of many others (Jacob, 1985). Jacob’s subsequent collaboration with
Donald Appleyard on the Urban Design Manifesto introduced a multi-
dimensional perspective that is more akin to ecological determinants of physical
activity (Jacobs & Appleyard, 1987). The Manifesto is an attempt at distilling
“essential urban physical characteristics” in historical context as well as in the
138

context of human values “such as health and comfort, self-reliance, opportunity,
and accessibility” (p. 112). These authors emphasize the opportunity that public
participation brings, for residents to “become involved, to decide what they want
and act on it” (p. 115). Despite the call for public participation and opportunity,
however, Jacobs’ later work tends to emphasize satisfactions as determinants of
physical activity, as in his Great Streets, which focuses on aspects of the
environment like “physical comfort” and “qualities that engage the eye,” which
presumably promote pedestrian activity by fulfilling preferences for certain
qualities of the urban environment (Jacobs, 1993).
It must be kept in mind that the interconnections between variables are not
confined to a single setting. That is to say, reduced affordance in one setting
might have an impact on a person’s ability to capitalize on opportunities in
another setting (Sen 2000 p. 4). For example, a lack of daycare in the
neighborhood might make it impossible for a single mother to take advantage of
increased job density. Or, a young child raised by an elderly grandparent may not
have access to park facilities if the grandparent is unable to leave the home to
accompany the child to play. Barker’s behavior setting survey method, vast in
scope, specifically captured these sorts of ecological interconnections of what Sen
would call capability deprivation.
It was only at the end of Good City Form that Lynch explicitly articulated
the Barkerian ecological model and its influence on his work. More than any other
139

urban design theorist of the 20
th
century, Lynch made the conception of the
human-environment ecological interaction fundamental to the development of city
form, policy, and their larger social implications. Although his earlier Image of
the City research was focused entirely on the visual-cognitive evaluation of city
form, by the end of his life he had come to embrace an ecological model––
adapted from Barker’s––and a description of collective well-being in the
environment that was quite close to what I would call Sen’s capability.
33

We know that Lynch was aware of the Theory of Justice, and in Good City
Form he acknowledges his debt to Rawls (p. 227). There is no evidence, though,
that Lynch was aware of the work of his contemporary Amartya Sen. The two
men were working in disparate fields, one in urban design and the other in
development economics, although they were developing arguments for placing
the individual at the center of evaluations of well-being (pp. 116, 192, 201, 230;
Sen, 2000, pp. 30-31).
Lynch recognized the growing influence of the Rawlsian Theory and its
definition of distributive justice, and understood that in Rawlsian fashion some
goods must be “unequally distributed” according to differences of need or ability
(pp. 190, 225). He also saw inherent contradictions that would need to be faced:
An equality of opportunities still leaves us unable to understand how the
individual’s personal and social constraints limit the ability to turn resources into

33
For an example of what I mean by this, consider Lynch’s culminating general description of settlement
form (Lynch, 1984, pp. 351-353).
140

well-being (p. 226). The contribution of built form to individual well-being could
not be only in the form of opportunity, but must include the ability to make use of
those opportunities (pp. 200-202). Lynch proposed to evaluate built form
according to how it “distributes life’s chances” in terms of “individual
development” (p. 230). And like Sen, Lynch recognized that there were strong
interconnections amongst the various dimensions of a person’s capability (p. 231).
By the time of Good City Form, Lynch questioned the dominant utilitarian
model and is emphasis on preference, judging it an “unreliable” appraisal of
distributive justice (p. 154). Like Lichfield, Lynch challenges the utility metric at
the heart of the cost-benefit method of assessing collective good (p. 226).
Ultimately, he attempted to discard both the utilitarian metric and cost-benefit
analysis as too limited in scope (p.369). Lynch strongly held to the notion that
Barker’s ecological model offered a new way to evaluate the human-environment
relationship, so that, “Activity itself is not falsely reified; the suitability of any
action is linked to the people engaged in it” (Lynch, 1984, p. 352). What Lynch
was looking for was not simply an evaluation of maximized choices, but rather an
evaluation of what Sen considered the real choices that an individual has reason to
make. Ultimately, Lynch proposed his seven “performance characteristics of good
city form” for normative evaluations of the environment. The characteristics
today appear in a number of AL audits and other evaluative frameworks (Day,
Boarnet et al., 2006; Ewing & Clemente, 2005; Jacobs & Appleyard, 1987). In
141

Chapter 3, I will make use of Lynch’s framework for the human-built
environment relationship in studying whether affordance data can be identified in
AL audits.
Like Lichfield, Lynch criticized the utilitarian metric but never quite let it
go. Like other researchers who were trying to conceive of something like
capability at this time, Lynch imagined mixing together the three kinds of
determinants that I have called satisfactions, opportunities, and affordance.
What Might an Affordance Audit Include?
Audits are practical decision-making tools. The information they gather is
meant to describe a built environment-physical activity relationship with
reliability and precision. I propose that because AL audits have operated without
an underlying general audit model, they have made use of various kinds of
determinants––satisfactions, opportunities and affordances––in various
combinations. This means it is highly likely that affordance determinants are to
some degree operating in AL audits already, although their full value is not
exploited.
Recognizing affordance determinants is the task taken up in the next
section. Here, though, I will review some AL primary research findings and
conclusions that suggest affordances, and consider how a determinant of
affordance might be developed from them.
142

1. Cohen, Finch et al. show findings indicating that even when children
have the ability to walk to destinations, if their neighborhoods lack a
kind of “social efficacy” the chances that the child will be physically
active decrease (Cohen, Finch et al., 2006). In this finding we can
recognize that deprivation in one category of the child’s life is
caused by its interconnection with deprivation in a separate category.
2. Amongst poor children, multiple disadvantages can have cumulative
effects on their ability to make use of the built environment.
Bauman, Silver et al. found that poverty, minority race/ethnicity, low
parental education, and not living with both biological parents were
strongly associated with a child’s poorer overall health, having a
chronic condition, and having a limitation of activity (Bauman,
Silver et al., 2006, p. 1326). Bauman, Silver et al’s “accumulation of
social disadvantage” measure recognizes the cumulative forces
acting on the child’s real lack of choices for physical activity. To
develop an affordance determinant of physical activity, the
researcher would need to first identify a set of opportunities for
activity offered by the built environment. The environment could
then be said to afford what activities the child can engage in beyond
the constraints of the accumulated disadvantages.
143

3. In their study of how the layout of suburban neighborhoods affects
physical activity, AL researchers Cervero and Duncan concluded
with a note of frustration. They recognized that neighborhood form
alone couldn’t represent all of the mediating environmental qualities
that the resident needs to support physical activity (p. 142). They
concluded that the reasons for walking are tied to local community
identity, to such a degree that their attempts to generalize about it
were confounded. In the end, they argued that it might be best if
neighborhoods were simply designed for those who are most likely
to use them, which begins to suggest an affordance-related
normativity.
Those authors concluded that in terms of urban design interventions to
increase physical activity, “…a greater public health benefit might accrue from
designing walkable neighborhoods that appeal to the niche-market characteristics
of different demographic groups versus microdesigning places in hopes of
swaying travel behavior” (Cervero & Duncan, 2003). Identifying the
neighborhood’s affordance for physical activity would require that the “niche-
market characteristics” describe both how they are physically active and what
environmental “hardware” supports it.
I have argued that Gibson’s “affordance” is an appropriate metric for
evaluating capabilities. In Chapter 3, I will examine a number of AL audits to
144

determine the types of observations they make, whether satisfactions,
opportunities, or affordances.
Conclusion
In this Chapter, I have attempted to describe the three types of information
gathered in the general audit model, as satisfaction, opportunity, and affordance. I
have described how each of those is viewed in the environment according to one
of two conceptual research models, the correspondence model and the ecological.
I argue that the AL ecological model, while uniquely suited for research on the
childhood obesity epidemic, ought to be extended to describe the human-
environment relationship as a reciprocal unit, one of affordance. I conclude that
because affordances describe ratio relationships between person and environment,
they are uniquely appropriate to describing capability in the built environment,
and that future AL audits ought to take affordance into account.
In the next section, I undertake analyses of six cases of AL audits. My
hope is that this exercise will aid AL researchers in clarifying their use of the
correspondence research model, to begin to inform audits with affordance-based
determinants.
145

Chapter 3
Evaluating the Audit’s Informational Basis: A multiple case study
Analyzing Audits Using the General Audit Model
This study is premised on the idea that there has been no conceptual model
for the development of audit instruments, at least not one that considers both the
dependent and independent variables of health and the built environment, and that
also takes the audit’s distributive function into consideration. Developing the
general audit model in Chapter 1 brought us through arguments and questions
about distributive justice and the philosophy of science, in order to distinguish the
difference between an audit and primary research. Chapter 2 took up the question
of the existing ecological conceptual model, and showed that it is the appropriate
one for describing capability data. There I introduced the concept of affordance.
The study thus far has been theoretical, but it is also meant to inform future
empirical research, and so it is time to relate the GAM and the affordance
ecological model to actual audit instruments.
As I said in the preface, I did not set out to do a theoretical study. I had
originally intended to conduct a mixed-methods study of the six cases of built
environment audits, to determine which ones gather more of which type of
evaluative data. However, my analytical plans were frustrated: Initially it was my
intention to re-categorize all of the audit’s “data points”––the individual items
146

that each audit asks for––according to the typology of satisfactions, opportunity
and affordance. My thought was then to use some quasi-statistical methods to rate
each of the audit information, and then “rate” the various audits as to what type of
information they contain. That effort was quickly frustrated by the fact that
although the audits gather different types of data, they gather it through different
methods, and those methods help to define the type of data. For instance, the
SAGE data is at face value a standard inventory of park environments, but when
the audit is considered as one step in a much larger evaluative process, SAGE’s
regional scope, interdisciplinary emphasis, and social justice program are
revealed.
The decision to use a multiple case study method was a conclusion made
in light of the fact that it is not the types of data that distinguish audits, but rather
it is the processes in which they were developed, and the standards and
assumptions at work, and of course the choice of conceptual model, that
ultimately tells us what sort of information an audit uses to evaluate the built
environment.
The research below takes the form of an instrumental, descriptive,
multiple case study. It is "instrumental" in the sense that the subject is the set of
six audits and the general audit model, and not an environment that will be
audited. The real focus is the general audit model, but the audits serve a
147

"supportive role, a background against which the actual research interests will
play out" (Berg, 2001, p. 229).
The "descriptiveness" of the study refers to its descriptive theory, which in
this case is the general audit model, and all of the underlying theory that has been
discussed in the first two sections (Yin, p. 111-112). I have adapted Yin’s
framework for descriptive case studies somewhat, and the protocol for the case
studies is extrapolated from the general audit model. The rubric can be seen as a
set of questions for interrogating each case. Ideally, the data from the cases
reflects back on the workings and limitations of the general audit model. The
protocol that will be use with each case is as follows:

1. How does the audit describe the individual?
o In terms of absolute/relative portion of the good?
Select cases
Develop theory
Design data
collection
protocol
Conduct first
case study
Develop policy
implications
Write
individual case
report
Write
individual case
report
Conduct
second case
study
Write cross-
case report
Modify theory
Draw cross-
case
conclusions
Figure 8. Multiple case study (after Yin) 2003)
148

o In relationship to some constraining characteristic or
circumstance?
2. How does the audit describe physical well-being?
o Physical well-being as an end-in-itself
o Physical well-being as means-to-an-end
o Physical well-being as satisfaction
o Physical well-being as opportunity
o Physical well-being as affordance
3. How does the audit describe the built environment?
o Designation of dependent variable (y)
o Designation of independent variable (x)
o Designation of proxies for the independent variable (x)
Before launching into the individual cases, I will offer a few observations
that apply to them overall.
How Audits Gather Information
There are two standard methods that audits use to gather data. The first is
the survey, conducted by phone. The other is field observation by trained
researchers. One AL audit departed that standard and introduced heart rate
monitoring of children as a triangulation method. Mostly, though, each tool
employs one of those two approaches, and none employs both methods. Still, a
number of the authors have suggested that there is good reason to triangulate
149

survey findings with different kinds of data, but to date the audits generally stick
to a single method. Chapter 1 argued that primary research and audits ought to be
considered as two separate procedures, with different aims and development
processes. Historically, primary research has made very clear distinctions between
subjective versus objective data.
34
I assert that when primary research and audits
are conflated, the value of the differences between subjective and objective data
can be lost to the auditor. For example, phone survey audits gather data from
respondents that includes not only “subjective” information such as data on mood
and affective states that affect exercise. Objective data is also considered to be
useful, as in the St. Louis audit, a phone survey, which rigorously tests the inter-
rater reliability of the responses. The purpose of that is to determine the reliability
of the choice of determinants: The audit’s high inter-rater reliability for
“Availability of walking trails/tracks/paths” means that the determinant’s results
are reproducible, one of the factors that establish the stability and validity of
variables. As we discussed in Chapters 1 and 2, a great deal more may be required
of audits than stable determinants, although those are the key

34
The subject-object distinction is the focus of much argument since the postmodern turn in qualitative
research, at least since the 1990s. We might also consider, though, that Barker was implicitly challenging that
distinction with his behavior setting method, which cuts across subject-object distinctions. Interestingly,
Gibson argument that “perception refers to something and is not caused by something” is a similar challenge
to the subject-object dichotomy, since it suggests a unity of person and environment (Lombardo, 1987, p.
122).
150

Case: Irvine-Minnesota Audit Instrument (IMI)
Description
The IMI is an inventory-type audit, conducted by trained researchers
through field observations. It is a 60-item audit of the built environment. Like
SPACES and SAGE, it does not observe how the built environment is used,
except in terms of observable evidence, such as facility maintenance. Children are
not taken into account as a separate group.
The audit is accompanied by a “code book” and instructions. Its unit of
observation is the neighborhood, the definition of which is left up to the user, be it
by census tract, political districts, or by a “coherent physical appearance,” or
according to “homogeneous populations (such as a senior community)” (Day,
Boarnet et al., 2005, p. 3). The setting is divided into smaller segments for
analysis, “typically defined as a block face” (ibid.).
The IMI proposes to deliver objective data in four categories or “domains”
of policy research on the built environment. What is of interest in each setting is
its “potential…to support physical activity” (p. 4). The four domains represent
four categories of determinants or “factors” that “may influence physical activity”
(p. 146):
1. Accessibility
2. Pleasureability
3. Perceived safety from traffic
151

