An aordance-based approach to architectural theory, design, and practice

Jonathan R. A. Maier and Georges M. Fadel, Clemson Research in Engineering Design and Optimization (CREDO) Laboratory, Department of Mechanical Engineering, Clemson University, 135 Fluor Daniel, EIB, Clemson, SC 29634-0921, USA Dina G. Battisto, Department of Architecture, Clemson University, Clemson, SC 29634-0921, USA The idea of aordance, borrowed from perceptual psychology, is applied to the domain of architecture. As to architectural theory, aordances can be used as a conceptual framework to understand the relationship between environments and occupants, especially with respect to form and function. Regarding architectural design, the concept of aordance allows for a common theoretical basis to improve the design process. Concerning architectural practice, aordances can be used as a tool to explore the connection between the intentions of the design with how the artifact is actually used, leading to archived knowledge, and the potential for avoiding common design failures. Published by Elsevier Ltd. Keywords: architectural design, design theory, design practice, design process, theory of aordances

Corresponding author: Jonathan R. A. Maier jmaier@ces.clemson. edu

rchitecture and design do not have a satisfactory theoretical basis, wrote psychologist James J. Gibson three decades ago. He then asked Can an ecological approach to the psychology of perception and behavior provide it? (Gibson, 1976). Clearly his opinion was yes, and we agree. In this article we expand upon this idea by applying Gibsons concept of aordance to the design of artifacts in general and in particular to the domain of architecture. In previous work we have applied the concept of aordance more specically to the eld of engineering design, where we have argued that the concept of aordance is more fundamental than other extant concepts, particularly that of function (Maier and Fadel, 2001, 2002; cf., Brown and Blessing, 2005). In this article we argue that, as in engineering, the concept of aordance is more fundamental to architecture than other often studied concepts, particularly that of form. One of our goals in this paper is therefore to show how the idea of aordances applies to a theoretical basis for architecture, in an answer to Gibsons provocative question.

www.elsevier.com/locate/destud 0142-694X $ - see front matter Design Studies 30 (2009) 393e414 doi:10.1016/j.destud.2009.01.002 Published by Elsevier Ltd.

393

Our application of aordances to architecture rests upon three main propositions, which we explore in the remainder of this paper. First, as to architectural theory, we assert that aordances can be used as a conceptual framework to understand the relationship between built environments and humans over time, especially with respect to the form, function, and meaning of architectural elements. Second, regarding architectural design, we propose that the concept of aordance allows for a common theoretical basis to improve the design process by oering a shared language among those involved in a design project, particularly architects and engineers. Third, regarding architectural practice, we believe that aordances may be used as an evaluation tool to explore the connection between the initial intentions or objectives of the design with how the artifact is actually used, leading to archived knowledge for use in future projects, and the potential for avoiding an array of common design failures.

In this regard we echo and expand upon some points made by Koutamanis in his application of the idea of aordance to building elements and spaces. He states Aordances promise integration of dierent viewpoints (architects, engineers, clients, users) and continuity, i.e., compatible expressions of functionality and usability throughout the lifecycle of a building (brieng, design and use). This holds promise for the codication of design knowledge: aordances could support direct matching of an existing building or type to a specic brief, thus allowing for early evaluation and renement of design or brieng choices (Koutamanis, 2006). Before expanding upon these ideas further, the concept of aordance needs to be explored and understood, as presented in the next section.

1 A generalized theory of aordances 1.1 History of the idea of aordance

The perceptual psychologist James J. Gibson rst put forward the theory of aordances. In other work, the present authors have expanded upon this theory, and identied new application areas (Maier and Fadel, 2001, 2002, 2003, 2005, 2007, in press). Following our introduction of the concept into the engineering design community, other authors have also begun using the concept of aordance within engineering design and industrial design research (e.g., Galvao and Sato, 2004, 2005, 2006; Brown and Blessing, 2005; Kim et al., 2007). In this section, we briey review our generalized theory of aordances (see Maier, 2005, Maier and Fadel, in press, for a more complete discussion) with a focus on its applicability to architecture. We begin with Gibsons original denition. Gibson coined the term aordance as follows (all emphases are his): The aordances of the environment are what it oers the animal, what it provides or furnishes, either for good or ill. The verb to aord is found

394

Design Studies Vol 30 No. 4 July 2009

in the dictionary, but the noun aordance is not. I have made it up. I mean by it something that refers to both the environment and the animal in a way that no existing term does. It implies the complementarity of the animal and the environment (Gibson, 1979). Gibsons book The Ecological Approach to Visual Perception is most concerned with how animals perceive their environment, which Gibson argues is through the perception of aordances in the environment. As such Gibsons theory of aordances is a descriptive formulation: it describes how animals perceive their environment. A decade later another psychologist, Donald A. Norman, took Gibsons theory of aordances and extended it into a prescriptive formulation: Norman gives some guidelines as to what certain objects should aord and should not aord. However, Norman, in his book The Psychology of Everyday Things, also published as The Design of Everyday Things (Norman, 1988), is concerned primarily with, as the title says, everyday things and not the design of artifacts in general or architecture in particular. Hence Normans theory culminates in two design-for-x methodologies (design-for-usability and design-for-error) but stops short of incorporating the concept of aordance as fundamental to the design of any artifact. Following Normans seminal use of the concept of affordance in interaction design, there has been considerable discussion of aordances in the context of human-computer-interaction (HCI) (cf., Norman, 1999). Other psychologists, such as Warren and his students, have applied the concept of aordance to design specic artifacteuser relationships, such as the height of stair steps (Warren, 1995). However, such an approach relies on the analysis of so-called invariant geometric relationships, such as the ratio of leg height to stair height, which may not exist in all cases. The idea of aordance has also been applied in the eld of articial intelligence, e.g., how to design robots that recognize aordances in their environment (as in Murphy, 1999), and in a somewhat similar vein, to the study of human childhood development (see, e.g., Gibson, 2000). Some authors in the Human Factors eld have recently applied an ecological approach to the design of user interfaces, known as Ecological Interface Design (EID) (Vicente and Rasmussen, 1992). Their approach emphasizes high-level processing of data by human users and speaks chiey to the layout and conguration of displays. However, their methods are not explicitly centered on aordances and do not address design projects in general. Meanwhile, a few other researchers have begun to apply the concept of aordance to architectural design explicitly. Tweed (2001) reports progress on identifying and highlighting aordances within computer-aided architectural design (CAAD) software. Koutamanis (2006) follows Normans application of the concept of aordance to individual building elements such as door