4. Perceived safety from crime
The codebook contains detailed instructions to guide the researcher in
each observation. Some additional data is required to thoroughly complete the
audit, including GIS information on density of origins, and job density.
The IMI is particularly concerned with the “perceived psychological
barriers” to physical activity, defined as something that “discourages people from
walking…such as a major 8 lane road” (p. 23). Both “tangible physical barriers
or perceived psychological barriers” are considered, and the criteria for barriers
are described. Psychological barriers are described simply as something that
discourages walking in a place even if walking there is physically possible, and
physical barriers are things standing between the pedestrian and a “possible
destination” (p. 23).
In keeping with the IMI’s focus on perceptions, it includes other
categories suggesting subjective perceptual data, such as “attractiveness” and
“interestingness,” defined as “remarkable, fascinating, or attention-grabbing” (p.
21). Still other categories suggest more objective data, like “land use.” “Safety,”
which appears is all the AL audits, in here as well.
The IMI is an inventory, and does not observe actual use of the built
environment. The observations are scaled in binary (yes/no), or three or four-point
Likert scale. Some items are scaled counts (“1-2 stories, 3-4 stories, 5 or more”
e.g.), or a single number (“Posted speed limit”). Attractiveness is on a three-point
152

scale (“attractive-neutral-unattractive”), and barriers on a four-point scale (“no
barrier-can not be overcome”). Psychological barriers have no particular listing of
their own in the audit, except perhaps a barrier score named “other.”
Subjective information on convenience and traffic safety are scored on
three-point Likert scales, and the historic status of buildings is also a three-point
scale (yes/no/NA). Although some additional GIS data is required of the auditor,
it is based no census data, and not on the data gathered in the 60-item inventory.
IMI, Satisfaction, Opportunity, and Affordance
The development process for the IMI, like that of many audits, involved a
variety of research methods including literature review, interviews, and consulting
with experts. In its development phase the IMI considered particular groups of
individuals and their relationships to the variables in relative fashion, during its
audit development process. By the end of the development process, however,
identifying characteristics and circumstances had made no real contribution to the
audit. The IMI developers conducted focus groups in order to include “specific
groups” who might have been overlooked in the existing literature, including low-
income people, teens, and non-white college students (Day, Boarnet et al., 2006,
p. 147). The focus groups were intended to identify built environment features
that might impact an active lifestyle (ibid.). Yet, they also note, “The purpose of
the focus group interview was to identify additional built environment features to
153

add to the inventory, and not to assess the relative importance of different features
to diverse groups” (ibid.).
As is the case with any audit, the kind of primary research that is
conducted prior to audit development has metrical implications within the audit
itself. In this case, data gathered from excluded groups was never meant to stand
separately as a control or mediating variable, but was from the start intended to be
averaged into the audit findings.
Ultimately, the IMI developers found that the focus groups produced only
redundant data, already included in a preliminary audit that the developers had
drawn up. The focus groups were ultimately abandoned. The IMI developers
ultimately became aware of the limitations caused by not taking identity into
consideration, and they noted it in their report on the development process (p.
147). In their conclusions they note,
Focus group questions might have generated more novel responses if
participants had been questioned about how their own identities (as teens,
low-income individuals) influenced their opportunities for active living.
(p. 150) [Emphasis added.]

This part of the IMI’s story suggests that there was a growing awareness
in the developers’ minds of the possibility of the audit evolving beyond its
original conception, to become something more than a straightforward inventory.
To the credit of the IMI developers, their satisfactions-based audit is
thorough in its set of determinants. Yet, because the audit offers a static
conception of the environment, it does not record synomorphy data as simple as a
154

foot path worn into the grass connecting two sidewalks. This broadens the
analytic focus to such a degree that very little is actually in the audit’s sights: The
IMI’s measure of “steepness” (“steep/moderate/flat or gentle”) is so general as to
apply as much to a mother pushing a stroller as to a tractor, or as much to an
elderly person as to a teenage jogger.
Consider the IMI data item on “Public Space.” Here, the implications of
“access” are suggested simply by whether public spaces are present or not, in the
form of parks, plazas, beaches, public gardens, and such. This conceives of the
human-environment relationship in linear terms, in the sense of assuming that if
certain equipment is in the landscape then people will exercise. This overlooks the
conception of public space as an integral part of the daily life of a neighborhood,
as a meeting place, a social hub for information and transportation and
conversation, as an alternative to home and work, as a kind of outdoor living
room, and of course as a place for exercise and physical activities, from jogging
and weight lifting to picnicking and feeding ducks and participating in community
activities. The IMI has two items on public space, the first to determine the type
of public space––a list of six possibilities is offered–– and the other item is called
“Use the Public Space?” This item is restricted to a binary response indicating
whether “it is possible for most people to get into or otherwise use the public
space.”
155

“Pleasurability” and Affordance
For each of the audits examined, I will draw one of its data points out of
context and re-consider it for its contribution to affordance data. My hope is that
this will ground the case studies in an applied context, albeit a hypothetical one.
In the case of the IMI, I have chosen to re-consider the heading “pleasurability” in
affordance terms, particularly the audit item on “attractiveness of views.”
It is important to keep in mind that I am not prescribing how one should
go about designing views. Rather, I am suggesting a way to re-consider how we
use the environment to define what is “good” in a good street.
A static audit views the street either as a field of activity stimuli, or else
as the sort of breathtaking rest-stop view one sees staked off in national parks.
This de-emphasizes the intense interactions between variables that one finds on an
urban street. I am not saying that pleasant views are worthless, on the contrary
they are extremely valuable. It’s just that the affordance perspective asks us to
consider the use value of that pleasure, rather than focusing on pleasure as an end
in itself.
The IMI generalizes its assumptions about why people are active, and then
oversimplifies the meanings that physical activity might have to people. The
audit’s conception of the street as an aesthetic magnet draws the analytic focus
away from a more subtle interplay between the environment and people’s lives.
On vacation we might travel long distances to experience a certain view, but
156

during rush hour traffic may not give a second thought to a view, however
stunning.
To gather affordance data on pleasureabilty, the IMI could observe the
times of day and year that the view is actually used, and also get some notion of
who uses it, and the degree to which different people have access to it. A pair of
lovers meeting for a fervid kiss, or a retired person resting on a stroll, may each
enjoy the view, but for different reasons, depending on the what the view reveals,
and who else is there, and a host of other possible factors. Affordance suggsts that
the value of a densely tagged graffiti wall is not necessarily less pleasurable than a
view of a manicured garden. It is a contextual matter, and context can include
time, and other people, and events, and not simply the physical form of cities.
Pleasurability should not be rejected out of hand as affordance data, but it must be
extended, in order to give it ecological application.
Case: The St Louis Audit Instrument
Description
The St Louis Environment and Physical Activity Instrument is a phone
survey that describes the responses of the average adult neighborhood resident to
104 questions about their general health physical activity, the built environment,
access, attitudes, ability, and institutional matters that affect exercise. The
dependent variable y is very clearly the “physical activity” of the average
neighborhood resident. The independent variable is described in terms of a variety
157

of factors, specifically social capital-type assets as well as transportation
resources, and also access to information and health programs. Emphasis is on
how the amounts and reasons for exercise might relate to the built environment,
social environment, and various personal characteristics, attitudes, and
circumstances. Like most of the AL audits, the St. Louis audit studies
neighborhood-sized segments. Its survey responses are meant to provide a
complete assessment of a community’s physical activity in terms of the built
environment. Children are not of specific interest to this audit, and in fact it is
designed to be targeted at adult residents of the household. The audit sub-divides
the 104 questions into seven categories and three sub-categories. The seven
categories and three sub-categories are:
1. Walking Assessment
2. Barriers
3. Infrastructure for Walking and Cycling
4. Neighborhood Surroundings
5. Neighborhood Safety
6. Social Assets
7. Community Assets
8. Policy Attitudes/Access
9. Sedentary Behaviors
10. Sociodemographic and Other Factors
158

The St. Louis audit’s survey data is meant to establish statistical averages
of physical activity and its associated influences, including mediating variables
like sedentary behaviors. In ecological fashion, it associates x variable influences
at multiple levels, such as personal characteristics (sedentary behaviors, e.g.) to
larger policy issues (health education programs and access to information,
neighborhood safety, e.g.). Although the audit data is clearly intended to create a
picture of the average resident, its scope is broad enough that it seems to cover all
three types of audit information described by the GAM: satisfactions,
opportunities, and affordance.
Satisfactions, Opportunity, Affordance, and the St. Louis Audit
Data on the average resident’s preferences for types of exercise is gathered
by the audit, and contrasted with data on why the resident might not exercise, in
terms of the sense of convenience of exercise facilities, and even the fear of
violent assault while using those facilities.
Yet before the data is processed, the St. Louis audit draws out information
that could require introspection on the part of respondents, regarding feelings
about being in crowds, or having the supportive company of friends, for example.
In a manner akin to Barker’s vision of the ecological human-environment
relationship, the audit gathers information on feelings of aggravation with the
presence of other people, and the feeling of fear that isolation can cause.
159

The survey also gathers information on types of exercise preferred, as well
as the sense of convenience of exercise facilities, and the sense of safety while
exercising. As was argued in Chapter 1, only certain kinds of goods can be
distributed with absolute equality, and in this case the goods are the levels of
pleasure or preference fulfillment that encourage or discourage physical activity
in the average rational adult.
The St. Louis audit also gathers data on what I have called general
resource opportunities. “Access” plays an important role in painting the picture of
what the average adult resident has to work with:
1. Access to transport
2. Access to information
3. Social capital
By “social capital” I refer to a rather lengthy inquiry into the individual’s
sense of being at home in the neighborhood, of sharing values with others of the
community, of depending upon neighbors, of helpfulness and mutual respect.
There are overlaps between the different kinds of information in this audit.
For example, I have categorized access and social capital both as opportunities––
means to an end––although clearly having friends and feeling supported in one’s
neighborhood could also be considered an end in itself. I have answered the
question of what kind of information this audit gathers by considering the larger
purpose of the survey, which is to determine what means the individual has on
160

hand. I call this an opportunity- or means-based focus, since it is more concerned
with an equality of means, rather than an equality of the ends to which such
means are applied.
Sources of information are of interest to this audit, as a resource that the
average resident can use in pursuit of health. Although the St. Louis audit does
not specifically indicate it, it would be reasonable to presume that the presence of
a social network for transport to exercise could also serve as a source of
information about opportunities for exercise, health and diet.
The St. Louis audit’s informational basis focuses primarily upon the
equality with which opportunities are provided. Yet, the St. Louis audit could also
be said to gather information that is necessary for an affordance audit: As was
described earlier, affordance is "opportunity plus agency," and so it is reasonable
to view this opportunity-based audit as containing information that could be
transformed into affordances. Further, this particular audit gathers information not
only on general resources/opportunities, but also on the factors that constrain the
individual’s use of them.
That is a key point for any affordance audit. In the case of the St. Louis
audit, information on such constraints includes the individual’s health; frequency
and intensity of exercise (proxies for health); location of exercise; personal
reasons for exercising; the sedentary behaviors that prevent exercise; the
individual’s moods and affective states that prevent exercise; physical states
161

(health) that prevent exercise. Yet, it seems this audit depends primarily upon
opportunity data, since it does not identify individuals to whom affordances
apply. An affordance audit would required defining the constraints that determine
agency, and the ability to take advantage of opportunities.
The St. Louis audit does not do this. It is true that this audit rigorously
tests the inter-rater reliability of responses, but that is to determine the reliability
of the choice of determinants. In affordance terms all that we can tell from a high
inter-rater reliability (IRR) rating on “Availability of walking trails/tracks/paths,”
for example, is that the field observation of this determinant is reproducible. IRR
does not ensure the validity of the determinant as a measure of affordance.
Information and Affordance
This audit is unique in the emphasis it places on information. Information
is valued as opportunity here, as a resource for physical activity. For example, the
audit focuses on the information available to the average resident about
employment-based health programs. It seems obvious that the pursuit of health
requires information about what the environment will support in terms of physical
activity, which suggests that information as opportunity could strongly contribute
suited to an affordance audit. Information as opportunity is information about
what one might do with the availabilitiy of facilities and programs for example.
Information as affordance is information about what one actually can do. I
suggest that the St. Louis audit stops at the conception of information as an
162

opportunity, but that such data could inform affordance data if the audit brought
more attention to the interaction effects between various determinants, and the
interlinkages between states of well-being.
At present, the St. Louis audit considers the possessing of access
information and fitness information as determinants of well-being. Yet it is not
difficult to imagine that people who are aware of fitness programs at work may
nonetheless never make use of such programs. In order to determine what the
environment actually affords in terms of physical health the audit would need to
connect data on available information to data on individual agency and constraints
on the acquiring and use of information: Do the elderly have access to the
Internet? Does a child have the transportation, orientation, and security needed to
take turn information into physical activity?
This audit could observe affordance if its conception of the reasons for
physical activity were extended, beyond the supply-side paradigm. This would
involve taking into account individual reasons for physical activity. Opportunities
to be physically active include much more than the programs and equipment that
are provided by cities and organizations; the ability to convert existing
opportunities to physical activity ought to be given some focus. In considering the
interlinkages between well-beings that are the hallmark of capability, this audit
could consider that physical activity takes place in ordinary life, and not only in
gyms and community centers and employee health programs: Having the
163

information needed to undertake a journey to visit friends across town could
increase the physical well-being of the elderly while also increasing social well-
being. The ability to bring a grandchild along on such a journey could raise a
number of well-beings in the child simply because of the educational value of the
adventure. One can imagine a variety of interconnections, in the sense that having
friends and having transportation not only requires information in terms of travel,
but having friends and taking journeys are themselves sources of information that
could make it possible to take advantage of still iother opportunities for activity.
This is in keeping with Sen’s idea of the interconnections or interlinkages
between freedoms (1999, pp. 4, 40, 43, 195, 284).
Case: Systematic Audit of Green-Space Environments (SAGE)/
Green Vision Plan
Description
The SAGE instrument is part of the larger Green Visions Plan (GVP)
regional study of parks, which among other things assesses the fairness with
which Southern California parks and park resources are distributed. SAGE’s
contribution to the GVP is an inventory of park environments. SAGE was used
for every park in Southern California, and that data was analyzed in Sister, Wolch
et al.
164

SAGE is a standard field audit, although instructions for auditors direct
them to also gather information on parks from park websites. The units of
observation are individual parks throughout the Southern California region.
The authors note that SAGE is largely based on determinants from audits
by other authors, including the Systematic Pedestrian and Cycling Environment
Scan (SPACES), which is one of the audit cases in the current study (Pikora, Bull
et al., 2002). SAGE is an inventory of facilities and services, landscape features,
the condition of various aspects of the parks, and objective measures of security.
It is a stand-alone inventory of detailed park data, and informs the additive indices
that Sister, Wolch et. al develop for their initial ratings of the provisioning of
Southern California parks. Those findings are ultimately combined with geo-
spatial research on the distribution of green space among different demographic
groups across Southern California communities. Combining the SAGE inventory
data with geo-spatial and demographic data via GIS creates a powerful tool with
which the GVP authors are developing a comprehensive description of the
relative equality with which Southern California park resources are distributed. It
is the hope of the GVP study authors that their work can explain the lack of park
use by inner city children, and also to inform recommendations for policy and
design interventions. The GVP is by no means focused primarily on the obesity
epidemic, although that is definitely one of its concerns.
165

All of the SAGE data are objective in the sense that they are gathered by
trained researchers who observe the park itself. No surveys or interviews with
park users are conducted. SAGE gathers data on what facilities and services are at
each park, what landscape features are there, as well as estimates of proportions
of paved surfaces, irrigated areas, and areas for organized recreation. Park
conditions include litter and graffiti, noise, overgrowth, conditions facilities and
signage, and an overall scaled estimate of levels of maintenance.
SAGE evaluates Southern California parks on a number of “dimensions”:
1. Park size
2. Park type
3. Safety
4. Condition
5. Landscape Diversity
6. Naturalness of Unpaved Park Areas
7. Runoff
The last two dimension, naturalness and runoff, reveal the GVP’s concerns
for larger environmental sustainability matters, although clearly these are also
relatable to park use.
The dimensions are scaled along a set of “continua”:
1. Small–large sizes
2. Single–many recreation uses
166

3. Safe–unsafe
4. Poor–excellent condition
5. Simple–complex landscapes
6. Low–high multiple use potential
As an example of how the dimensions are described by the continua,
consider the dimension of “safety,” which is scaled on the continuum
“safe/unsafe.” This represents data on three phenomena: (a) the presence of
emergency call boxes, (b) security staff, and (c) park staff. The continua are
additive, so that the presence of any of the three phenomena indicates a greater
amount of safety, and an absence of any of the phenomena indicates less safety.
Other categories of data are similarly aggregated and scaled.
SAGE data is divided into four major categories, which represent the
audit’s independent variables:
1. Facilities and services
2. Landscape features
3. Condition
4. Safety
The dependent variable is park use by the average neighborhood resident.
Along with their fieldwork training, SAGE auditors are given clear
descriptions of each level of the scale in each section (e.g., “Excellent: Site is very
167

well-maintained, facilities are new, no litter or graffiti,” and 90 photographs are
offered as a reference for each data point, and for each of the various scalings).
SAGE itself makes no such distinctions between park resources and how
they are used, assuming rather that certain levels of particular determinants
equates with increased physical activity. As mentioned above, SAGE is one step
in a much larger regional analysis within the GVP, the scale of which covers the
entire Southern California region (see figure 9). SAGE offers a fine example of
how audit data is used in subsequent research, and in policy development, and it is
worthy of study for this reason alone. On the other hand, children are not set aside
as a specific concern in SAGE, although they become a demographic factor in the
larger GVP, of which SAGE is a part. In the context of the obesity epidemic, it is
noted that SAGE does not specifically take children into account, but the larger
GVP study is very much concerned with children’s opportunities for park play
(see Sister, Wolch et al., p. 115). Also, SAGE/GVP illustrates how primary
research, audits, and built environment interventions are interrelated as a sequence
of activities that inform policy. When all of the SAGE audits of all Southern
California parks were complete (some 1200 in number), Sister, Wolch et al.
generated from that data a set of additive indices to describe the level at which
each park provisions its resources.
168

These indices represent SAGE’s independent variables over all of
Southern California. They are averaged for each subregion (figure 9). With the
subregion averages in hand, the GVP authors then determined what resources are
available to each region across the entire Southern California region. That in itself
provides an archive of valuable information that can be used by researchers,
community groups, and park professionals, however it is only a second step in a
larger research program. In the next step, researchers determined who,
demographically speaking, was using those resources, by sub-region (Figure 10).
What this reveals is the disparities of park resources by demographic
characteristics, by subregions, for all of Southern California.

Figure 9. Green Vision Plan study area, Southern California region (Sister,
Wolch et al. 2007).

169

These indices are additive, drawn from the SAGE data (p. 37). They are:
1. Basic Facilities Index
2. Active Recreation Index
3. Passive Recreation Index
4. Community Index
5. Landscape Diversity Index
6. Safety Index
7. Condition Index (p. 37)
One finding was that low-income Latino neighborhoods tended to have
fewer parks overall. Another was that a negative relationship exists between the
number of low-income children in a subregion and accessible park resources per
1000 kids (p. 112).Future GVP studies include a "park pressure" study that will
investigate level of service and park congestion around particular parks, and a

Figure 10. Average index values for Green Vision Plan parks, across all the
categories of the park audit (Sister, Wolch et al., 2007, p. 37)

170

third study will establish a park "need rating" that takes recreational programs into
account alongside the depletion of access caused by local crime, and also the lack
of physical fitness amongst neighborhood children.

Comparison: SAGE/GVP and the IMI
It is uncertain whether SAGE is as opportunity-focused as the larger GVP
study. By its developers’ own descriptions, SAGE and the GVP are ultimately
focused on evaluating the equity with which park resources are distributed in
Southern California, which they call “park supply” (p. 15). That equity is what I
am calling “opportunity” along Rawlsian lines. SAGE/GVP and the IMI can be
compared in terms of the “satisfactions” and “opportunities” lines of valuation in
the general audit model, but it is necessary to keep in mind that we are not
studying how parks ought to be designed and provisioned, but rather how audits
evaluate the presence of some equally distributed good.
In comparing SAGE/GVP and IMI, I note that SAGE/GVP conceives of
the environment as a site of contest regarding the equitable distribution of rights
and opportunities and access. The presence of various park elements, such as
safety, are framed in terms of the physical health resources residents have a right
to expect. There is no question that SAGE/GVP employs determinants of physical
activity, but it does so under the assumption that their presence represents the
equitable distribution of rights and opportunities.
171

The IMI audit, on the other hand, conceives of the environment
instrumentally, as a set of determinants that trigger certain feelings in residents
(safety, interest, delight, etc.), and these inspire physical activity. The IMI
inventories neighborhoods to evaluate their instrumental ability to trigger physical
activity.
Thinking about these two audits along the "lines of valuation" proposed in
the general audit model in Chapter One, we see a clear distinction between the
IMI’s “satisfactions” and the SAGE/GVP’s “opportunities.” The IMI conceives of
the good in consequentialist terms, as a feeling, and an end in itself. By contrast
SAGE/GVP conceives of the good in deontological terms, as a free-standing set
of opportunities and rights. SAGE/GVP assumes that a resident’s desire to be
healthy and active exists in the person independent of the presence or absence of
the resources, and the role of the audit is to evaluate the level of those rights and
opportunities. For SAGE/GVP, quality of life is the presence of supports for
physical activity. Conversely, for the IMI, quality of life is the presence of the
stimulants that cause physical activity.
SAGE/GVP assumes that the resident wants an active life, and merely
incidents of activity, and certain kinds of resources are required to get it, such as
access and opportunity. For example, a lack of safety determinants in parks
signals to SAGE/GVP a denial of access. Although satisfactions-style rational
choice plays a role in SAGE, those choices are not ends-in-themselves, but
172

opportunities for pursuing active living and health. The distinction may be subtle,
but it has implications for how the SAGE/GVP inventory might be used in
arriving at built environment interventions. A rational choice/satisfactions
approach like the IMI would provision parks according to an ideal set of
determinants that the average resident rationally prefers. SAGE/GVP provisions
parks with the things that residents need in order to pursue whatever sort of active
life they prefer, and these may be actual objects, or they may be institutional
frameworks.
SAGE, Preference, and the Rawlsian Priority of Liberty
In his criticism of the utilitarian approach to evaluating collective good,
Rawls charges that the utilitarian definition of the good, which averages
preference, does not allow the evaluator to “take seriously the difference between
persons” (Rawls, 1999, p. 163). That is to say that in the utilitarian conception
some persons––those with preferences that do not conform to the general one,
presumably––will always be averaged out of the good. Rawls argues that this is
an unfair arrangement. His alternative, “justice as fairness,” does not propose to
dismiss rational choice altogether, but rather to move the focus of rational choice
from end-state utilities to means-based primary goods (general resources) (Sen,
2000, p. 72). This is the Rawlsian idea of “priority of liberty,” which argues that
when an issue of fairness comes up between people, the parties involved are on
some level aware of the fairness (or unfairness) of the “background” principles
173

that structure what anyone can do in the situation, and if pushed to it, would give
priority to the liberty to pursue ends––to “realize higher order interests”––rather
than protect only, say, “their desires for money and wealth, for food and drink…”
(Rawls, 1993, p. 75-76). Justice as fairness recognizes the predictive power of
rational choice, but it is rational first in its prioritizing of the structures that
guarantee the pursuit of one’s preferences. In Rawlsian terms, satisfactions are
clearly linked to opportunities, but the appropriate measure of collective well-
being must be the person’s ability to pursue her own ends, rather than the ends
themselves. On this point, the ideas of Rawls and Sen are in agreement.
The priority of liberty idea extends into the realm of audits. SAGE/GVP
probably very reasonably assumes that the average resident will choose the
presence of security guards over their absence, and the upkeep of park facilities
over their neglect. However, it is not the pleasure or satisfaction that people take
from these things that SAGE/GVP values, but rather the fairness with which these
things are distributed.
SAGE and the Ecological Model
I argued in Chapter 2 that determinants of satisfaction operate more in a
correspondence model of perception, and that opportunities and affordance are
more ecologically oriented. If SAGE/GVP emphasizes opportunities, then we
ought to be able to recognize its ecological orientation, either in terms of the
174

trans-disciplinarity that is the focus of the AL ecological model, or in terms of the
reciprocality of person and environment that we saw in Barker and Gibson.
One of the SAGE/GVP indices assesses the “Multiple Use Potential” of
parks, referring to their role as mechanisms for restoration of the larger urban
habitat, with playgrounds and ball fields doubling as bioswales and runoff
catchment areas, for instance (Sister, Wolch et al., p. 17).
35
In this, the disciplines
of park planning and environmental sustainability planning intersect in a creative
and vital way. The quality of life as a matter of public health is extended to larger
and longer-term scales, so that a child’s afternoon in the park is conceived in the
context of the urban environment’s impact on her over the course of a lifetime,
and even into subsequent generations. SAGE/GVP is the only one of the audit
studies to consider the sustainability of parks to bridge those disciplines, and to
consider the human-environment ecology within the context of sustainability.
I argued earlier, with Sen, that different types of human well-being are
interlinked, and that an evaluation of quality of life ought to reflect this. A child’s
social, emotional, material, even political well-being (in terms of participation in

35
Walzer’s concept of “open-minded space” is relevant here (1992). To my knowledge Walzer has not
developed the concept into an applied model for city planning and policy, but possibilities for such a model
can be imagined. For example, in its current form SAGE evaluates “multiple use potential” according to pre-
set categories of recreation activities. The reader may recognize Whyte’s argument for the multiple-uses of
the built environment in the plazas and parks of New York City. In the case of both SAGE and Whyte,
however, the inventory-style audit only reveals the presence of absence of resources, and overlooks what
might be done with those resources. An affordance audit can be imagined as useful here, for cataloguing
various features of the environment that give the built environment multiple uses. The difference between the
resource-based inventory audit and the affordance audit in this case is that SAGE is taking an inventory of
features, while an affordance audit would be gathering data on what might people might actually do with the
features. For example, although Whyte’s study did not explicitly state it, there is no question that adults are
the study population. It would be interesting to gather data on the environment that could be sifted and re-
combined in various ways, to determine what the environment will support for different sorts of individuals.
175

planning and design) could be interrelated in the ecological model. SAGE/GVP
emphasizes this somewhat, although does not give it great emphasis.
This is not to disparage the SAGE/GVP emphasis on opportunities. The
study connects the child’s well-being to a sustainable environment is a way that is
a call to creative thinking and advocacy in planning and design, to maximize the
child’s opportunity for health. The ecological focus of SAGE/GVP is suggested
by its conclusion that planners and designers must develop unused pieces of land–
–what Lynch called “wastelands” (Lynch, 1984, pp. 217, 445-447)––as recreation
space.
A fuller embrace of the ecological model regarding the diversity of
landscape would have to take individual use in mind, which SAGE/GVP does not
do. If the study were to consider the question of how each feature is used by
different sorts of persons, then diverse features could be linked to the abilities and
purposes of groups of individual park users, such as those children who are most
at risk for overweight and diabetes.
Regarding the question of developing waste spaces, SAGE/GVP would do
so in order to equalize the resources that are available to any child in the city.
Given the arguments in his ecologically oriented "Where Learning Happens,"
Lynch might argue that it is the individual's ability to turn that resource to well-
being that should be our evaluative focus (Lynch & Carr, 2002). That would be a
176

measure of the equality of resources plus the agency to make use of those
resources, or what Sen calls capability.
The SAGE/GVP authors exercise the strengths of the general resources
approach, but they also recognize its limitations. They note in their conclusions
that simply increasing park acreage and the provision of equipment in poor
neighborhoods is not enough, since access to opportunity can take more subtle
forms, such as the mediating variables of neighborhood crime rates and other
trans-disciplinary issues that they note are important for future research to take up
(Sister, Wolch et al., 2007, p. 21).
The SAGE/GVP authors exercise the strengths of the general resources
approach, but they also recognize its limitations, and get closest to an affordance
perspective when the SAGE data reveals distributions of resources according to
demographic information (Figure 11). That is a simple depiction of which ethnic
group has more or less park acreage in each of the subregions, but more complex
graphs could be extracted from the same data. If data were available on what age
groups, ethnicities, and household incomes actually use which park facilities and
equipment, hypotheses could be investigated that determine how certain children
actually make use of certain resources. One could propose, "Do children whose
local parks have an abundance of green space engage in a greater number of
activities per child?" or, bringing in neighborhood census data, once could
investigate whether being poor and raised by a grandparent makes children less
177

likely to use parks. Other household data could be factored in, such as whether a
disabled person resides in the home, for instance. In any case, gathering
affordance data on all of Southern California would be an enormous task, but it
might be possible to do with select groups and comparative neighborhoods. My
point is that as useful as the SAGE/GVP analysis of park resources is, it only tells
us what children have access to, and not how they might or might not use those
resources in the pursuit of well-being. That is a matter of capability, or in terms of
the ecological human-environment relationship, affordance.
Does SAGE/GVP observe affordance? Affordance, it will be recalled, is
what the environment offers a person in support of some activity, including not
only qualities of the physical environment but also the support given by the
Figure 11. Green Vision Plan park acres by demographic (Sister, Wolch et al).,
p. 112)
178

presence of other people, and even the effects of the individual’s presence on the
environment. In an affordance measure, certain constraining characteristics and
circumstances of the individual exist in relation to resources of the environment
that the person has reason to want to use. With that in mind, I point out that
SAGE/GVP most certainly describes the resources part of the affordance
equation, by documenting unfair distributions of park resources in ways that
acknowledge the identities of children, and their potential to suffer from a lack of
resources. Yet, even the SAGE/GVP authors recognize that simply increasing
green space and park facilities will not cause children to magically take on active
lives (pp. 19-21). If they did, they would be gathering affordance data. They begin
to get at affordances in their discussion of why some children are unable to make
use of available parks, and how a park’s distance from transit stops affects use,
and the likelihood of traffic accidents in relation to park use.
Planned SAGE/GVP studies include one that will investigate
“environmental justice dimensions of park supply by examining park access of
people living in communities dominated by specific race/ethnic populations, as
well as communities characterized by a diverse socioeconomic class” (p. 115).
Audits that study opportunities in terms of access to particular commodities
would be one step closer to a fully ecological affordance audit.
179