An aordance-based approach to architectural theory, design, and practice

395

handles, and proceeds to apply aordances to architectural spaces in which he introduces the technique of aordance mapping to help architects think about the aordances of their architectural designs. In a similar vein, Kim et al. (2008) study the actual aordances of a building lobby and how they vary between dierent users. Whereas the above authors tend to focus on techniques for using the concept of aordance in architecture, our focus in this paper is a thorough application of aordances on a theoretical level which should underlie subsequent practical applications.

1.2

Examples of aordances

The concept of aordance is perhaps most easily understood through some simple examples. Gibson gives the following examples:  If a terrestrial surface is nearly horizontal (instead of slanted), nearly at (instead of convex or concave), and suciently extended (relative to the size of the animal) and if its substance is rigid (relative to the weight of the animal), then the surface aords support  Terrestrial surfaces, of course, are also climb-on-able or fall-o-able or getunderneath-able or bump-into-able relative to the animal. Dierent layouts aord dierent behaviors for dierent animals  Air aords breathing, more exactly, respiration. It also aords unimpeded locomotion relative to the ground. when illuminated and fog-free, it affords visual perception. It also aords the perception of vibratory events by means of sound elds and the perception of volatile sources by means of odor elds  Solids aord various kinds of manufacture, depending on the kind of solid state. Some, such as int, can be chipped; others, such as clay, can be molded; still others recover their original shape after deformation; and some resist deformation strongly (Gibson, 1979) Pertaining to architecture specically, the following examples of aordances may be helpful:  Buildings have many high-level aordances, including aording shelter to occupants from the exterior environment, aording aesthetics to occupants and passers-by, aording storage of goods, aording comfort to occupants through climate control, etc. More detailed aordances can better be analyzed by looking at specic building elements.  Windows aord the transmission of light, and hence illumination of the interior environment as well as a view of the exterior environment. Operable windows may also aord the exchange of air, and in extreme cases even defenestration.  Floors aord the support of occupants weight, as well as furniture, the attachment of nish materials, the routing of utilities, and in some cases even drainage.

396

Design Studies Vol 30 No. 4 July 2009

Notice how some of the above aordances are positive, i.e., benecial to the user, such as the lights, air and views through windows. Other aordances are negative, i.e., harmful to users, such as defenestration through windows. With both Gibsons and our examples of aordances in mind, for both environmental things and architectural elements respectively, we can now generalize the concept in order to apply it to the design of artifacts in general and architecture in particular. This is accomplished in the context of two distinct classes of aordances, artifact-user aordances and artifact-artifact aordances, as discussed next.

1.3

About artifacteuser aordances (AUA)

Aordances always express a complimentary relationship between two separate systems. In Gibsons original formulation, one is the environment, and the other is an animal situated in it. For design, and especially for architecture, we can view the environment as the built environment, i.e., artifacts, and consider the typical animals in them to be in fact human users. Hence the usual aordances of interest exist between artifacts and users. We call these artifact-user aordances (AUA). Aordances are distinguished from other types of interaction by the potential usefulness of the relationship. Other types of relationships, such as ownership, or even the artifacts actual use by the user are not aordances. An aordance indicates the potential for a behavior, but not the actual occurrence of that behavior. As Gibson pointed-out, individual properties of either the artifact (color, density, size, etc.) or the user (strength, age, height, etc.) are not in and of themselves aordances, but taken together can determine whether a specic aordance exists, such as the ability of a specic person to walk on a specic oor. Note also that an aordance must rst exist before the behavior aorded can ever be exhibited. For example, the aordance of visibility through a window is one type of interaction, while the behavior of a person looking through the window is a dierent type of interaction, but the two are related because the window must (rst) aord visibility before it can (second) ever actually be gazed through.

1.4

About artifacteartifact aordances (AAA)

An aordance does not need to be perceived for that aordance still to exist. This means that the human user does not need to even be present for the aordance to exist. Taking another step of abstraction, we recognize that an aordance expresses a relationship between two (or more) subsystems in which a behavior can manifest between the two subsystems that either subsystem cannot manifest in isolation. Examples exist of course between artifacts and users (e.g., turnability of a door-knob, readability of a sign) between multiple users (e.g., conversations, mating, ghting, etc.), and nally between multiple artifacts (e.g., walls aording support to roofs, sprinklers aording suppression of res). We call the latter relationships artifact-artifact aordances (AAA).

An aordance-based approach to architectural theory, design, and practice

397

Obviously buildings must be designed to aord desired uses to its occupants and other stakeholders, however, the components of the building must be designed to have aordances as well. The behaviors of these aordances may be realized in practice, such as walls supporting roof loads, or may be designed simply as contingencies that may never be realized, such as re suppression, or extreme loading by snow, hurricanes, or oods. It is important to note that whereas an AUA expresses a relationship that is directly useful to users (including the meaning of architectural elements as discussed in Section 2.3), an AAA expresses a relationship that is indirectly useful to users. Floors must support users walking on them, however, walls must support roofs, but this is ultimately to protect users within the building.