Case: Systematic Pedestrian and Cycling Environment Scan
(SPACES)
Description
There are parallels between the SPACES walkability/bikeability audit and
a number of other audits, and for good reason. The SAGE developers tell us that
they drew their determinants from SPACES, for instance. Others such as the IMI
acknowledge the influence of SPACES, and the IMI attempts to build on it as one
of the field’s early accomplishments. SPACES was developed in 2001, making it
the earliest of the Active Living audits made available to the public by the Robert
Wood Johnson Foundation. At the time that it was developed, SPACES was the
first of its kind (p. 7). By 2006 AL audit developers were still referring to
SPACES as the “most widely known audit tool” (Day, Boarnet et al., 2006,
p.145). Like SAGE, SPACES uses only objective determinants, auditing the built
environment with trained researchers and also “desktop” research from secondary
sources regarding pollution levels and traffic volume and other such data.
Information on destinations are also coded as background GIS data for the audit
(Figure 12). SPACES is an inventory of objective determinants associated with
walking and cycling. In addition, though, the auditor is also asked to provide
subjective impressions about the various elements of the environment, such as its
“attractiveness,” “difficulty for walking,” “attractiveness for cycling,” and
“difficulty for cycling.” The audit responses are scaled in a few different ways.
180

Typologies of things are indicated with four-to-eight-item checklists, for instance.
Some of those checklists observe:
1. Type of buildings/features
2. Transport infrastructure
3. Housing
4. Office
5. Convenience stores
6. Industrial
7. Educational
8. Service
9. Natural features
The presence or absence of physical resources and services, such as
streetlights, destinations, and auto parking, are indicated with a binary scale
(yes/no). Qualities and conditions of paths and roads are indicated with 4- or 5-
point ordinal lists of qualities (from “a lot of bumps, cracks, holes” to “very few
bumps, cracks, holes” e.g.). Quantities of things like trees or driveway crossovers
are given four-or-five-point ordinal scales of quantities (from “one driveway” to
“no driveway,” e.g.) The auditor's manual includes a set of detailed photographs
illustrating the terms used (Figure 13). SPACES also gathers the auditor’s overall
subjective impressions of the study area on three-point Likert scales, either “very
attractive-not attractive at all” or “easy-very difficult.” SPACES is generally
181

assumed to have high reliability, although that has been brought into question
recently. Like other AL audits such as the “Measuring Urban Design Qualities”
(MUDQ), the SPACES development process included an early preliminary list of
determinants based on a literature review, interviews, and a Delphi process with
an expert panel (Pikora, Bull et al., 2002, p. 2). However, McCormack, Masse et
al. took SPACES to task in 2006, in a study of how indices might be developed
from actual SPACES audit information. Those authors used a Rasch model to
determine how well SPACES actually captures the observations that it attempts to
gather. They concluded that the process of developing indices would always be
more complex than the simple summing of scores for a set of determinants, and
that audit developers should rethink how they determine the reliability of their
instruments. It is very important, they argued, to bring index validation measures
(such as a Rasch model) into play right from the start of primary research
(McCormack, Mâsse et al., 2006). The first version of SPACES was given a
rather extensive trial run––sixteen observers audited 1,987 kilometers of urban
roads–and that information was then put to the various reliability tests that AL
audit developers depend upon, such as the kappa statistic and test-retest reliability
measures. From that reliability work, the current instrument was developed. The
use of reliability tests in SPACES seems to have set a reliability standard for
audits, nearly allof which use reliability tests post- audit. The audit developers
note that the SPACES audit is meant to be accompanied by a set of indices to
182

Figure 12. Systematic Pedestrian and Cycling Environment Scan
(SPACES)
183

summarize the findings and public policymakers, although no such indices have
been created (p. 8).
SPACES is firstly an inventory of hardware (speed bumps, street lighting,
e.g.), and of facilities (auto and bike parking, e.g.). It also estimates--in
correspondence model terms--residents' perceptions of the attractiveness and
general safety of the area (tree height, similarity of building designs, and the
degree to which a pedestrian can be observed from neighborhood homes, e.g.).
Finally, it offers the researcher's subjective opinion on the attractiveness of the
area, and also the difficulty it might pose for walking and biking.
SPACES, Satisfaction, Opportunity, Affordance
The value of satisfaction is in what it has given, but the value of opportunity is in
what it could give. From the perspective of urban design policy, one difference
between audit information about pedestrian safety in the satisfactions framework
and pedestrian safety in the opportunities framework is that in the former the
designer develops the neighborhood in ways that please residents by making them
feel safe, which is a satisfying end in itself. Presumably, in feeling safe residents
become more likely to walk, and by implication become healthier. In the case of
the opportunities framework the designer develops the neighborhood in ways that
make residents safe so that they have the means to pursue what ends satisfy them.
The difference between these is that in the satisfactions framework the design
provides an environment that is in itself satisfying, while in the opportunity
184

framework resources are provided for the person to pursue physical well-being in
whatever way they wish. The evaluation of satisfaction is averaged across a
population, and it is presumed that everyone has some equal level of satisfaction
regarding crosswalks, or “interestingness” for example. In order to be fair, the
satisfactions audit must pre-determine what is equally satisfying to everyone,
while the resources audit determines the degree to which people have the
opportunity to pursue diverse satisfactions.
A small number of data points in SPACES regarding pleasant views and
the subjective feeling of safety suggest that a portion of the audit emphasizes
feelings of pleasure and satisfaction. On the other hand, the stockroom-style
checklist inventories in SPACES suggest that it is an audit of resources. Yet
because these are not general resources, or the things that “it is supposed a
rational man wants whatever else he wants” it is unlikely that SPACES
emphasizes opportunity (Rawls, 1993, p. 181). Unlike SAGE/GVP, the SPACES
audit is stand-alone, meaning that it is meant to index the walkability/bikeability
of a single neighborhood, and is not designed for inter-area comparisons.
This audit’s observations of curb types and streetlights and trail slope
angles and even maintenance issues are not meant to evaluate the fairness with
which these things distribute the opportunity for a healthy lifestyle. Rather,
SPACES evaluates amounts of equipment and qualities and attractiveness, to
represent the likelihood that residents will walk and cycle. Presumably, the more
185

of SPACES’ 37 data point items that are found in the neighborhood, the greater
the pleasure residents will take in walking and cycling.
SPACES and the Ecological Model
Neither AL transdisciplinarity nor Gibsonian reciprocity are explicitly
observable with the SPACES audit. The resident’s perception of the environment
is linear, with the environment acting on the individual, never the other way
around. No interconnections between well-beings are considered. SPACES offers
no method for developing an index of quality-of-life from its findings, but if it
did, that index would no doubt point to correct levels of provisioning of attractive
things in the neighborhood, which taken together stimulate walking and biking, in
correspondence model fashion.
The audit assumes that if people do not walk in the neighborhood, then
there are not enough determinants in place. This is a standard inventory style
audit. SPACES assumes it understands the average resident better than she does
herself, which was one of the Rawlsian critiques of utilitarianism. Because it
evaluates in a utilitarian-satisfactions framework, SPACES prioritizes the things
“that a person wants” rather than the fair distribution of the things that make it
possible for them to pursue the choices that they prefer. The satisfactions
approach is time-honored in urban planning and design, in the form of the demand
model for example, but the equality of satisfactions in this audit loses sight of
186

Figure 13. SPACES audit, path material assessment (Pikora, Bull et al., 2002)

187

groups like obesity-prone children, whose reasons for walking and biking may be
quite different than those of the average resident. The data gathered by SPACES
does not address equality of rights either within a community or between
communities, except to average utility. The average of utility is a kind of equality,
but not the sort that considers relative differences in an ecological way.
Case: SOPLAY and SOPARC
Description
Two AL audit instruments that can be described as observing synomorphy
are the System for Observing Play and Leisure Activity in Youth (SOPLAY) and
the System for Observing Play and Recreation in Communities (SOPARC). The
stated intentions of SOPLAY and SOPARC are to provide “…an objective tool
for quantifying physical activity in ‘open’ environments, such as recreational and
leisure settings” (McKenzie, 2006, p. 2).
SOPLAY’s dependent variable of physical activity is modified by three
descriptive categories, “Sedentary,” “Walking,” and “Very Active.” The
instrument gathers data on the context in which behaviors occur along four
environmental/play dimensions:
1. Types of organized play facilities
2. Improvements to areas

188

3. Overlaps between areas
4. Surface materials
Time is a factor in SOPLAY, not only because the behavior sampling
method is time-based, but also in terms of the time of day in relationship to other
activities. Behaviors occur relative to three contextual times, “before school,” “at
lunchtime” and “after school.” Ambient temperature is also recorded. SOPLAY
sub-divides the contextual data still further, into five “conditions” in which
behavior occurs, according to:
1. Whether the area is physically “accessible” in terms of locked gates
or displacement by other groups
2. Whether it is unusable because of weather conditions
3. Whether the area is supervised
4. Whether organized activities are supposed to take place there
5. Whether equipment is provided
There are four sets of data that are gathered to conduct the audit, including
observations of actual behaviors, documentation of facilities and ‘hardware’ of
the environment, and access to facilities. These allow the auditors to evaluate
what opportunities are available to students, and more, the degree to which
students actually take advantage of those opportunities, and why. With these, the
auditors can evaluate physical activity in parks and schoolyards in terms of where
189

activity occurs, who is involved in the activity, and which characteristics of the
environment are used, or not, and why.
The audit gathers data on physical activity behaviors using a "momentary
time sampling" method which allows researchers to gather an "unbiased
estimation of behavior duration," in the sense of documenting occurrences of
behaviors over time in a particular place (Ary & Suen, 1983, p. 143). Because the
technique employs samples of occurrences, rather than documenting each and
every occurrence, researchers are able to gather reliable observations without the
need for unwieldy amounts of data (Figure 14).
The ultimate product of the audit is a set of “mapping variables” with
which researchers can tell, in specific and spatial terms, which children’s behavior
engages the built environment, what activity, when, and why (Figure 15).
SOPLAY/SOPARC, Satisfaction, Opportunity, Affordance
Using SOPLAY, auditors observed physical activity “opportunities”
available to children of certain ages and characteristics. These opportunities
included access to play spaces, play partners, ambient conditions and the time to
play. Because SOPLAY is a “direct observation” tool that observes what children
are actually able to do with their opportunities for play, auditors were also able to
determine some factors that likely prevented students from making use of
opportunities. I would argue that these are the “conversion” factors that Sen
describes as preventing individuals from taking advantage of opportunities (Sen,
190

2000, p. 74). In the case of SOPLAY, these include the children’s not being old
enough to drive; gender disparities; lack of organized activities; and lack of sports
equipment (McKenzie, Marshall et al., 2000, p. 75).
SOPARC’s developer tells us the audit is designed to evaluate either a
single activity space, or to compare activity spaces. This echoes the comparative
equality measure used in SAGE/GVP, and suggests that SOPARK might evaluate
the equality with which actual use is distributed between activity spaces,
neighborhoods, regions.
Figure 14. SOPLAY Momentary Time Sampling Data Form (McKenzie,
2006)

191

SOPARC/SOPLAY and the IMI
The difference between the IMI and SOPARC/SOPLAY in this regard is
the difference between satisfactions and affordance, in the sense that it is the

Figure 15. SOPLAY mapping variables observation sheet (McKenzie,
2006).
192

difference between evaluating the choices that a person prefers and those that she
can actually make use of. Both utility (satisfactions) and capability (affordance)
are consequentialist in that they are focused on the end-state of the persons in
question. Yet satisfaction is not a functioning, but a feeling. Functionings may
have attendant feelings of satisfaction, but the feeling of satisfaction is not the
point, but rather to have the freedom to attain desired functionings. A state of
well-being is what Sen would call an "achieved functioning." Evaluating
capability means assessing the person’s opportunity and agency to pursue a
functioning. A "perfect" affordance equates with a fully achieved functioning.
Because SOPLAY/SOPARC is ecological in its conception of the child
and the environment, it evaluates the “actual opportunities” that a child has to be
physically active (Sen 2000 p. 17). This refers to Sen’s concept of capabilities,
described as “what we are actually able to do or be” and “the freedom to pursue
various lifestyles” (Nussbaum 2000, p. 12; Sen, 2000, p. 75).
SOPLAY/SOPARC and the Ecological Model
SOPLAY observes physical activity behaviors of groups of children in
school settings, and SOPARC gathers the same sorts of data on communities.
Both gather data on what could effectively be called synomorphy, referring to
Barker’s definition, as a sameness of structure between an activity and its context
(Barker, 1968, p. 19). In the SOPLAY/SOPARC audits, the synomorphy is a
necessary relationship between children and schoolyards, and community
193

members and public parks. That relationship tells the story of what people can
actually do or be as a result of their engagement with the environment.
While the SOPLAY/SOPARC audits are not using Barker’s behavior
setting method per se, they represent an important step toward a methodological
description of children and their environments as truly ecological relationships.
The other aspect of Barker’s ecology suggested by this audit is the idea that
behavior settings are interconnected. SOPLAY/SOPARC recognizes that the
various places where children spend their days are interlinked in important and
dynamic ways: In this audit, the time at which the sixth grade class lets out
constrains the use of the playground by second graders. The linkages between
behavior settings is one of the criteria, in fact, that Barker uses to define a
behavior setting (Barker, 1968, pp. 25-26). SOPLAY/ SOPARK does not use
Barker’s method of measuring interlinkages, however, and even Barker admits
that method is highly complex (p. 26), but the relationship between synomorphs is
acknowledged in SOPARC/SOPLAY, which is not the case in other audits.
The SOPLAY/SOPARC author tells us that the audit was designed to
“assess the physical activity participation of a group in a specific setting” (ibid.).
Yet I would argue that the setting in this case includes not only the most general
description of the space (“playground”) but also the times that it may be used, by
whom, the surfaces of the play areas, the equipment and official personnel
involved, programming, as well as the influence of events in other spaces.
194

This audit overlooks the trans-disciplinary data that the AL ecological
model stresses, but on the other hand it is the first discussed here to represent
something of the reciprocality of Gibsonian/Barkerian ecological perception. It
tells us how the things that children actually do might affect other things that they
and other children can do.
Discussion
SOPLAY and SOPARC consider the built environment and physical
activity as a relationship, in ways that other audits do not. It can be said that this
audit observes the agency with which children may take advantage of
opportunities. It should be noted that although SOPLAY and SOPARC raise very
specific issues about what prevents children from making choices, the audit’s
suggestions to correct these are not built environment solutions, but
recommendations for education-oriented programs. On the other hand, the “lack
of organized activities” and “lack of sports equipment”––seemingly management
issues and not overtly related to design––are nonetheless very much issues of
design if we consider the places where that equipment will be used, what is
required of those places, the patterns of activity that occur there, their effect on
the individual, and the degree to which all of this supports an active lifestyle in
children, on or off the schoolyard. The idea of environmental programming is
relevant here, particularly if our definition of programming includes the limiting
effects that privately owned public spaces can have on individual freedoms, such
195

as in the case of outdoor malls and themed entertainment streets in the US, for
example.
If we consider the other two factors that keep children from using play
resources as a way of recognizing those children as individuals––their not being
old enough to drive, gender inequities between boys and girls, poverty, etc.––
these factors speak to the sorts of excluded groups that the IMI developers
suggested should be considered in audits. Because SOPLAY/SOPARK frames the
human-environment relationship is an ecological, synomorphic way, it reveals not
only which groups are excluded, but the highlighting of the conversion factors
tells policymakers what must be taken into account in order to address the
exclusions.
I concur with the Active Living audit developers that the ecological model
is the right choice for built environment-physical activity research. I add, though,
that the conception of the ecological model ought to be revised in ways that
Barker and Gibson describe. Using a revised ecological model to audit the built
environment, I argue, would require an affordance metric, which uses a different
sort of information than is currently used in quality of life measures.
Case: Measuring Urban Design Qualities
Description
MUDQ is an inventory-style audit instrument that gathers data on 27
observations of the built environment. The dependent variable is “walking
196

behavior,” and the 27 items are clustered under five proxy categories of “urban
design qualities” (Ewing, Handy et al., 2006, p. 223)
36
:
1. Imageability
2. Enclosure
3. Human Scale
4. Transparency
5. Complexity
All of the data gathered is objective, and collected by trained researchers
during field observation. No interviews are conducted during the audit, and no
activities are observed. Each researcher is provided with a detailed field manual
that provides graphically enhanced photographs to illustrate each of the urban
design qualities. The 27 data points were developed in an extensive process that
included a literature review of human perceptions of the urban environment,
drawn from across a variety of disciplines including “architecture, landscape
architecture, park planning, environmental psychology, etc. (p. 225). That review
produced a very interesting list of “perceptual qualities” from which the audit
developers would ultimately choose: “[A]daptability, ambiguity, centrality,
clarity, compatibility, comfort, complementar-ity, continuity, contrast, deflection,
depth, distinctiveness, diversity, dominance, expectancy, focality, formality,
identifiability, intelligibility, interest, intimacy, intricacy, meaning, mystery,

36
A preliminary list of eight qualities appears in their conceptual model. The additional qualities are
“Legibility,” “Enclosure,” and “Coherence.”
197

naturalness , novelty, openness, ornateness, prospect, refuge, regularity, rhythm,
richness, sensuousness, singularity, spaciousness, territoriality, texture, unity,
upkeep, variety, visibility, and vividness” (ibid.). The two criteria for inclusion in
the above list were that they were “most frequently discussed” in the literature
and were “shown by empirical evidence to be important to users of urban space”
(ibid.). From the above list a set of nine qualities was culled. Operational
definitions of these were arrived at after consultation with a panel of urban design
experts who were shown walk-through video clips of sample streetscapes. The
material from that consultation was subjected to a battery of statistics and
reliability tests.
37
Eventually the development team arrived at the final five
qualities listed above.
MUDQ, Satisfactions, and the Ecological Model
The conceptual model with which the audit was developed is specifically
designed to consider perception as a “mediating variable” in the urban
inhabitant’s decision and ability to walk (p. 224) (see Figure 16). The MUDQ
developers make no claim for the model being specifically ecological, however
the initial literature review from which their determinants were ultimately drawn

37
Specific steps included: (a) recruitment of a panel of urban design and planning experts; (b) creation of a
library of video clips of streetscapes; (c) selection of video clips; (d) rating of urban design qualities of
streetscapes by the expert panel; (e) measurement of physical features of streetscapes through a content
analysis of video clips; (e) inter-rater reliability testing of physical measurements and urban design quality
ratings; (f) statistical analysis of relationships between physical features and urban design quality ratings; (g)
selection of qualities for operationalization, and (h) development of operational definitions and measurement
protocols for urban design qualities based on statistical relationships (Ewing, Handy et al., 2006, p. 227).
198

included a variety of fields, as noted above. Yet, their panel of experts were from
the planning and urban design fields, and excluded input from the community.
MUDQ is concerned with those perceptual interactions with the
environment that predict physical activity. In the audit’s development process it
was concluded that if a certain level of imageability, enclosure, human scale,
transparency, and complexity, can be established in an environment, then it is
more likely that people will engage in physical activity there. Thus the audit
simply counts the incidences of those five things in the environment.
Unlike SOPLAY/SOPARK, which verified its findings by triangulating
them with heart rate monitoring of children in the same activities, MUDQ’s

Figure 16. Measuring Urban Design Qualities Conceptual
Framework (Ewing, 2005)

199

findings rely upon statistical tests of reliability, which is to say, the likelihood that
if the audit were reproduced it would achieve the same results.
MUDQ’s view of environmental perception is from the inside of the
pedestrian’s mind only, where incoming information is layered with the person’s
previous experiences, and based on this thickened information “reacts” to the
qualities of the built environment. The audit observes impressions of the
environment that come through the five traditional “exteroceptive” organs of
sense, which Gibson traces to Sherrington’s 1906 model of perception (Gibson,
1966, p. 33). While the Sherrington model has not persisted in environmental
psychology, robotics, or neurophysiology, it seems to be a persistent and
unshakable aspect of audits. In Barker’s ecological model, the presence of other
pedestrians represents an interlinking system of activity patterns that are sustained
by unspoken rules that automatically resist deviance, in part because people are so
dependent upon those patterns as the stuff of daily life.. A behavior setting’s
pattern of activity represents a set of options that tell the person entering the scene
what he can and cannot do, as surely as a street light or a posted ordinance.
Because MUDQ represents a mental process of perception rather than one that is
obtained as well as imposed, in Gibson’s terms, the audit references pedestrians
simply as an additive element of the block’s “complexity,” along with the number
of buildings and their accent colors, outdoor dining options and the number of
pieces of public art. It cannot be denied that public art and outdoor dining may
200

very likely add to the complexity and the "walkability" of a city block, but only if
we conceive of the block’s qualities as imposed upon the pedestrian, rather than
obtained.
Using the general audit model from Chapter One as a case study
framework, I have examined a number of built environment audits with an eye to
describing the degree to which they gather affordance data. A few audits can be
said to gather affordance data, and one of those gathers almost nothing but
affordance data, namely SOPLAY/SOPARC. That audit, however, is focused
heavily on behaviors, and less on the human-built environment relationship. In
truth, none of the audits examined here can be said to offer a clear set of
determinants for affordance, although as a group, the audits might suggest a
framework for how we might go about developing an affordance audit.
Affordance allows the auditor to describe quality of life with information
on what the environment supports in terms of the individual’s physical well-
being. This suggests a different kind of quality of life index than the more
common additive indices that represent counts of the things that an environment is
supposed to have in order to be satisfying.
201

Affordance calls for an approach to urban design intervention that
prioritizes equality of human well-being in the design of the built environment.
Like capability, affordance demands that we consider well-being, the
environment, and evaluative audits in a new and even radical way.
For what we gain with affordance we also surrender too, however. The
context-specific nature of affordance evaluation lessens the ability to easily
generalize the role of physical activity determinants in the lives of heterogeneous
persons. By way of recompense, though, affordance offers a new conception of
places in which the competencies of the individual and her social and physical

Figure 17. Tai Chi in the picnic area. (Photo by author)

202

environments might intensify exponentially with their increased interconnection
and interrelationship.
“Spot the Affordance”
Consider a tai chi teacher who has been asked to drive out to the suburbs
to tutor a couple in the tai chi form. The teacher suggests that they meet at the
local park at 7 a.m., while the park is still empty. Arriving early, the teacher takes
a moment to find a good place to hold his class. Spotting the picnic area, with its
flat grassy floor and open space, he decides that this spot “affords” the kinds of
work that he and his students will be doing that day. It may seem unlikely that the
park’s architect designed that spot for “picnic/tai chi,” but it would seem likely
that in thinking about what the park needs to support in terms of human activity, a
plural conception of park use might be part of the park’s conception (Figure 17).
If square footage and funding were limited, and if potential use of the park
exceeded existing space, it seems likely that a clever designer might think of
multiple uses for the space, and park administrators might calculate the intensity
of park use in a similar fashion.
The teacher’s criteria might have included “flat open space,” “smooth,
slightly soft surface underfoot,” “quiet,” “fresh air,” “shade,” and other things that
are also desirable in a picnic area. That assessment of affordance may be
intentional or pre-conscious, but it is the same in either case. Affordance is not
something that we can see in the physical landscape, because it is a relationship
203

between an individual and some aspect of the environment. Even observing the
use of the environment, which tells us that affordance is taking place, cannot
necessarily also describe affordance itself, unless we know what it is the
individual needs from the environment: We cannot know that a man wandering in
the park at 7 a.m. is looking for tai chi space unless we ask him.
One of the most delightful examples of affordance is what I call “running
and screaming downhill.” Both Barker and Kevin Lynch speak of this
phenomenon: Children, coming upon a grassy hill, almost inevitably run down it,
screaming in exhilaration. This author’s experience observing park use around
Los Angeles corroborates that “running and screaming” in indeed ubiquitous
(Figure 18). There is something in the slope of the hill, the softness of the surface,

Figure 18. Grassy Hills afford “Running Down and Screaming”
(photo by author)
204

the lack of obstacles, the size of the child, and perhaps even the child’s
developmental stage, that makes a grassy hill afford a child’s “running and
screaming downhill.” There are examples of grassy park hills that end in bike
paths or deep water, or both, where “biking” and “fishing” affordances would be
rated as high, but “running and screaming downhill” would be low. On the other
hand, we cannot know if fishing and biking are along the neighborhood’s path to
well-being, unless they say so. Unlike audits for satisfaction or opportunity,
affordance takes the individual into account. That is in fact the whole point of
affordance.
Equality of satisfaction addresses the differences between people by
averaging their happiness; equality of opportunity addresses difference by
changing the evaluative focus to an equal set of resources with which individuals
may pursue what ends they wish. Affordance not only implies, but also requires
that we take the differences between persons into account. Of the audits examined
here, only SAGE and SOPARC/SOPLAY specifically factor identity into their
metrics. As I noted before, in considering the individual and her particular context
we surrender generalizability to some degree in auditing the built environment,
because no set of physical activity determinants can possibly support the activity
needs of everyone.
Also, as SAGE points out (the only audit to do so), neighborhoods
change, in demographic terms, but also in a kind of overall “character” that
205

influences use. Many of the audit developers here note the need to consider the
mediating variables that neighborhood character implies, but few take it up.
Jennifer Wolch, one of the SAGE authors, points up one of the reasons for that
when she notes that even though it is essential to understand what kind of park
resources particular neighborhoods need, policymakers and designers must not
fall into the trap of essentializing, in the sense that “Latinos want soccer fields,”
or other such stereotypes (Wolch, 2004, p. 65). My point is that although we may
be fairly certain that children under a certain age run-and-scream-downhill
whenever they get the chance, we cannot be so certain about biking and fishing
and other physical activities, without considering a variety of mediating variables,
including cultural ones, institutional ones, and those of the physical landscape. In
the affordance conception, identity signals the need for public empowerment and
participation, which is explicit in the work of the moral philosophers such as
Rawls and Sen, and amongst planning theorists such as Lichfield, Harper and
Stein, and Lynch. As researchers undertake to audit the built environment, it
should be considered that in the same way that identity can serve as a source of
political as well as psychological meaning, so can an identity-based ecological
measure of affordance inform policies in ways that recognize difference, and in
ways that could translate directly to the everyday lives of urban residents.
38

38
The question of identity politics and the built environment suggests a notion of what could speculatively be
called a ‘radical affordance.” Two respected theoretical statements on what we might call the “city of
206

Is the Affordance Metric Feasible?
It is clear to me that one of the limitations of any affordance audit is that it
must be individual-specific. Its data might made generalizable to the degree that
the individual’s identity is generalizable, but it would then have to be assumed
that the other side of the ratio––the built environment––is equally general.
Because my draft affordance audit emphasizes the role of the elderly in
facilitating at-risk children’s park use, if it were used to evaluate any parks for
any user its would likely overlook entire swaths of the population. This problem
parallels the critiques of the capability approach that claim its emphasis on the
well-being of one group of individuals could unfairly overlook another, and puts
the evaluator in the difficult position of choosing between which group’s best
interest equals the public interest.
39

It would be an interesting project to develop an affordance audit general
enough to consider any human being in any public space, for example. The
problem is that generalizing in this way defeats the underlying point of deploying

difference” are Madanipour (1996) and Sandercock (1998). However, neither of those discusses actual
auditing of the built environment in terms of difference.

39
As I mentioned in Chapter 2, Sen counters this charge with the argument that such comparative judgments
can be justly made using an efficiency criteria for comparing which group of individuals ought to be the focus
of development (See Sen, 1992, pp. 141-147). Affirmative action programs would be an example of how
efficiency is used in this way. The question is somewhat trickier in built environment terms, and in this regard
I am grateful to Dr. Banerjee for introducing me to Walzer’s concept of “open-minded space,” which parallels
Sen’s argument, albeit more in terms of public space (Walzer, 1995). Sen does not argue that such
comparative efficiencies ought to be monolithic and intractable, but rather that comparing the needs of groups
in this way makes values explicit and sensitizes us to the justice of the particular situations involved.
Similarly, Walzer is not arguing that all public space ought to be “open-minded” in the sense that its physical
features are temporary, but rather that in a liberal democracy the needs of different groups take precedence at
different times, and planning for public space ought to keep this in mind.
207

capability as affordance: Evaluations of quality of life tend to generalize away the
difficult question of whose life is in question, and what they ought to be able to do
with the quality of it.
It would also be interesting to conduct concurrent affordance audits for
groups that compete for the same resources. Yet difficult questions remain about
how to aggregate and compare the well-beings of different groups into an overall
quality of life rating. In any case, that matter cannot be taken up here. I have
attempted to show the many conceptual and methodological parallels between the
capability approach, the ecological model of perception, and affordance. More, I
have argued that the two scales of capability/affordance are interlinked: Sen’s
capability approach evaluates quality of life at the scale of entire countries, and
affordance evaluates the negotiations between one person and her environment.
Clearly there will be hurdles, theoretical and practical, in applying the affordance
idea to AL research, and to developing quality of life measures for the built
environment. The correspondence model is persistent and ubiquitous because our
instinct tells us it is true, and it partly is. It is simply not all of the truth. We
humans are almost never entirely passive, we are always exploring investigating,
even sitting quietly on a coffee house step watching the world go by, we are
obtaining something of that world. At least since Descartes we have taken the
distinction between subject and object––individual and environment, in this case–
–for granted. Affordance is a conceptual bridge over that fundamental conceptual
208

split. Affordance has its instinctive pull too, although getting a feel for it will take
some effort and imagination.
An additive quality-of-life index is a simple and time-honored tool. Is an
affordance metric equally simple, flexible, and expressive? The simplicity and
power of an additive index is that it expresses amounts of things in a way that can
be a powerful policy tool. It tells its truth plainly. An affordance index would tell
us how much or how little of a relationship exists, which is a considerably more
complex and potentially cumbersome idea.
The notion of affordance is a powerful one. The meaning of things is
discerned all the time, as we investigate the environments in which we live, and
those investigations become part of the meaning of our lives and our places. Our
moment by moment negotiations with the built environment, with other people,
with gravity and weather, with fears and threats, and with the effects we ourselves
have in the world, can all be described in terms of our individual relationships to
our places. With the concept of affordance, built environment audits might
transcend checklists of qualities and things, to become documents that record how
we actually live in the world, and how the world actually lives in us.
Disparities and Fairness
If urban design adjusts built form in response to a change in the character
of a neighborhood, what is to be done when the character of that neighborhood
changes yet again? Built form has a much longer half-life than human
209

communities. Michael Walzer’s “open-minded space” is space not locked into
one use or even one idea of use, but can contain––or as Lynch might say,
encourage, contain, and surprise us with
40
––a plurality of uses and identities. If
urban design policies could be conceived in open-minded terms it might go a long
way toward addressing some of the issues raised above.
Also, since any community can be sub-divided into an infinite number of
demographic groups, we must ask if it is fair to develop the built environment
with one group in mind, and not another. Where is the fairness in focusing scarce
city resources on one sub-group over another? Is developing with the capability of
one group in mind not in effect depriving another group? These are questions that
remain to be addressed in the affordance approach.
There is also the practical issue of whether the process of gathering data
on affordance would be too cumbersome to be practical. Barker’s original
behavior settings survey is methodologically massive, requiring at least a year to
gather data, and another year to study dozens and dozens of behavior settings and
their interlinkages. The potential for that method remains spectacular, but
unwieldy and expensive. Bechtel, Wicker and others have successfully