1.5

Aordances and system behavior

Whether an aordance exists, and what quality the aordance is, depends upon the structure of both of the subsystems involved. A thicker, stronger oor affords support to heavy occupants better than a thinner, weaker oor. However, a thinner, weaker oor may be adequate to support lighter users, such as children. In the case of AUA, designers have control over the structure of the artifacts they design, and thus over what aordances exist with respect to specic users, over whom they usually do not have control. In the case of AAA, designers have control over both artifacts, but still need to design the aordance and resulting behavior ultimately to benet or to protect human users. At this point three basic categories can be identied as essential to any aordance relationship. The rst is structure (of artifacts and/or users), the second, behavior (again, or artifacts and/or users), and the third, purpose. The fundamental relation between these categories is that systems aord behaviors via their structure for a purpose. This is a more detailed adaptation of the dictum from general systems theory that structure inuences behavior (cf., Senge, 1990). Essentially, we have used the generalized concept of aordances to describe how structure inuences behavior, and to what ends. Structure determines what aordances exist. The aordances indicate what behaviors are possible, whether or not they are ever expressed. The ultimate usefulness of the aordance to users (directly in the case of AUA, or indirectly in the case of AAA) is the purpose of the system and its organization. In the next subsection we oer a specic architectural example, using the theory of aordances to understand some well publicized failures of modern architecture.

A motivating example using the concept of aordance to understand building failure

Hermetically designed concepts for high-rise housing towers provide an example of how unintentional social behaviors develop in structures that inherently

1.6

398

Design Studies Vol 30 No. 4 July 2009

aord those behaviors. Modern masters such as Le Corbusier, in his austere high-rise multi-family housing projects, signicantly inuenced a generation of housing schemes in urban environments in the International Style. Modernist ideals were evident in the design of the failed public housing developments such as Pruitt-Igoe in St. Louis and Cabrini Green in Chicago, although the lessons learned following the occupation of the buildings by the tenants clearly show that there was a gap between the intentions of the design and the ensuing behaviors of the tenants (cf., Hillier, 2007). The building achieved its intended aordances of providing high density, inexpensive housing. Its failure was due to the multiple macro-scale unintended negative aordances that resulted in such bad actual living conditions. Although it is a myth that Pruitt-Igoe ever received any kind of architectural prize, some of its design features that were initially praised, such as the design of the elevators and hallways (Figure 1a), soon proved to be opportune environments for violent crime (Bristol, 1991). Due to a variety of factors, including demographic shifts within the city, poor management and maintenance, as well as actual design aws, three of the high-rise towers of Pruitt-Igoe were intentionally demolished in 1972 by the St. Louis Housing Authority (Figure 1b), just 18 years after their completion in 1954. We suggest that if the aordances of the proposed buildings had been better understood in the design stage, then the building design could have been modied so the building would not aord the undesirable behaviors later experienced by the tenants. Recently developed strategies for identifying and understanding aordances in the design process are discussed in the next section.

Strategies for identifying and understanding aordances in the design process

Two design tools have recently been developed to help designers identify and manage aordances with respect to user behaviors and artifact structures in the design stage. The Function Task Interaction Matrix (FTIM) proposed by Galvao and Sato (2005) identies aordances as the intersection between artifact structure and user tasks. While Galvao and Sato (2005, 2006) originally demonstrated the FTIM with consumer products such as cell phones and digital cameras, Kim et al. (2007) have applied the FTIM to identify 28 aordances to guide their design of the interior of a conference room. They conclude It is more systematic and eective to use these relationships in the design of [a] conference room to meet specic needs of users rather than to only rely on designers experience and knowledge. Another matrix based design tool to aid designers in understanding and managing the impact of artifact structure on aordances is the Aordance Structure Matrix (ASM) which is being developed by the present authors (Maier et al., 2007b, 2008). Using an ASM, changes to artifact structure can be traced to the aordances that depend on each structural element. Thereby

1.7

An aordance-based approach to architectural theory, design, and practice

399

Figure

Demolition

of

Pruitt-Igoe housing project ( Newman, 1996)

possibilities for improving positive aordances and mitigating negative aordances can be explored during conceptual design. Methods for designing individual aordances have also been introduced by the present authors (Maier and Fadel, 2003). Finally, Tweed (2001) has begun investigating explicit representation of aordances in computer-aided architectural design software. Although research into applying aordance-based design theory to the design of architectural elements is on-going, as discussed further in Section 3, we suggest that the careful application of even these early concepts and tools can produce better architecture with better aordances and less tendency for failure.

2 Application to architectural theory and design 2.1 Historical separation of form and function in architecture
As the preceding example illustrates, an understanding of the relationship between form (or more generally, structure) and function (or more generally,

400

Design Studies Vol 30 No. 4 July 2009

user behavior) is essential for a successful design project. An understanding of the separation between form and function in architecture helps to explain the theoretical dilemma that designers face today. The need to provide appropriate form and function has been recognized since antiquity, e.g., in the writings of Vitruvius (Kruft, 1994). However, for Vitruvius, and for centuries of architects that he inuenced, form (rmitas) and function (utilitas) were considered separate but competing requirements, among others, such as beauty (venustas) (Vitruvius, 1960). We suggest that the concept of aordance can be used as a conceptual basis to unite the originally separate Vitruvian ideas of form and function, as explained in the following sections.

2.2 The inuence of aordance in existing architectural theory

In the domain of architecture, the concept of aordance has not been applied to design directly, although as a concept it has inuenced work within the integrative eld of Environment and Behavior (EB) research, which attempts to integrate form and function more than earlier conceptions which separated form and function. EB research explores the relations between people and their surroundings and is a multi-disciplinary eld spanning sociology, environmental psychology, humanistic geography, natural resources, urban planning and architecture. EB research is oriented toward understanding human characteristics at all levels of human experience (from physiological responses to social and cultural phenomena) and at all scales of the everyday physical environments (from the micro-scale of interiors to the macro-scale of regions.) One of the primary objectives of EB research is to discover underlying mechanisms that link environment and behavior together (Moore, 1987). We believe the concept of affordance is just such a mechanism, as aordances link the structure of the environment with the capabilities of human users to determine what behaviors are possible and even likely. Furthermore, current EB research practices can aid designers in anticipating aordance perception and behavior response. Established research instruments such as the environmental assessment technique (Astin and Holland, 1961) and environmental descriptor scales (Kasmar, 1970) can be used to quantify attributes of built environments or concepts and test hypotheses about usages of the building (Craik and Feimer, 1987). Kims recent evaluation of the aordances of a building lobby demonstrates this approach in action (Kim et al., 2008). The concept of aordance can also be applied to understand basic ideas within person-environment theories in the behavioral sciences and gerontology. Theoretical models such as those by Lawton and Nahemow (1973) and Lawton

An aordance-based approach to architectural theory, design, and practice

401

(1982), and Kahana (1982) involve an interaction approach to studying older adults and their environments. Much of this work emphasizes a quantitative view of the person-environment interaction building on the work of Lewin (1951). This approach focuses on the functional aspects of the personenvironment relations and views the interaction as a system that is divisible into separate entities of person and environment.1 Both person and environment are believed to be uid, one changing as the other changes, with cause and eect being dicult to separate. The concept of aordance provides a useful framework to understand just such a relationship, where the (sometimes uid) properties of both person (or user) and (articial) environment determine the aordances that exist and the consequent behavior that manifests. To the extent that these properties can be quantied, so too might the resulting aordances and behavior. Such was the basic approach of Warren (1995) in his studies of stair step height. The concept of aordance can also be applied to person-environment theories within place-based research. This work focuses on a qualitative view of the interaction. The disciplines interested in understanding this type of interaction involve elds of environmental psychology (see, e.g., Canter, 1977), architecture (see, e.g., Schulz, 1988), and humanistic geography (see, e.g., Tuan, 1980). The term theory of place is a term often used to encompass the meanings and emotional ties associated with physical environments. The person-environment relationship is then viewed as a unied interaction, from which individual components of the environment and behavior experience cannot be separated. The researchers view the interaction in terms of the experiential nature of the place and the consciousness of the individual. Interpreted as aordances, this type of interaction emphasizes the fact that aordances depend on both the artifact (i.e., environment) and user (i.e., person) and do not exist in isolation from the other. This research also takes into account more of the psychological properties of the user than the quantitative person-environment theories discussed above.

2.3

On form and meaning in architecture

As we have just seen, the study of aordances obviously entails an understanding of the interaction between people and their (articial, i.e., architectonic) environment. Similarly, the study of form and meaning within architecture has addressed the connections between an environmental setting and the observer. Thus, it seems plausible that meaning can be considered as a separate aordance, dependent on the physical form of architectural elements and the past experiences, beliefs, aesthetic preferences, etc. of the observer. We explore this idea in the remainder of this section, beginning with a brief review of the history of meaning in architecture. Since the 20th century, the investigation of the symbolic quality of the build environment has been a focus of architectural discourse particularly during

402

Design Studies Vol 30 No. 4 July 2009

the Post Modern Era. This can be seen in historical studies in architectural iconography as seen in e.g., Venturis (1966) Complexity and Contradiction in Architecture and Moore and Bloomers (1977) Body, Memory and Architecture. In other work, symbolic qualities have been linked to the signicance or production of meaning (semiotics) by the sociologist Umberto Eco (1979) in his Theory of Semiotics. Building upon readings from Edmund Husserl, Martin Heidegger and Gaston Bachelard, Pallasmaa formulates a theoretical position about experiences reliance in memory, imagination, and the unconsciousness. Pallasmaa asserts that meaning in architecture depends on its ability to symbolize human existence or presence, and as modern architects appear to have overlooked, on the spatial experience of the work.Forms themselves are meaningless, but can transmit meaning via images enriched by associations.Science and reason, have contributed limiting mindsets like analysis, elementarism, and reductionism, with unfortunate consequences for architecture.By contrast, the experience of architecture is synthetic, operating at many levels simultaneously: mental/ physical, cultural/biological, collective/individual, etc. (Pallasmaa, 1986: p. 447) As noted previously, meaning in the context of architecture has been shown to be an important component in linking people to places (Schulz, 1976, 1988). There are multiple ways in which a building can convey meaning. For example, meaning may be associated with an architectural forms historical signicance or specic cultural, political, or social symbolism. Columns, for example, while performing the function of supporting a roof load, to some mean power and prestige hearkening back to the Grecian and Roman empires, however, to other people these columns may symbolize American colonial racial repression. Marble on the oor of a bathroom, while aording support to walk on and a water-proof seal above the sub-oor, may be perceived as luxurious because of its relative expense. Such meanings are not captured in computer-aided-design (CAD) systems or any other form-based representation (Maier et al., 2007a). They are culturally embedded; they are taken with reference to the human users and their cultural identity. As such, meaning may be considered as another category of aordance. In order to understand the aordances, both positive and negative, within a particular context, the study of meaning may shed light on the connections between artifact and user. What is meant (or not meant) in a particular architectural setting depends, as theorists such as Pallasmaa have realized for years, on the prior experiences of the individual. The meaning of any architectural

An aordance-based approach to architectural theory, design, and practice

403

attribute, therefore, depends on the individual, just as does any other aordance of that same feature.

3 Application of aordances to architectural practice 3.1 An aordance-based view of the design process
In an eort to understand the design process in general, we rst need to take a step back and address the question of why people design things at all. Once again, Gibson applies the powerful concept of aordance to answer this fundamental question: Why has man changed the shapes and substances of his environment? To change what it aords him (Gibson, 1976). Hence the impetus for any design project can be understood in terms of creating and changing aordances d neither creating artifacts to do certain things, as a functional view of design would hold, nor creating artifacts solely on the basis of creating a beautiful form, but rather to create artifacts that can be used and that have meaning. The concept of aordance provides an alternate way of viewing the design of environments, emphasizing the complementarity of the relationship between environments and their users, i.e., between the form of buildings and the resulting behavior of their occupants as the building functions in practice. We envision architectural design as fundamentally being no dierent from the design of any other artifact. First, the desired (positive) aordances and the undesired (negative) aordances must be identied. As usual, these aordances can be divided into two broad categories, artifact-user aordances (AUA) and artifact-artifact aordances (AAA) as discussed previously (cf., Maier and Fadel, 2003, 2007). In architecture, AUA are extremely important yet often neglected to the extreme consternation of buildings occupants and visitors. In The Design of Everyday Things, Norman points out the frustration experienced by many users of doors in buildings which suer from common (yet easily avoided) pitfalls such as failing to indicate whether they need be pulled or pushed (by poor choice of handles), or failing to properly indicate that they are doors at all (e.g., glass doors). Modern buildings also sport an abundance of signage to explain to people how the buildings should be used. As Norman has pointed-out about objects in general, if they require written instructions, they are probably bad designs. A better design indicates, by virtue of its structure, how it should be used (what it aords). No explanation should be necessary (Norman, 1988). Meanwhile, AAA are also extremely important in architecture, for architects are also responsible for various modern technical systems of buildings: their lighting, heating, cooling, ventilation, electrical provision, lightning protection, plumbing, endurance of severe weather, maintainability, longevity,

404

Design Studies Vol 30 No. 4 July 2009

re-resistance, etc. Currently, these various design activities share no common basis. If architects and engineers were to approach these systems from a common theoretical basis, that of aordances, it would be easier for them to communicate and concurrently design their systems. Unfortunately, usually it is not merely sucient to design an artifact that possesses certain aordances; as Norman has pointed-out, that artifact must also not possess certain undesired aordances. Hence, as we have discussed in prior work (Maier and Fadel, 2001; Maier and Fadel, in press; Maier, 2005) the design process can be viewed simply as the search for an artifact that:  possesses all the desired aordances, and  does not possess any undesired aordances. Note that the vitally important determination of what exactly the artifact is supposed to aord and not to aord is an early and crucial phase of the design process itself. Within the domain of architecture, this phase of the design process is often referred to as architectural programming.

3.2 Applying aordance-based design to architectural programming

In architecture generally, we see aordance-based design as most benecial in this architectural programming phase of a project. It is important to note, however, that the concept of aordances should not be limited to just programming but relates to all phases of architectural design. One of the most referenced texts, Problem Seeking by Pena (2001), denes architectural programming as an analysis stage and is distinct from the later design phase dened as problem solving. Penas initial approach to architectural programming (published as early as 1977) has signicantly inuenced the practice of architecture and provided a basis for subsequent writings on architectural programming (e.g., Kumlin, 1995; Cherry, 1998; Hamilton, 1999; Hershberger, 1999). During pre-design phases architects and other professionals on the planning team generally employ several interactive methods and techniques to collect information on the internal and external forces that impact a given project. Generally speaking, the planning team may conduct environmental evaluations and assessments, collect data, research best practices, interview dierent user groups, establish goals, and generate a list of technology and equipment needs. During this inquiry process, objectives are established, the context of the design problem is identied and the performance requirements are dened (Duerk, 1993). Ultimately, a large amount of information is distilled into several key concepts that are intended to inform the later design process. Most architectural programming approaches utilize a divided programming and design process and this is often reinforced in the organizational structure of large architectural rms. A limitation of this

An aordance-based approach to architectural theory, design, and practice

405

fragmented approach is the ineective transfer of information to the designers and the subsequent disconnect throughout the entire design process. The architectural programming process is a highly analytical process that tends to focus on things that are measurable. Such a highly analytical process of design often creates a more functional view of a design problem. It is important, however, that the functionality of design must be balanced with an artistic or poetic view of design in order to achieve Quality Architecture (Pirsig, 1974). By providing a common understanding of the design process, means of representing requirements, and means of analyzing candidate structures, the concept of aordance can provide a framework to create a more uid transfer of information through the multiple stages of the design process, and a more coordinated eort in the planning and designing eorts among multiple disciplines. Note also that the determination of aordances directly requires the expertise of designers who have knowledge of the context in which the artifact or building will be used. This includes the meaning suggested by individual elements as well as everything that will need to be accomplished with the artifact (which leads to everything the artifact needs to aord), not only what the artifact will need to do itself (function) or look like (form). Recall that an important dierence between aordance-based design and other views of the design process is the formal identication of things that the design should not aord. In other words, it is not sucient to design an artifact to accomplish certain goals (e.g., meeting budget, attractive appearance) if that artifact also accomplishes undesired goals (e.g., criminal activity, short life-span).

3.3 Aordance-based design and an architectural knowledge base

A key advantage of the concept of aordance is that it can be used from the very early stages of a design project through to evaluation after the project is completed. Understanding early information in terms of aordances can help designers determine appropriate measurable goals or hypotheses that may serve as guiding principles in the design process. These hypotheses can then be evaluated following the occupation of the building. Doing so may aid in developing a knowledge base on the success and failure of designs and replace the tradition of transferring information and knowledge through oral history. The need for such a knowledge base has been articulated at national conferences on the design of buildings and landscapes. For example, the 2003 National American Institute of Architects (AIA) Convention theme was Design Matters e Poetry and Proof (Aiarchitect, 2003) a provocative title which conveys that there is not a sucient knowledge base to inform design decisions and help justify design decisions when communicating with the client or

406

Design Studies Vol 30 No. 4 July 2009

other stakeholders involved in the design process. According to Hamilton (2003), architects are accustomed to casual investigation as opposed to scientic rigor, i.e., not establishing any clear hypotheses with subsequent claims that can be measured to build a knowledge base. When the design intentions are not clearly articulated in the design process, it makes it dicult to conduct a post-occupancy evaluation to determine if the design is meeting measurable objectives, e.g., with regard to safety, comfort, productivity, exibility, aesthetics, etc.

To respond to clients expectations, those involved with designing environments are seeking ways to advance their thinking and position themselves as leaders capable of blending the creative arts and a credible knowledge base. The appropriate application of the concept aordances may help to build this knowledge base, by providing a means for comparing actual behaviors with the intended aordances of a structure, and documenting solutions to problems encountered in practice, so these problems can be avoided in future projects. Such is the approach taken by the authors in the recent design of a community wellness center (Battisto et al., 2006).

3.4 Applying aordance-based design to understanding architectural failures

There are numerous examples of unintentional consequences that may be discovered after an artifact is built. These consequences may be revealed after the people begin to use the building, or as programs and operations evolve, and they may change throughout the lifecycle of the artifact. Unintentional design consequences within the domain of architecture include unexpected behaviors, adaptation, interpretation error, signage, and obsolescence.

Unexpected behaviors occur when the aordances of designed structures were not correctly understood, or when structures create an environment in which novel behaviors can occur. Consequently, users interact with the designed building in ways the designers either did not or could not anticipate. Just as jumbo-jet air-liners were not designed to be missiles, but unfortunately aord this use in the hands of terrorists, spectacular and symbolic architectural landmarks such as the World Trade Center in New York and the Pentagon building in Washington, D.C. aord targets for terrorism. The hallways of Pruitt-Igoe were designed to foster community interaction but instead aorded a haven for criminals through which their prey had to traverse.

Adaptations are needed when a building structure does not aord desired behaviors, or aords undesired behaviors. Levy and Salvadori (1994) provide many examples of adaptations to buildings over time due to unintentional design consequences.

An aordance-based approach to architectural theory, design, and practice

407

Interpretation error occurs when a building is designed to aord one behavior, but in practice aords a dierent behavior. The external inuences that shape how an artifact is used over time may also alter the interpretation of that artifact. For example, the symbolic front door in American homes has changed over time with the introduction of the automobile; meanwhile new entryways to the house have inuenced the reading of the house within the contemporary culture. The physical nature of the front door has not changed, although the pathway to the house is now viewed in terms of the automobile and has been emphasized with the garage. The door still aords entry to the house, but is now rarely used for that.

Signage is often used to compensate for the lack of properly designed aordances. Within any large institutional, educational or commercial building one can see signs placed all over these facilities for the purpose of orienting the user, e.g., watch your step, or do not enter. Better designs are those that do not need signs because they indicate by their structure how they are to be used. For example, the front entry doors to the Cooper Library at Clemson University were recently changed. In the old design, the exterior handles aorded both pulling and pushing, although the doors only opened inward. To remedy the problem of patrons consternation over frequently trying to pull the door open, signs were installed instructing the user to PUSH (Figure 2a). However, the handles themselves were much larger than the signs, and users continued to try to pull the doors open. Recently, the problem was resolved by replacing the handles with push plates that only aord pushing and not pulling (Figure 2b). Note that with the new push plates the signage for PUSH is in fact redundant because the plates only aord pushing, not pulling.

Obsolescence in artifacts may result when buildings aord specic uses, but do not aord change. Often times, it is too expensive to rectify or to modify the design of an artifact and it becomes prematurely demolished. In many cases, these buildings are built with (i.e., to aord) a specic intended purpose, for a temporary time period.

3.5 Additional ramications of aordance-based design for architectural practice

The concept of aordance-based design also suggests a natural metric against which various designs may be compared. Any design that aords its intended purpose may be called a successful design. However, a design that aords that same purpose while aording other desirable features (such as user comfort, safety, durability, recyclability, etc.) may be viewed as better. Furthermore, a design that does aord everything it is supposed to do, while also aording something it is not supposed to do, is worse than a comparable design that only aords what it is supposed to do. It is also important to remember that aordance is not always a clear has or

408

Design Studies Vol 30 No. 4 July 2009

Figure 2 Front entry doors to the Cooper Library at Clemson University

has not distinction. One artifact may aord a desired use better than another. Moreover, the aordances of the artifact can dier with respect to dierent users. The concept of environmental role developed by the British psychologist David Canter clearly supports this premise. Building from his general theory of place as mentioned earlier, environmental role is a particular set of associated behaviors and rules within a particular place that vary according to the relationship between an individual and place (Canter, 1977). He argues

An aordance-based approach to architectural theory, design, and practice

409

that people have dierent relationships with places depending on their role within that particular place. Thus, people have particular purposes with regard to place that shape their behavior and their conceptualizations of that place (Canter and Walker, 1980). In our language, the aordances of the building environment would be dierent for individuals in dierent user groups. The aordances may vary for individuals within each user group as well. It is important to notice here how the concept of aordance supports experiential aspects of good design.

It is also interesting to note how the concept of aordances applies over a wide range of scales from the very large macro-scale of projects such as Pruitt-Igoe to more human-scale objects such as entry doors to smaller micro-scale elements such as door handles. This observation emphasizes the general applicability of the concept of aordances across the eld of architecture as we have advocated in this paper.

Summary remarks

In this article we have proposed that the idea of aordance, an established concept from ecological psychology, may serve as a conceptual basis and unifying framework for architectural theory, design, and practice. As many researchers have realized, architectural design, as other elds of design, lacks a rigorous theoretical basis. The concept of aordance can provide that basis, as Gibson originally suggested, because it ties together human occupants (users) and the built environment (artifacts) and explains why and how users behave the way they do. These ideas unite various perspectives from architectural theorists from Vitruvius to Norburg Schulz, Heidegger, Canter, Pirsig, and other contemporary thinkers. We have also shown how the concept of aordance provides a useful conceptual framework for understanding past architectural failures, and how an understanding of aordances can help prevent such failures in the future. More broadly, the scheme we have outlined of aordancebased design for architecture can be used as a broad framework uniting the work and understanding of architects, engineers, future occupants, and other stakeholders. Such a framework will also facilitate knowledge capture and re-use.

We realize that these ideas are presented at a high-level and are chiey theoretical in nature. It is our hope that this article provokes a lively debate within the architectural and broader design community, and helps to undergird parallel eorts by other researchers who are actively developing techniques for applying aordances to architectural practice. The potential benets of a rigorous, testable, and practicable design theory are enormous. So too are the costs of continuing ad-hoc practices which frequently result in dissatisfaction and occasional outright failure. We invite and look forward to the opinions and contributions of other researchers in the pursuit of a unifying and mature theory

410

Design Studies Vol 30 No. 4 July 2009

for architectural design and practice. We commend the concept of aordance as an important tenet within such a theory.

Acknowledgements
We would like to thank Chris Pagano of the Department of Psychology at Clemson University for his many valuable comments on drafts of the paper. We would also like to thank Peg Tyler, Associate Librarian in the Clemson University Libraries for graciously providing the photographs of the door handles on the Cooper Library. We also thank the anonymous reviewers of earlier drafts of this paper for their helpful comments, especially with respect to the Pruitt-Igoe example and the scale of aordances.
1. Lewin argued that behavior in general could be viewed generally as a function of the interaction between people and the environment. Based on his research a formula was derived: B f (P, E, P E), where behavior (B) is a function (f) of person (P) and environment (E) and the interaction between the two (P E).

References
Aiarchitect (2003) AIA national convention and exposition, <http://www.aia.org/ aiarchitect/thisweek03/tw0509/convention2003.asp> (accessed 12.10.07) Astin, A W and Holland, J L (1961) The environmental assessment technique: a way to measure college environments, Journal of Educational Psychology Vol 52 pp 308e316 Battisto D, et al (October 2006) A green community wellness center: expanding the scope of design for health, AIA Academy Journal Bristol, K G (1991) The Pruitt-Igoe myth, Journal of Architectural Education Vol 44 No 3 pp 163e171 Brown, D C and Blessing, L (2005) The relationship between function and affordance, in ASME Conference on Design Theory and Methodology, Longview, CA, Paper No. DETC2005e85017 Canter, D (1977) The psychology of place Architectural Press, London Canter, D and Walker, E (1980) Environmental role and the conceptualizations of housing, Journal of Architectural Research Vol 7 No 3 pp 30e35 Cherry, E (1998) Programming for design: from theory to practice John Wiley & Sons, New York Craik, H K and Feimer, N R (1987) Environmental assessment in Handbook of environmental psychology, John Wiley & Sons, New York pp 891e918 Duerk, D (1993) Architectural programming: information management for design John Wiley & Sons, New York Eco, U (1979) A theory of semiotics Indiana University Press, Bloomington, IN Galvao, A B and Sato, K (2004) Human-centered system architecture: a framework for interpreting and applying user needs, in ASME Conference on Design Theory and Methodology, Salt Lake City, UT, Paper No. DETC2004-57545 Galvao, A B and Sato, K (2005) Affordances in product architecture: linking technical functions and user requirements, in ASME Conference on Design Theory and Methodology, Long Beach, CA, Paper No. DETC2005-84525 Galvao, A B and Sato, K (2006) Incorporating affordances into product architecture: methodology and case study, in ASME Conference on Design Theory and Methodology, Philadelphia, PA, Paper No. DETC2006-99404