40
“An environment for growth would be more exposed, accessible, and diverse, more open both physically
and psychologically, more responsive to individual initiative and control. It would invite exploration and
reward it; it would encourage manipulation, renovation, and self-initiated changes of many kinds. It would
contain surprises and novel experiences, challenges to cognition and action. It would not be the most efficient
and safe environment. Nor would it offer maximum stability and security. It would certainly not be extremely
comfortable, nor even very beautiful, unless we look for beauty in the process of interaction rather than in
static form” (Lynch, 1990, pp. 426-427).
210

abbreviated the Barker method, and one can see vestiges of it the work of Zeisel
and others.
Lynch was strongly influenced by Barker’s research. He argued in Good
City Form that describing the city required Barker’s behavior settings, because
they view activity and people as one thing rather than as separate.That vision is of
a human-environment interrelationship in which the built environment is assessed
less in terms of level-of-service, and more in terms of relationships between the
city and human activity (Lynch, 1984, p. 230).
Were there time and space enough, I could say a great deal more about the
parallels between Lynch, Barker and Sen, but that will have to wait for another
time.
Conclusions
The six audits above were each examined with the concepts of the general
audit model. The general audit model seems to have helped to categories and
clarify some of the underlying ethical assumptions within the audits, evidenced by
the comparative differences and similarities that were noted between the different
cases. The three categories of satisfactions, opportunity, and affordance were able
to capture and describe all of the data from the audits. This suggests that the
“three goods” of satisfactions, opportunity, and affordance describe at least a
majority of the ethical frameworks involved in audits. On the other hand, as these
are case studies, it is only possible to say for certain what has been observed
211

within the cases themselves. Yet the ability of the general audit model to describe
the six cases in terms of their underlying ethical assumptions argues that this
theoretical research may have practical use in planning urban design policy.
This theoretical and conceptual work is meant to inform future empirical
studies. Audits should inform indices of quality of life, however as is argued by
McCormack, Massey et al., audit indices cannot simply be a summing of the
numerical presence of physical activity determinants.
41
Rather, the statistical
reliability of the indices must be established during the same process that
validates the audit’s determinants. This approach seems wise, and the Rauch
model employed by McCormack, Massey et al. offers a way to undertake such
validating of indices during the determinant development phase.
The AL primary research model ought to fold in the concept of Gibsonian
reciprocity. Reciprocity is information that moves in two directions rather than
only one, one is imposed on the perceiver from outside and the other is obtained
by the perceiver through inquiry. It not only conceives of the human as a passive
receiver upon whom inputs are imposed, but also as an active agent of the
environment. The introduction of two additional axes across through the various
domains of the AL model, in two directions, would make it easier to describe
affordance and capability in the built environment.

41
We can note too that SAGE/GVP creates additive indices in the manner that McCormack, Massey et al.
warned against.
212

The models recognize no hard separation between kinds of determinants––
transportation overlapping with medicine, overlapping again with policy, etc. I
argue it should also see through methodological assumptions about barriers
between physical well-being and other kinds. We interact with the environment at
multiple levels at once, and determinants are interconnected at a variety of levels
at once. So should we consider the categories of well-being, as a way of
reconceiving the goals of audits. The goal is the total development of what Sen
calls capability. The child’s emotional and material and other well-beings are all
interconnected, and research methods must consider the categories of built
environment performance criteria in a multi-dimensional way, and reciprocally
interconnected.
Sallis, Cervero et al. offer their transdisciplinary ecological model in
response to what they describe as shortcomings in built environment-physical
activity research to date. They note frequently that there is a strong need for
seemingly disparate research fields such as leisure research and transportation
research to work in an interdisciplinary way (p. 314). That kind of effort, it is
argued, can lead not only to more insightful research data, but also to a combining
of the interests of advocacy groups, and a compounding of benefits to the
neighborhood resident. Too often, they tell us, non-ecologically focused research
observes behavior but without considering the human purposes behind that
behavior (p. 315). I have come to believe that they are right about this, that it is
213

easy to overlook the target in built environment-physical activity research. Yet I
would add that an additional reason that the target is overlooked is because
research investigates the behavior of residents for the purposes of triggering
physical activity, and overlook the triggers that already exist in the individual’s
dynamic relationship with the environment. If the AL ecological model could be
expanded to include the ecological ideas discussed in this study, the pursuit of
well-being as a human purposive activity could bring a powerful new dimension
to the conception of the human-environment relationship.
Again, I in no way dismiss the Sallis, Cervero et al. model, but simply
argue for expanding its scope to embrace a fuller ecological conception. Those
authors recognize that the field has focused mostly on “middle class, mostly white
adults living in urban and suburban settings,” while the obesity epidemic threatens
mostly low-income families in inner city areas (p. 316). They also recognize that
the economic impacts and costs of policy changes must be considered in trans-
disciplinary, multi-level terms of “real estate, transportation, electronic
entertainment industries, and various government agencies” (ibid.). The
ecological research agenda must be expanded “to encompass the environmental
factors, policies, and programs shown to promote active living” (p. 317). They
call for the development of new methodologies, pointing out that although
“survey research dominates” the field, “there is a recent trend to include
interpretive, qualitative, and case studies” (p. 310).
214

Our concern should be that new AL research will be conducted in the
manner of the old demand model, perhaps augmented with supply-side
information about what mediating variables inform the average person’s
preference. If the study of ecological psychology tells us anything it is that the
decision to engage the environment cannot be described as either supply or
demand, nor as subject or object, nor as desire and commodity, but as something
in between those. Making the adjustment to that idea begins with the AL
conceptual model.
I concur with AL researchers they we cannot blithely overlook the
powerful cultural-community impacts of recreation they argue (ibid.). I agree that
cost-benefit models are currently limited in describing the relationships between
exercise and the daily lives of residents. Sallis, Cervero et al. incisively note that
cost-benefit models can be insensitive to the tradeoff decisions that people make
in deciding how to use their time and energy (p. 312). Yet we must not forget that
merely expanding the definition of “cost” and “benefit” to include non-
commodity goods does not necessarily expand the cost-benefit continuum beyond
traditional conceptions of rationality. Human values––such as commitment to an
idea or a feeling––can inform choices in ways that defy the traditional view of
economic rationality. If researchers do not understand the actual opportunities that
people have, and how opportunities interact as interlinked constraints, they risk
215

overlooking the powerful role that values and commitment play in decisions about
physical activity.
Ecological AL models ought to describe a reciprocal human-environment
relationship, such as the one that is so characteristic of both Barker and Gibson.
Without the concept of reciprocity, the AL ecological model could emphasize
how the environment mediates people’s decisions to be active, but not how those
decisions affect their actual ability to be well. Barker describes behavior settings
as constraining decision-making and access in powerful ways, and not simply as
‘the places where things happen.’ Like behavior settings, affordances are neither
entirely on the supply side nor on the demand side, but something that is both, and
neither.
Like the behavior setting, affordance shows us that we choose to engage
the environment within a critical domain of what we have the real opportunity to
choose––according to what is afforded. For example, in terms of the role of
leisure and educational programming, Sallis, Cervero et al. astutely suggest that
the physical spaces of parks ought to be seen not simply as equipment for activity,
but also as “a unique amenity for promoting physical activity” (p. 310). That
expanded view of trans-disciplinary partnerships between leisure and other fields
would benefit from a reciprocal understanding of why residents do and do not
exercise in particular places and times.
216

A reciprocal understanding of physical activity education and
programming would not only inform residents about what they ought to do, and
not only discern what residents want to do in parks, but would also inform park
design/education partnerships about the what residents are motivated to actually
be and do. How urban residents use their bodies and minds and emotions and
social networks and material resources and political opportunity is not only a
matter of what everyone ought to know about health, or what equipment is
available to anyone. The researcher’s understanding of affordance must be
informed by cultural and institutional practice, for example, and gender roles, and
how individuals understand their world and their place in it. Particularly in the
high-stakes context of the childhood obesity epidemic, design policy cannot
afford to exclude such concerns from what it means by ‘quality of life.’
Sallis, Cervero et al. suggest that researchers are now attempting to
ecologically understand not only what physical activities children engage in, but
also how they do so and why. That could represent a significant new direction in
AL research, if it allows researchers to conceive of the human-environment
relationship as more than simply supply or demand. As Chemero was earlier cited
as saying, a glass can only be said to be ‘fragile’ if conditions exist in which the
glass might be broken; that is neither a question of the demand for non-fragile
glasses nor an assessment of the glass-shattering environment, but a ratio of the
glass/environment relationship. Similarly, we can only say that the environment
217

affords the individual well-being if we consider the environment in direct relation
to the individual, rather than as service delivery.
Again, my purpose here is not to dismiss the Active Living ecological
model, but simply to augment it with the idea of ecological reciprocity between
the individual and the environment. Bringing reciprocity into the Active Living
ecological model begins to suggest an expanded vision of urban design in
collaboration with a variety of disciplines, and more importantly, in collaboration
with residents. I am arguing that this is not only a trans-disciplinary matter,
however, but also a question of the reciprocal relationship between residents and
their environments.
42

42
The term “environment” here connotes not only the physical environment, but in ecological terms also
includes the policy and domestic and other levels of the trans-disciplinary model. Augmenting that model
with the concept of reciprocity does not necessarily add a new level, but rather suggests that the
environmental influences on the decision to be active is not simply caused by the environment, but also how
the various levels of constraints are changed by the presence of individuals and their activities. Focusing on
trans-disciplinarity alone overlooks how one human activity affects another, and how one constraint affects
another, and how the inability to make use of one resource renders some other resource unusable.
218

Bibliography
Active Living Research. St. Louis environment and physical activity instrument.
Robert Wood Johnson Foundation. Retrieved 1/29/07, 2006, from
www.activeliving.org.

Active Living Research. Tools and measures. Robert Wood Johnson Foundation.
Retrieved 1/29/07, 2006, from http://www.activelivingresearch.org/
index.php/Tools_and_Measures/312.

Ainsworth, B. E. Environmental Supports for Physical Activity (Social and
Physical Environment Survey). Prevention Research Center, Norman J.
Arnold School of Public Health, University of South Carolina. Retrieved
1/22/06, 2007, from http://prevention.sph.sc.edu/tools/Env Supports for
PA.pdf.

Alexander, C. (1987). A New Theory of Urban Design. New York: Oxford
University Press.

Alexander, C., Ishikawa, S., & Silverstein, M. (1977). A pattern language:
Towns, buildings, construction. New York: Oxford University Press.

Alkire, S. (2002). Valuing freedoms : Sen's capability approach and poverty
reduction. Oxford; New York: Oxford University Press.

Angrosino, M. V., & Mays de Perez, K. A. (2000). Rethinking observation. In
Denzin, N. K. & Lincoln, Y. S. (Ed.), Handbook of Qualitative Research.
pp. 673-702. Thousand Oaks, CA: Sage Publications.

AOA. AOA Fact Sheets: Obesity in youth. American Obesity Association.
Retrieved 4/07, 2007, from
http://www.obesity.org/subs/fastfacts/obesity_youth.shtml.

Archibugi, F. (2006). Pitfalls in Planning and Plan-evaluation. In Alexander, E. R.
(Ed.), Evaluation in planning: Evolution and prospects. Aldershot,
England ; Burlington, VT: Ashgate Pub.

Aristotle. (1962). Nicomachean ethics. Indianapolis, IN: Bobbs-Merrill.

219

Ary, D., & Suen, H. K. (1983). The use of momentary time sampling to assess
both frequency and duration of behavior. Journal of Psychopathology and
Behavioral Assessment 5(2), 143-150.

Banerjee,T., & Baer, W. C. (1984). Beyond the neighborhood unit : Residential
environments and public policy. New York: Plenum Press.

Barker, R. G. (1968). Ecological psychology: Concepts and methods for studying
the environment of human behavior. Stanford, CA: Stanford University
Press.

Barker, R. G. (1978). Habitats, environments, and human behavior (1st). San
Francisco, CA: Jossey-Bass.

Barker, R. G., & Wright, H. F. (1971). Midwest and its children: The
psychological ecology of an American town. Hamden, CN: Archon Books.

Bauman, L. J, Silver, E. J., & Stein, R. E.K. (2006, April). Cumulative social
disadvantage and child health. Pediatrics, 117(4), 1321-1328.

Bechtel, R. B. (1977). Enclosing behavior. Stroudsburg, PA: Dowden,
Hutchinson & Ross.

Bedimo-Rung, A. L, Mowen, A. J., & Cohen, D. A. (2005). The significance of
parks to physical activity and public health: A conceptual model."
American Journal of Preventive Medicine, 28(2S2), 159-168.

Bell, S. (2002). Spatial cognition and scale: A child's perspective. Journal of
Environmental Psychology, 22, 9-27.

Boarnet, M. G, Anderson, C. L, Day, K., McMillan, T., & Alfonso, M. (2005).
Evaluation of the California Safe Routes to School Legislation. American
Journal of Preventive Medicine, 28(2S2), 134-40.

Booth, K. M., Pinkston, M. M., & Poston, C., & Walker, S. (2005). Obesity and
the built environment. Journal of the American Dietetic Association,
105(S5), 110-117.

Borders, T. F, Rohrer, J. E, & Cardelli, K. M. (2006). Gender-specific disparities
in obesity." Journal of Community Health, 31(1), 57-68.