An aordance-based approach to architectural theory, design, and practice

411

Gibson, J J (1976) The theory of affordances and the design of the environment Symposium on Perception in Architecture, American Society for Esthetics, Toronto, October 1976. Republished as Chapter 4.9, Part VIII in E Reed and R Jones (eds) Reasons for realism: selected essays of James J. Gibson (1982), Lawrence Erlbaum Associates, Hillsdale, NJ, USA Gibson, J J (1979) The theory of affordances in The ecological approach to visual perception, Houghton Miin, Hopewell, NJ, USA Gibson, E J (2000) Perceptual learning in development: some basic concepts, Ecological Psychology Vol 12 No 4 pp 295e302 Hamilton, D K (1999) Innovations in planning for healthcare The Center for Innovations in Health Facilities, College Station, TX, USA Hamilton, D K (2003) The four levels of evidence-based practice, Healthcare Design Vol 3 No 4 Hershberger, R (1999) Architectural programming and predesign manager McGraw-Hill Professional, New York Hillier, B (2007) Space is the machine: a congurational theory of architecture Space Syntax, London, UK Kahana, E (1982) A congruence model of personeenvironment interaction in M P Lawton, P G Windley and T O Byerts (eds) Aging and environment: theoretical approaches, Springer, New York Kasmar, J V (1970) The development of a usable lexicon if environmental descriptors, Environment and Behavior Vol 2 pp 153e169 Kim, Y S, et al (2007) Affordances in interior design: a case study of affordances in interior design of conference room using enhanced function and task interaction, in ASME Conference on Design Theory and Methodology, Las Vegas, NV, Paper No. DETC2007-35864 Kim, Y S, et al (2008) Personal cognitive characteristics in affordance perception: the user activity case study in a building lobby, in ASME Conference on Design Theory and Methodology, New York, NY, Paper No. DETC2008-50063 Koutamanis, A (2006) Buildings and affordances in J S Gero (ed) Intl conference on design computing and cognition, Springer, New York pp 345e364 Kumlin, R (1995) Architectural programming: creative techniques for design professionals McGraw-Hill Professional, New York Kruft, H (1994) A history of architectural theory from Vitruvius to the present Princeton Architectural Press, New York Lawton, M P (1982) Competence, environmental press and adaptation in M P Lawton, P G Windley and T O Byerts (eds) Aging and environment: theoretical approaches, Springer, New York Lawton, M P and Nahemow, L (1973) Ecology and the aging process in C Eisdorfer and M P Lawton (eds) Psychology of adult development and aging, American Psychological Association, Washington, DC Levy, M and Salvadori, M (1994) Why buildings fall down W.W. Norton & Company, New York Lewin, K (1951) Field theory in social sciences University of Chicago Press, Chicago Maier, J R A (2005) Foundations of aordance based design, Ph.D. Dissertation, Department of Mechanical Engineering, Clemson University, Clemson, SC 29634 Maier, J R A, et al (2007a) A computational framework for semantically rich design problems based on the theory of affordances and exemplar technology in International Conference on Engineering Design, Paris, France, Paper No. ICED07/510 Maier, J R A, et al (2007b) The affordance structure matrix e a concept exploration and attention directing tool for affordance based design, in ASME

412

Design Studies Vol 30 No. 4 July 2009

Conference on Design Theory and Methodology, Las Vegas, NV, Paper No. DETC2007-34526 Maier, J R A, et al (2008) Extending the affordance structure matrix e mapping design structure and requirements to behavior, in 10th International Design Structure Matrix Conference, DSM 08, Stockholm, Sweden Maier, J R A and Fadel, G M (2001) Affordance: the fundamental concept in engineering design, in ASME Conference on Design Theory and Methodology, Pittsburgh, PA, Paper No. DETC2001/DTM-21700 Maier, J R A and Fadel, G M (2002) Comparing function and affordance as bases for design, in ASME Conference on Design Theory and Methodology, Montreal, Canada, Paper No. DETC2002/DTM-34029 Maier, J R A and Fadel, G M (2003) Affordance-based methods for design, in ASME Conference on Design Theory and Methodology, Chicago, IL, Paper No. DETC2003/DTM-48673 Maier, J R A and Fadel, G M (2005) A case study contrasting german systematic engineering design with affordance based design, in ASME Conference on Design Theory and Methodology, Long Beach, CA, Paper No. DETC2005-84954 Maier, J R A and Fadel, G M (2007) Identifying affordances in International Conference on Engineering Design, Paris, France, Paper No. ICED07/591 Maier, J R A and Fadel, G M Aordance based design: a relational theory for design, Research in Engineering Design, in press, doi:10.1007/s00163-008-0060-3. Moore, C W and Bloomer, K C (1977) Body, memory and architecture Yale University Press, New Haven, CT Moore, G T (1987) Environment and behavior research in North America: history development, and unresolved issues in Handbook of environmental psychology, John Wiley & Sons, New York pp 1359e1410 Murphy, R R (1999) Case studies of applying Gibsons ecological approach to mobile robots, IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans Vol 29 No 1 pp 105e111 Newman, O (1996) Creating defensible space U.S. Department of Housing and Urban Development, Washington, DC Norman, D A (1988) The design of everyday things Currency Doubleday, New York Norman, D A (1999) Affordance, conventions and design, Interactions 1999; May/June Issue pp 38e43 Pallasmaa, J (1986) The geometry of feeling: a look at the phenomenology of architecture Chapter 9, in K Nesbitt (ed) Theorizing a new agenda for architecture. An anthology of architectural theory 1965e1995, Princeton Architectural Press, New York Pena, W (2001) Problem seeking: an architectural programming primer (4th edn) John Wiley & Sons, New York Pirsig, R (1974) Zen and the art of motorcycle maintenance Quill William Morrow, New York Schulz, N (1976) The phenomenon of place Rizzoli, New York Schulz, N (1988) Architecture, meaning and place: selected essays Rizzoli, New York Senge, P M (1990) The fth discipline: the art and practice of the learning organization Currency Doubleday, New York Tuan, Y (1980) Rootedness versus sense of place, Landscape Vol 24 No 1 pp 3e8 Tweed, C (2001) Highlighting the affordances of designs, in B de Vries, J van Leeuwen and H Achten (eds) Computer aided architectural design futures 2001, Kluwer, Dordrecht pp 681e696 Venturi, R (1966) Complexity and contradictions in architecture Museum of Modern Art/Doubleday, Garden City, NY

An aordance-based approach to architectural theory, design, and practice

413

Vicente, K and Rasmussen, J (1992) Ecological interface design: theoretical foundations, IEEE Transactions on Systems, Man, and Cybernetics Vol 22 No 4 pp 589e606 Vitruvius (1960) The ten books on architecture (M. Morgan, Trans.) Dover, New York Warren, W H (1995) Constructing an econiche, in J Flach, P Hancock, J Caird and K Vicente (eds) Global perspectives on the ecology of humanemachine systems, Lawrence Erlbaum Associates, Hillsdale, NJ

414

Design Studies Vol 30 No. 4 July 2009