220

Boslaugh, S. E., Luke, D. A., Brownson, R. C, Naleid, K. S., & Kreuter, M. W.
(2004). Perceptions of neighborhood environment for physical activity: Is
it 'Who you are or where you live?" Journal of Urban Health: Bulletin of
the New York Academy of Medicine, 81(4).

Bradley, J., & Schaefer, K. C. (1998). The uses and misuses of data and models :
The mathematization of the human sciences. Thousand Oaks, CA: Sage
Publications.

Brain, D. (2002). From good neighborhoods to the sustainable city: Social science
and the social agenda of the new urbanism. International Regional Science
Review, 28(2): 217-238.

Brownson, R. C, Chang, J. J., Eyler, A. A, Ainsworth, B. E, Kirtland, K. A,
Saelens, B. E., & Sallis, J. F. (2004). Measuring the environment for
friendliness toward physical activity: A comparison of the reliability of 3
questionaires. American Journal of Public Health, 94(3), 473-483.

Byrne, J., W., Jennifer, Swift, J., & Ryan, C. SAGE Systematic Audit of Green-
Space Environments. USC Center for Sustainable Cities. Retrieved
5/18/08, 2008, from www.greenvisionplan.net.

Campbell, H., & Marshall, R. (2002). Utilitarianism's bad breath? A Re-
evaluation of the public interest justification for planning." Planning
Theory, 1(2), 163-187.

Cervero, R., & Duncan, M. (2003). Walking, bicycling, and urban landscapes:
Evidence from the San Francisco bay. American Journal of Public Health,
93(9).

Chawla, L., & Heft, H. (2002). Children's Competence and the Ecology of
Communities: A functional approach to the evaluation of participation.
Journal of Environmental Psycholog, 22, 201-216.

Chemero, A., Klein, Colin, & Cordeiro, W. (2003). Events as changes in the
layout of affordances. Ecological Psychology, 15(1), 19-28.

Clark, C., & Uzzell, D. L (2002). The affordance of the home, neighbourhood,
school and town centre for adolescents. Journal of Environmental
Psychology, 22, 95-108.

221

Clitheroe, H. C., Stokols, D., & Zmuidzinas, M. (1998). Conceptualizing the
context of environment and behavior. Journal of Environmental
Psychology, 18, 103-112.

Charter for the New Urbanism (CNU). Congress for the new urbanism. Retrieved
Oct. 14, 2006, from
http://www.cnu.org/aboutcnu/index.cfm?formaction=charter.

Cohen, D., Finch, B.K., Bower, A., & Sastry, N. (2006). Collective efficacy and
obesity: The potential influence of social factors on health. Social Science
and Medicine, 62(3), 769-778.

Craig, C. L., Brownson, R. C., Cragg, S. E., & Dunn, A. L. (2002). Exploring the
effect of the environment on physical activity: A study examining walking
to work. American Journal of Preventive Medicine, 23(2, Supplement 1),
36-43.

Cuthbert, A. R. (2007). Urban design: Requium for an era––review and critique of
the last 50 years. Urban Design International, 12.

Day, K.. Audit tools for research on built environment features ties to active
living. Robert Wood Johnson Foundation. Retrieved 4/26/08, 2008, from
www.activelivingresearch.org/files/AuditToolsComparisonTable.pdf.

Day, K., Boarnet, M., & Alfonzo, M.. Irvine Minnesota Inventory for Observation
of Physical Environment Features Linked to Physical Environment: Code
Book 1.
Active Living. Retrieved 2/12/07, 2007, from www.activeliving.org.

Day, K., Boarnet, M., Alfonzo, Mariela and Forsyth, Ann (2006). The Irvine-
Minnesota Inventory to measure built environments: Development.
American Journal of Preventive Medicine, 30(2), 144-152.

Derr, V. (2002). Children's sense of place in Northern New Mexico. Journal of
Environmental Psychology, 22, 125-137.

Dixey, R. (1998). Improvements in child pedestrian safety: Have they been gained
at the expense of other health goals?" Health Education Journal, 57, 60-
69.

222

Drukker, M., & Van Os, J. (2003). Mediators of neighborhood socioeconomic
deprivation and quality of life. Social Psychiatry and Psychiatric
Epidemiology, 38, 698-706.

Duany, A. Plater-Zyberk, E., & Alminana, R. (2003). The new civic art :
Elements of town planning. New York: Rizzoli.

Ebbeling, C. B, Pawlak, D. B., & Ludwig, D. S. (2002). Childhood obesity:
Public-health crisis, common sense cure. The Lancet, 360(9331), 473-482.

Ellen, I. G., Dillman, K.,& Mijanovich, T. (2001). Neighborhood effects on
health: Exploring the links and assessing the evidence. Journal of Urban
Affairs, 23, 391-408.

Ellen, I. G.m, & Turner, M. A. (1997). Does neighborhood matter? Assessing
recent evidence. Housing Policy Debate, 8, 833-866.

Emery, J., Crump, C., & Bors, P. (2001). Reliability and validity of two
instruments designed to assess the walking and suitability of sidewalks
and roads. American Journal of Health Promotion, 18(1).

Evans, G. W, Lercher, P., & Kofler, W. W (2002). Crowding and children's
mental health: The role of house type. Journal of Environmental
Psychology, 22, 221-231.

Ewing, R., Handy, S., Brownson, R. C., Clemente, O., & Winston, E. (2006).
Identifying and measuring urban design qualities related to walkability.
Journal of Physical Activity and Health, 3(Suppl 1), S223-S240.

Ewing, R., & Cervero, R. (2001). Travel and the Built Environment-Synthesis."
Transportation Research Record, 1780 (2001), 87-122.

Ewing, R., & Clemente, O. Measuring urban design qualities related to
walkability: Final report. Active Living Research Program of the Robert
Wood Johnson Foundation Retrieved 4/25/08, 2008,
http://www.activelivingresearch.org/node/10636.

Ewing, R., Schmid, T., Killingsworth, R., Zlot, A., & Raudenbush, S. (2008).
Relationship between urban sprawl and physical activity, obesity, and
morbidity. In Marzluff, J.M., Shulenberger, E., et al (Ed.), Urban
Ecology: An international perspective on the interaction between humans
and nature, ( pp.567-582). Springer US
223

Fleischaker, G. R. (1984). The traditional model for perception and theory of
knowledge: Its metaphor and two recent alternatives. Behavioral
Science,29(1), 40-50

Flournoy, R., &Yen, I. (2007). The influence of community factors on health: An
annotated bibliography. PolicyLink and the California Endowment.
Retrieved 5/6/07, 2007, from http://www.policylink.org/CHB/default.html.

Frank, L. D., Schmid, T. L., & Sallis, J. F., Chapman, J., & Saelens, B. E. (2005).
Linking objectively measured physical activity with objectively measured
urban form findings from SMARTRAQ. American Journal of Preventive
Medicine, 2005(28 2S2),117-125.

Gans, H. J. (2002). The sociology of space: A use-centered view. City &
Community, 1(4): 325-335.

Georgiou, D., Carspecken, P. F., & Willems, E. P. (1996). An expansion of Roger
Barker's behavior setting survey for an ethno-ecological approach to
person-environment Interactions. Journal of Environmental Psychology,
16, 319-333.

Gibson, J. J. (1966). The senses considered as perceptualsSystems. In. (xiv, 335
p.). Boston, MA: Houghton Mifflin.

Gibson, J. J. Purple perils: A note on substances, surfaces, places, objects, events.
Human-web interaction cycle. Retrieved 6/8/2007, 2007, from
http://huwi.org/gibson/ index.php. http://huwi.org/gibson/substances.php.

Gibson, J. J. (1979b). The ecological approach to visual perception. Boston, MA:
Houghton Mifflin.

Gieryn, T. (2002). Give place a chance: Reply to gans. City & Community, 1(4),
341-343.

Godbey, G. C, Caldwell, L. L., Floyd, M., & Payne, L. L (2004). Contributions of
leisure studies and recreation and park management research to the active
living agenda." American Journal of Preventive Medicine, 28(2S2).

Harper, T. L., & Stein, S. M. (2006). Dialogical planning in a fragmented society:
Critically liberal, pragmatic, incremental. New Brunswick, NJ: Center for
Urban Policy Research.
224

Hart, R. (1997). Children's participation. New York: UNICEF/Earthscan
Publications.

Hart, R. (Ed.). (1979). Children's experience of place. New York, Knopf.

Hayslip, B., & Kaminski, P. (2005). Grandparents raising their grandchildren: A
review of the literature and suggestions for practice. The Gerontologist,
45(2), 262-269.

Heft, H. (1988a, Fall). Affordances of children's environments: A functional
approach to environmental description. Children's Environments
Quarterly, 5(3), 29-37.

Heft, H. (1988b, Fall). Affordances of children's environments: A functional
approach to environmental design. Children's Environments Quarterly,
5(3), 29-37.

Heft, H. (2001). Ecological psychology in context : James Gibson, Roger Barker,
and the legacy of William James's radical empiricism. Mahwah, NJ: L.
Erlbaum.

Hoehner, C. M., Brennan Ramirez, L. K, Elliott, M. B., Handy, S. L., &
Brownson, R. C (2005). Perceived and objective environmental measures
and physical activity among urban adults. American Journal of Preventive
Medicine, 28(2S2), 105-116.

Hoover, K. R., & Donovan, T. (2003). The elements of social scientific thinking
(8th). Belmont, CA: Thomson/Wadsworth.

House, E. R. (1980). Evaluating with aalidity. London: Sage Publications.

Howard, E., & Osborn, F. J.(1965). Garden cities of to-morrow (6th
printing,1976). Cambridge, MA: M.I.T. Press.

HUD. (1996). Beyond shelter: Building communities of opportunity––the United
States report for Habitat II. US Department of Housing and Urban
Development:

HUD/CNU. (2004). Principles for Inner City Neighborhood Design. Department
of Housing and Urban Development, from www.hud.gov.

225

Humpel, N., Owen, N., & Leslie, E. (2002). Environmental factors associated
with adults’ participation in physical activity: A review. American Journal
of Preventive Medicine, 22(3).

Inam, A. (2002). Meaningful urban design: Teleological/catalytic/relevant.
Journal of Urban Design, 7(1), 35-58.

Jacobs, A., & Appleyard, D. (1987, Winter). Toward an urban design manifesto.
APA Journal, 112-120.

Jacobs, A. B. (1985). Looking at cities. Cambridge, MA: Harvard University
Press.

Jacobs, A. B. (1993). Great streets. Cambridge, MA: MIT Press.

Jacobs, J. (2002). The death and life of great American cities. (2002). New York:
Random House.

King, A. C, Stokols, D, Talen, E, Brassington, G. S., & Killingsworth, R. (2002).
Theoretical approaches to the promotion of physical activity: Forging a
transdisciplinary paradigm. American Journal of Preventive Medicine,
23(2S), 15-25.

King, A. C, Bauman, A., & Abrams, D. B. (2002). Forging transdisciplinary
bridges to meet the physical inactivity challenge in the 21st century.
American Journal of Preventive Medicine, 24(2S), 104-106.

Kirtland, K. K., Porter, D.E, Addy, C.L., Neet, M. J., Williams, J. E., & Sharpe, P.
A.
(2003). Environmental measures of physical activity supports: Perception
versus reality. American Journal of Preventive Medicine, 24(4), 323-331.

Klosterman, Richard E. (1978). "Foundations of Normative Planning." Journal of
the American Institute of Planners 44(1): 37-46.

Korpela, K., Kytta, M., & Hartig, T. (2002). Restorative experience, self-
regulation, and children's place preferences. Journal of Environmental
Psychology, 22, 387-398.

Kumanyika, S. K. (2005). Obesity and physical activity among minority
populations. diversity workshop. Harvard School of Public Health.

226

Kumanyika, S. K., Jeffery, R. W., Morabia, A, Ritenbaugh, C., & Antipatis, V. J.
(2002). Obesity prevention: The case for action. International Journal of
Obesity, 26, 425-436.

Lee, R. E., Booth, K. M, Reese-Smith, J. Y, Regan, Gail, & Howard, H. H.
(2005). The physical activity resource assessment (PARA) instrument:
Evaluating features, amenities and incivilities of physical activity
resources in urban neighborhoods. International Journal of Behavioral
Nutrition and Physical Activity, 2(13).

Lichfield, N. (1971). Cost-benefit analysis in planning: A critique of the Roskill
commission. Regional Studies, 5(3), 157-183.

Lombardo, T. J. (1987). The reciprocity of perceiver and environment : The
evolution of James J. Gibson's ecological psychology. Hillsdale, NJ: L.
Erlbaum Associates.

Lynch, K. (1960). The image of the city. Cambridge, MA: Technology Press.

Lynch, K. (1984). Good city form (11th paperback edition). Cambridge, MA: MIT
Press.

Lynch, K. (1990). Where learning happens. Cambridge, MA: MIT Press.

Lynch, K., & Carr, S. (2002). Where learning happens. In Banerjee, T., &
Southworth, M (Eds.), City Sense and City Design (pp. 418-429).
Cambridge, MA: MIT Press.

Madanipour, A. (1996). Design of urban space: An inquiry into a socio-spatial
process. Chichester, NY: Wiley.

Madden, K., & Love, K. (1982). User Analysis: An approach to park planning
and management. Davies, S. Washington, DC: American Society of
Landscape Architects.

McCormack, G. R, Mâsse, Louise C, Bulsara, M., Pikora, T. J, & Giles-Corti, B.
(2006). Constructing indices representing supportiveness of the physical
environment for walking using the Rasch measurement model.
Internatioanl Journal of Behavioral Nutrition and Physical Activity, 3(44).

McKenzie, T. L. SOPARC: Protocols. Robert Wood Johnson Foundation Active
Living Research. Retrieved January 8, 2008.
227

McKenzie, T. L. (2006).SOPLAY: Description and procedures manual. Robert
Wood Johnson Foundation Active Living Research. Retrieved 1/08, 2008,
from http://www.activelivingresearch.org/files/SOPLAY_Protocols.pdf.

McKenzie, T. L. (2007).SOPLAY: Mapping pdf file. Robert Wood Johnson
Foundation Active Living Research. Retrieved January 8, 2008, from
http://www.activelivingresearch.org/files/SOPARC-
SOPLAY_MappingStrategies_0.pdf.

McKenzie, T., L, Marshall, S. J, Sallis, J. F., & Conway, T. L. (2000). Leisure-
time physical activity in school environments: An observational study
using SOPLAY. Preventive Medicine, 30, 70-77.

McLaughlin, J. A., & Jordan, G. B. (2004). Using logic models. In Wholey, J.
S.,Hatry, H.P., et al (Eds.), Handbook of Practical Program Evaluation.
San Francisco, CA: Jossey-Bass.

Miles, R. & Panton, L. (2006). The influence of the perceived quality of
community environments on low-income women's efforts to walk more.
Journal of Community Health, 31(5), 379-392.

Mill, J. S. (1993). On liberty and utilitarianism. New York: Bantam Books.

Min, B., & Lee, J. (2006). Children's neighborhood place as a psychological and
behavioral domain. Journal of Environmental Psychology, 26, 51-
71.Moroni, S. (2006a). Planning, evaluation and the public interest. In
Alexander, E. R. (Ed.), Evaluation in Planning : Evolution and Prospects.
Burlington, VT: Ashgate Pub.

Moroni, S. (2006b). The ethics behind evaluation: Lichfield's approach and
utilitarianism. In Alexander, E. R. (Ed.), Evaluation in Planning:
Evolution and prospects. Burlington, VT: Ashgate Pub.
Morris, M. (1999). Research on evaluation ethics: What have we learned and why
is it important. New Directions for Evaluation. (82), 15-24.

Moudon, A. Vernez, & Lee, C. (2003). Walking and bicycling: An evaluation of
environmental audit instruments. American Journal of Health Promotion,
18(1).

228

Moudon, A., Vernez, L., Chanam, C., Allen, D., Garvin, C., Johnson, D. B.,
Schmid, T. L., & Weathers, R. D. (2007). Attributes of environments
supporting walking. Health Promotion, 21(3).

Mueller, D. C. (2003). Public choice III. Cambridge, MA: Cambridge University
Press.

Musil, Carol M., Warner, C. B., Zauszniewski, J. A., Jeanblanc, A. B., & Kercher,
K. (2006). Grandmothers, caregiving, and family functioning. Journal of
Gerontology, 61B(2), S89-S98.

Nagel, T. (1979). Mortal questions. Cambridge, MA: Cambridge University
Press.

Nasar, J. L. (1998). The evaluative image of the city. Thousand Oaks, CA: Sage
Publications.

Nasar, J. L. (2007). Assessing environmental perceptions. Presentation prepared
for the Active Living Research Annual Conference Feb 21-24, Coronado,
CA.

National Research Council. (2005). Does the built environment influence physical
activity? Examining the evidence. Transportation Research Board
Committee on Physical Activity Health Transportation and Land Use,
Washington, D.C. Transportation Research Board, Institute of Medicine
(U.S.).

Newman, D. L., & Brown, R. D. (1996). Applied ethics for program evaluation.
Thousand Oaks, CA: Sage Publications. Newman, O. (1972). Defensible
space: Crime prevention through urban design. New York,: Macmillan.

Newman, O. (1976). Design guidelines for creating defensible space.
Washington. DC: National Institute of Law Enforcement and Criminal
Justice, Law Enforcement Assistance Administration, U.S. Dept. of
Justice.

Nussbaum, M. C. (2000). Women and human development: The capabilities
approach. Cambridge, MA: Cambridge University Press.

Nussbaum, M., & Putnam, H. (1995). Changing Aristotle's mind. In Nussbaum,
M, & Rorty, A. O. (Eds.), Essays on Artistotle's De Anima. Oxford, CN:
Oxford University Press.
229

Owen, N., Humpel, N., Leslie, E., Bauman, A., & Sallis, J. F. (2004).
Understanding environmental influences on walking: Review and
research. American Journal of Preventive Medicine, 27(1).

Piko, B. F., & Keresztes, N. (2006). Physical activity, psychosocial health, and
life goals among youth. Journal of Community Health, 31(2).

Pikora, T., Bull, F., Jamrozik, K., Knulman, M., Giles-Corti, B., & Donovan, R.
(2002). Developing a reliable audit instrument to measure the physical
environment. American Journal of Preventive Medicine, 23(3), 187-194.

Pikora, T., Giles-Corti, B., Bull, F., Knulman, M., Jamrozik, K., & Donovan, R.
Systematic Pedestrian and Cycling Environmental Scan (SPACES)
Instrument. Robert Wood Johnson Foundation. Retrieved 1/29/07, 2006,
from
http://www.activelivingresearch.org/index.php/SPACES_instrument/323.

Pogge, T. Can the capability approach be justified? Retrieved 5/18/08, 2008,
from
http://mora.rente.nhh.no/projects/EqualityExchange/ressurser/articles/po
gge.pdf.

Prezza, M., Alparone, F., Romana, Cristallo, C., & Luigi, S. (2005). Parental
perception of social risk and of positive potentiality of outdoor autonomy
for children: The development of two instruments. Journal of
Environmental Psychology, 25(2005): 437-453.

Purcell, A. T. (1986). Environmental perception and affect: A schema discrepancy
model. Environment and Behavior, 18(1), 3-28.

Ramirez, L. K. Brennan, Hoehner, C. M., Brownson, R. C., Cook, R., Orleans, T.,
Hollander, M., Barker, D. C., Bors, P., Ewing, R., Killingsworth, R.,
Petersmarck, K., Schmid, T., & Wilkinson, W. (2006). Indicators of
activity-friendly communities an evidence-based consensus process.
American Journal of Preventive Medicin, 31(6), 515-524.

Rauschmayer, F. (2001). Reflections on ethics and multi-criteria analysis in
environmental decisions. Journal of Multi-Criteria Decision Analysis, 10,
65-74.

230

Rawls, J. (1974). Reply to Alexander and Musgrave. The Quarterly Journal of
Economics, 88(4), 633-655.

Rawls, J. (1993). Political liberalism. New York: Columbia University Press.

Rawls, J. (1999). A theory of justice. Cambridge, MA: Belknap Press of Harvard
University Press.

Rawls, J., & Kelly, E. (2001). Justice as fairness: A restatement. Cambridge, MA:
Harvard University Press.

Ritenbaugh, D, Kumanyika, S. K., Jeffrey, R. W., Morabia, A., & Antipatis, V. J.
(1999). Caught in the causal web: A new perspective on social factors
affecting obesity. Healthy Weight Journal, 13, 88-89.

Rocchini, A. P. (2002). Childhood obesity and a diabetes epidemic. The New
England Journal of Medicine, 346(11), 853-854.

Saelens, B. E., & Sallis, J. F. (2007). Neighborhood environment walkability scale
(NEWS). Robert Wood Johnson Foundation. Retrieved 1/29/07, 2007,
from http://www.activelivingresearch.org/.

Saelens, B. E., Frank, L. D., Auffrey, C., Whitaker, R. C., & Burdette, H. L.
Guidebook for EAPRS direct observation tool. Active Living Research
Program. Retrieved January 1. 2007, from www.activeliving.org.

Saelens, B. E., Frank, L. D., Auffrey, C., Whitaker, R. C, Burdette, H. L., &
Colabianchi, N. (2006). Measuring physical environments of parks and
playgrounds: EAPRS instrument development and inter-rater reliability.
Journal of Physical Activity and Health, 3(Supplement 1), 190-207.

Saelens, B. E., Sallis, J. F., Black, J. B., & Chen, D. (2003). Neighborhood-based
differences in physical activity: An environmental scale evaluation.
American Journal of Public Health, 93(9), 1552-1558.

Sallis, J., Cervero, R. B, Ascher, W., Henderson, K. A, Kraft, K. M., & Kerr, J.
(2006). An ecological approach to creating active living communities.
Annual Review of Public Health, 27.

Sallis, J. F., Bauman, A., & Pratt, M. (1998). Environmental and policy
interventions to promote physical activity. American Journal of Preventive
Medicine, 15(4), 379-397.
231

Sallis, J. F., Linton, L., & Kraft, M. K. (2005). The first active living research
conference: growth of a transdisciplinary field. American Journal of
Preventive Medicine, 28(2S2), 93-95.

Sallis, J., & Glanz, K. (2006). The role of built environments in physical activity,
eating, and obesity in childhood. The Future of Children, 16(1).

Sandercock, L. (1998). Towards cosmopolis: Planning for multicultural cities.
New York: J. Wiley.

Schoggen, P., Barker, R. G., & Fox, K. (1989, August). Behavior settings : A
revision and extension of Roger G. Barker's ecological psychology.
Stanford, CA: Stanford University Press.

Sen, A. (1977). Rational rools: A critique of the behavioral foundations of
economic theory. Philosophy and Public Affairs, 6(4), 317-344.

Sen, A. (1980). Equality of what? In McMurrin, S. M. (Ed.), Tanner Lectures on
Human Values. Salt Lake City, UT: University of Utah Press and
Cambridge University Press.

Sen, A.. (1990, Spring). Justice: Means versus freedoms. Philosophy and Public
Affairs, 19(2).

Sen, A. (1992). Inequality reexamined. Cambridge, MA: Russell Sage
Foundation; Harvard University Press.

Sen, A. (2000). Development as freedom. New York: Random House.

Shibley, R., Axelrod, E., Farbstein, J., Welch, P., & Wener, R. (2001). Making
and sustaining excellent urban places. In Watson, D., &Plattus, A. et al
(Eds.), Time-Saver Standards for Urban Design. New York: McGraw-
Hill.

Sister, C., Wolch, J., Wilson, J., Linder, A., Seymour, M., Byrne, J., & Swift, J.
The green visions plan for 21st century Southern California: Park and
open space resources in the green visions plan area. University of
Southern California GIS Research Laboratory and Center for Sustainable
Cities. Retrieved January 6, 2007, from
http://www.greenvisionsplan.net/html/documents/ParksReport.pdf.

232

Spence, J. C., & Lee, R. E. (2003). Toward a comprehensive model of physical
activity. Psychology of Sport and Exercise, 4, 7-24.

Stein, S., M., & Harper, T. L. (2005). Rawls's 'justice as fairness': A moral basis
for contemporary planning theory. Planning Theory, 4(2), 147-172.

Stoffregen, T. A, Gorday, K. M., & Sheng, Y. (1999). Perceiving affordances for
another person's actions. Journal of Experimental Psychology: Human
Perception and Performance, 25(1), 120-136.

Stokols, D. (1992). Establishing and maintaining healthy environments: Toward a
social ecology of health promotion. American Psychologist, 47(1), 6-22.

Talen, E., & Coffindaffer, M. (1999). The utopianism of children: An empirical
study of children's neighborhood design preferences. Journal of Planning
Education and Research, 18, 321-331.

Timperio, A., Crawford, D., Telford, A., & Salmon, J. (2004). Perceptions about
the local neighborhood and walking and cycling among children.
Preventive Medicine, 38, 39-47.

Troped, P. J., Cromley, E. K, Fragala, M. S, Melly, S. J., Hasbrouk, H. H,
Gortmaker, S. L. & Brownson, R. C. (2006). Development and reliability
and validity testing of an audit tool for trail/path characteristics: The path
environment audit tool (PEAT). Journal of Physical Activity and Health, 3
(Supplement 1).

Uhm, J. A, Lewis, F., & Banerjee, T. (2007). Children's capabilities: A view from
urban planning and design. Presented at the Workshop on Children's
Capabilities, University of Florence, Italy.

Walzer, M. (1995). Pleasures and costs of urbanity. In Kasinitz, P. (Ed.),
Metropolis: Center and symbol of our times. New York: New York
University Press.

Warren, W. H. (1984). Perceiving affordances: Visual guidance of stair climbing.
Journal of Experimental Psychology: Human Perception and
Performance, 10(5), 683-703.

Whyte, W. H. (1980). The social life of small urban spaces. Washington, DC:
Conservation Foundation.

233

Wicker, A. W. (1979). An introduction to ecological psychology. Monterey, CA:
Brooks/Cole Pubublishing Co.

Wolch. (2004). Green visions plan for 21st century Southern California: A guide
for habitat conservation, watershed health, and recreational open space.
Scientific Workshop on Strategies for Integrated Habitat, Watershed, and
Recreational Planning, Los Angeles, University of Southern California,
GIS Research Laboratory and Center for Sustainable Cities.

Wolch, J., Wilson, J. P., & Fehrenbach, J. (2005). Parks and park funding in Los
Angeles: An equity-mapping analysis. Urban Geography, 26, 4-35.

Wolff, J. (2007). Equality: The recent history of an idea. Journal of Moral
Philosophy, 4(1).

Woolley, H., & Johns, R. (2001). Skateboarding: The city as playground. Journal
of Urban Design, 6(2), 211-230.
Yin, R. K. (2003). Case study research : Design and methods (3rd). Thousand
Oaks, CA: Sage Publications.
Zabinski, M. F., Saelens, B. E., Stein, R. I., Hayden-Wade, H. A., & Wilfley, D.
E. (2003). Overweight children's barriers to and support for physical
activity. Obesity Res., 11(2), 238-246.

Zeisel, J. (1981). Inquiry by design : Tools for environment-behavior research.
Monterey, CA: Brooks/Cole Pub. Co.

Zhang, J. J. (2000, in press). Distributed cognition, representation and
affordances: Cognition and pragmatics Manuscript in author's
possession.

Abstract (if available)

Abstract Primary research, policy recommendations, and audit instruments are three responses to the U.S. obesity epidemic offered by built environment-physical activity researchers. This study is concerned with the theoretical foundations of audits.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Walkability as 'freedom': the ecology of school journey in inner city Los Angeles neighborhoods

PDF

The built environment, tour complexity, and active travel

PDF

Nonprofit organizations and the environmental policy outcomes: a systematic inquiry into the role of different types of nonprofits to influence the processes and outcomes of the environmental policy

PDF

San Francisco Recreation & Park Department Climate Action Plan: repositioning to a sustainable parks & open space system

PDF

Healthy mobility: untangling the relationships between the built environment, travel behavior, and environmental health

PDF

The production of public spaces: design dialectics and pedagogy

PDF

Governance and urban design: Omaha by design

PDF

Alternative ways, locations, and partners to meet the recreational needs of underserved communities: the case of Florence-Firestone

PDF

Foreign-related activities of the Chinese local governments and agents of globalization: a case study of 31 provinces in mainland China

PDF

Spaces of market-culturalism: the case of Chungking Mansions, Hong Kong

PDF

The Wenchuan earthquake recovery: civil society, institutions, and planning

PDF

Household carbon footprints: how to encourage adoption of emissions‐reducing behaviors and technologies

PDF

The future of work: defining a healthy ecosystem that closes skills-gaps

PDF

Homegrown violent extremism: designing a community-based model to reduce the risk of recruitment and radicalization

Asset Metadata

Creator Lewis, Ferdinand (author)

Core Title A capability-based approach to defining performance characteristics of the built environment

School School of Policy, Planning, and Development

Degree Doctor of Philosophy

Degree Program Policy, Planning, and Development

Publication Date 08/14/2008

Defense Date 06/10/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag active living,built environment audit instruments,built environment research,childhood obesity prevention,distributive justice,ecological model of perception,ecological psychology,Lynch, Kevin,OAI-PMH Harvest,quality of life measures,Rawls, John,Sen, Amartya,urban design policy,urban design theory,US obesity epidemic

Place Name USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Banerjee, Tridib (committee chair), Irazabal, Clara (committee member), McCann, Edwin (committee member)

Creator Email ferdinal@usc.edu,ferdinand.lewis@sbcglobal.net

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1578

Unique identifier UC1300741

Identifier etd-Lewis-2256 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-111084 (legacy record id),usctheses-m1578 (legacy record id)

Legacy Identifier etd-Lewis-2256.pdf

Dmrecord 111084

Document Type Dissertation

Rights Lewis, Ferdinand

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu