Knowledge-based Design

Abstract

An implemented design of an urban area imposes long-term conditions on societal processes, such as the opportunities people have to organize their lives in temporospatial respects in a healthy and safe living environment, and the way social, cultural and economic institutions and organisations can function. In view of the fact that both people and institutions experience recurrent problems - ranging from getting lost in new housing estates to the awkward accessibility of workplaces - it is striking that in a world as ours laden as it is with scientific knowledge and its applications, the design/redesign of urban areas is based on scarcely any substantive-scientific knowledge in the area of urban & regional design. Designers are supplied with knowledge, particularly from social-spatial sciences, which is explicit, well-founded and open to critique, but design itself is not considered to be scientific. As to the professional field, there has been very little concern to develop a scientific foundation for urban & regional design, particularly during the last few decades. The widely held notion that each design is unique and based on individual creativity has hitherto left little room for thinking about urban & regional design as a science, whilst the immense complexity of urban areas plays a role as well. The assumptions underlying my research are that urban & regional design can be developed into a societally relevant science, that this depends on the view held regarding the significance of urban & regional design to society, and what is considered to be the object of the discipline derived from this view. I base these assumptions on the knowledge and insights I have acquired during the last fifteen years; the first ten years within the Chair of Urban & Regional Design, and after that within the Chair of Spatial Planning, both of the Faculty of Architecture of the Delft University of Technology. The research can therefore be characterised as an interpretative-theoretical study, a term from the methodologist A.D.de Groot ((1961) 1968: 325ff.). The characteristics of such a study are that within a particular collection of data (tentative) connections are made, that it must be impossible to solve the problem directly by experimental tests, and that the interpretation is not the only one possible. This makes the research an 'intellectual quest'. The first question that needs to be answered is where urban & regional design fits into the field of sciences, if a substantive-scientific approach is indeed possible. The standpoint from which I start my quest is that the real world exists independently of us as knowing subjects. As the cognitive power of human beings is limited by 'nature' and 'nurture', the real world is only knowable by approximation (the correspondence theory of truth). People are selective in their perception in accordance with general organization principles of simplification, categorisation and generalisation. They make connections, so forming a picture of how reality 'fits together', or can be fitted together. Reality, including urban reality, can therefore be approached as an (open) system, or a system of (open) systems of equal and unequal order. Elements of a system derive their significance (location value) from the position they hold in that system. Processes in a system can be either linear or cyclic. Whether changes are perceived in a spatial or temporal sense depends on the spatial or temporal grain of the perception (Jong 1992: 16). Within urban systems we can distinguish physical urban systems, made up of spatial elements such as buildings, streets, parks, sewers, stations, or made up of configurations of spatial elements like buildings, streets etc. which have certain characteristics in terms of form, physical state and function. These urban spatial objects, in mutually coherent combinations and in coherent combinations with natural spatial objects, have been and are constructed or reconstructed in order to fulfil a carrying function and an information function, on behalf of the urban society. The visual manifestation at a particular moment is called the composition of the urban area or urban landscape. Scientific knowledge is based upon rational considerations. Language, which includes visual language, has an organizational function in the thought process and is a means for conveying scientific ideas. Science limits 'chance' in the sense of 'random events'. Science presupposes generalization. For this purpose we must simplify systems and focus on similarities rather than differences. In order to communicate about and/or reflect on these systems we have to use models: conscious simplifications of (past, present and future) reality. Relevant to urban & regional design are pragmatic (analogue) models, particularly spatial ones, and from a functional viewpoint particularly descriptive, intentional-projective and exploratory-projective models. I therefore do not regard a model as an 'example to follow' as some (if not most) architects do. An urban or regional design is a proposal for a coherent package of spatial interventions in a certain urban or urbanescent area, and always affects more than one sector. Sciences may be divided into: (1) formal sciences without empirical content, (2) empirical sciences which concentrate on 'that which is (or was) the case' and therefore 'that which will probably be the case' and (3) practical, action-oriented sciences that have the application of science as their object of scientific research (Peursen 1986: 61) (Fig.A). The findings of practical sciences are then applied in concrete cases (Fig.B). Practical science is not knowledge that is acquired in practice (Drenth 1995: 157; Gunsteren 2001): practice only generates questions. The ultimate question for the practical sciences is 'does it work? i.e. which effects are to be expected and under what conditions. This involves both insight into constructive options and utilitarian options. Given the extrascientific problem statement a monodisciplinary approach is unlikely to be fruitful. On the basis of the above we can conclude that urban & regional design as a science would have to be categorised among the practical sciences. The same scientific rules and standards apply in empirical and practical sciences. In the views regarding these rules two main approaches can be distinguished: an objectivistic one and a subjectivistic/relativistic one, the primary exponents of which are, respectively, Karl Popper and Thomas Kuhn. Kuhn's conclusion that science should be conducted in a primarily non-rational, consensus-driven manner may be correct in a descriptive sense, but from a realistic view of scientific knowledge - the only one possible for practical sciences - this is not tenable as a goal. However 'natural' the inductive verification of hypotheses/theories may be, we do not necessarily have to follow this tendency, no more than we do the tendency to prefer 'certainty' over '' doubt'. Popper's views in their strictest sense are not tenable either. From an objectivistic point of view criticism has primarily been levelled at the fact that he regards the context of discovery as scientifically irrelevant: also hypotheses and theories can have a rational foundation. Popper's student, the objectivist Imre Lakatos - to some extent influenced by Kuhn - also criticised in particular the 'unnaturalness' of the exclusive focus on falsification instead of on verification/corroboration. This led him to develop a research approach, which due to his emphasis on heuristics, abduction and plausible reasoning, shows interesting similarities with (urban & regional) design processes. This approach offers perspectives for the development of urban & regional design into a practical-scientific discipline. In practical sciences the context of justification is however viewed in a different light from in empirical sciences. Ethical and financial considerations as well as the time factor may make it impossible to test a practical science hypothesis experimentally, under controlled, repeatable conditions. In these cases one will have to draw plausible conclusions on the basis of a series of applications, regarding the necessary conditions and effects that arise. Because the term 'context of justification' has much less significance in these cases, the term 'application' is preferable. In urban & regional design the above-mentioned testing limitations apply in a cumulative manner. Acquiring information from the context of application moreover is hindered because the conditions under which proposals are implemented in practice show relatively few similarities and these conditions cannot be manipulated. Based on this, the emphasis in the practical-scientific approach to urban & regional design lies on the context of discovery: what is assumed to be possible, and what are the probable effects, under which conditions. Empirical and formal scientific knowledge, in part derived from the context of application, should provide the necessary constraints. The next question is whether, in the development or non-development of 'urban and regional design' into a substantive-scientific discipline, what one regards as the object of that discipline - arising from what one considers to be the significance of urban & regional design to society - plays a part. To answer that question I position urban & regional design in relation to the disciplines of 'spatial planning' and 'architecture'. Lines of approach are (1) the systems approach to (future) urban reality, (2) the limitations inherent to working with spatial models of which the most important is the time factor and consequently that processes can only be shown indirectly (Fig.C), and (3) an examination in further detail of the concept of the 'carrying function' and the 'information function' (subjectively use value and experiential value). With regard to the latter two terms I argue, for example, that to be able to experience a physical urban system, unlike a building, the ability to use that system is a necessary condition. This positioning makes it clear that the general description of the physical urban system as the object of urban & regional design can be interpreted in two ways and in practice is indeed interpreted in two ways. The physical urban system can be seen on the one hand as an autonomous system, with the other components of the urban system as the system environment, and on the other hand as an inextricable component of the urban system as a whole. The first approach mentioned focuses on the composition of the physical system and on the linear processes in this system (characterized by a large temporal grain). I call this the pattern-oriented approach. This type of urban & regional design focuses on the so-called 'transformation' of urban areas. Influenced by the discipline 'architecture', and usually based on a quantitative programme of functional requirements, the creation of an experiential value guided by personal form concepts is seen as the main task for urban & regional designers. In part due to this emphasis on the design, pattern-oriented designs have the character of blueprints, which makes them fairly inflexible. In the second approach mentioned, which I call process-oriented, the focus is primarily on small-grained cyclic urban-societal processes with a spatial dimension. These processes do of course also have a large grain linear component. This design view emphasises the use value, with the experiential value as an essential, functional support of this value. Important components are the functional-spatial structures, which are supported by relevant visual design that facilitates any desired processes, and the potential user bases needed for the functioning of collective institutions. For process-oriented design it is sufficient to indicate the functional-spatial structure and a number of essential indications with regard to visual design. Urban & regional designers are insufficiently aware of distinction between these two types of approach. In addition to having shared concepts with similar definitions, pattern-oriented and process-oriented designs each have their own terminology as well as different definitions for the same concept. They also make use of different types of spatial models. This results in a lack of understanding in the field, confusion in the language used and insufficient insight into the societal significance of urban & regional design. We can conclude that the perspectives for a practical-scientific approach of urban & regional design differ according to which standpoint is adopted. Pattern-oriented design offers little perspective for a practical-scientific approach. We cannot ask 'does it work?' with regard to possible uses, unless the elements of physical urban systems are seen in mutual relation. 'Perception' therefore also has no functional significance. What is more, the emphasis lies on cultural and aesthetic aspects as well as personal form concepts. This is different for the process-oriented approach in which the physical urban system cannot be regarded as We can conclude that the perspectives for a practical-scientific approach of urban & regional design differ according to which standpoint is adopted. Pattern-oriented design offers little perspective for a practical-scientific approach. We cannot ask 'does it work?' with regard to possible uses, unless the elements of physical urban systems are seen in mutual relation. 'Perception' therefore also has no functional significance. What is more, the emphasis lies on cultural and aesthetic aspects as well as personal form concepts. This is different for the process-oriented approach in which the physical urban system cannot be regarded as separate from the urban system as a whole, and in which the use value is the primary point of interest and the perception of the physical urban system supports the use. I concretize these perspectives by outlining the relationship between urban & regional design and research in practice. In general, when making proposals for the spatial development of cities, there is a division of tasks whereby empirical scientists supply knowledge and insight into spatial planning, which is integrated by designers in a 'creative leap' into a design for a specific situation. In the Netherlands the term 'the unity of town planning' is applied (Lohuizen 1948: 3). This division of tasks also applies to means-oriented design, whereby the possibilities of the situation provide direction and the design result is evaluated ex ante. In practice, however, there seems to be a gap between empirical science and urban & regional design. There are a number of reasons for this 'applicability gap'. An important one is that increasingly more knowledge of a varied nature has become available, and designers as a consequence no longer have a comprehensive grasp of this knowledge, the more so as much of the information is irrelevant (Hillier, Musgrove & O'Sullivan 1972). Designers are consequently unable to let go of their preconceptions; on the contrary, they become more dependent upon them (ibid.). In the nineteen sixties the rising interest in the scientific approach to urban & regional design focused not only on the procedural side of design but also for a while on the substantive side (e.g. J. Jacobs 1961; Doxiadis 1968; Alexander 1977). However, there was and still is much resistance against the scientific approach to urban ®ional design: a rational, systematic approach is thought to adversely affect the essential creativity needed when making designs. Also the guild-like manner in which the community of urban & regional designers is organized, similarly to that of architects, does not stimulate the development of a scientific approach. An education and working climate that is characterized by a master-apprentice relationship, in which often no account is given of the resulting product, where the validity of claims are derived from the status of the speaker, and where debates regarding views held in the field are avoided rather than sought, is not the most conducive for a scientific development of that discipline. For so far as (realistic) scientific research is carried out in this context, it is almost without exception empirical descriptive research of a (cultural) historical nature: design research. For practical-scientific research one must not think in terms of each design being unique; in other words it is necessary to dissociate the object of design from the specific design context. This opens the doors for the design of theoretical models with spatial organization principles as 'building blocks': designs that in spatial-ecological and/or socio-cultural and/or economic-technical terms are independent of the situation. The activity of design acts to serve research and has become a research method: research by design. In research by design knowledge is not integrated directly and individually into each localized design, but general, integrated scientific urban & regional design knowledge is developed in an additional design phase. Explicit scientific knowledge is essential in this practical-scientific approach to urban & regional design as this makes a critical-rational debate regarding this knowledge possible. Theoretical models bridge the previously mentioned applicability gap. Creativity is crucial in both the development of this knowledge and its application in specific situations. If the research approach of Lakatos is related to the approach developed at the Chair of Urban & Regional Design, then I regard the following as elements of the hard core of the practical-scientific research programme: · Viewing the built (future) reality as an open system; · Approaching this physical urban system as an organized complex system; · Viewing the physical urban system as part of the urban system as a whole; · The fact that an element of this physical system derives its significance from its position in the system on the one hand and contributes to making the system what it is on the other; · The distinguishing of various temporal grains in societal processes; · The distinguishing of levels of scale within the physical urban system on the basis of societal processes that are characterised by a relatively small temporal grain; · The consequently necessary cohesion between the system levels; · The boundaries of design areas at various scales being defined on the bases of societal processes; and · Unlinking the design object from a specific design context; · Regarding design not just in the usual sense but also as a method of research. The essence of the research approach can be described as follows. Start with a number of basic elements from the object under study and manipulate them, in part on the basis of organizational principles, in such a way that the resulting theoretical models of physical urban systems are logically plausible and internally-consistent: constructions which, in the light of our available formal and empirical knowledge, are likely to function stably when implemented. Contextual conditions and effects analyses then have to be carried out for these basic theoretical models, in part based on empirical research. During the process of research by design there is also a continual ex ante evaluation. In order to limit the theoretically infinite number of possibilities, the breadth of useful research is determined by situations that occur in reality. Counter examples (Lakatos's 'monsters') play an important part: they increase the theoretical content of theoretical models. Urban & regional design entails the question of whether plausible spatial planning principles and plausible theoretical models can be derived from partially inconsistent information with heuristic and creative abduction as mechanisms (Schomburg 1991: 59; Magnani 2001: 78). Theoretical models are not ready-made templates for creating localized designs, but 'tools'. The task of the designer of a localized design is on the one hand to retain as much quality as possible of the chosen theoretical model - this can even be expanded with the help of specific situational potentials - and on the other to utilize the spatial individuality of the site in the design. The purpose of the latter is to bring about the spatial diversity that is necessary even if it is only to make the most of the information function. Theoretical models can be seen metaphorically as being made of elastic. The intended effects will have to be continually checked in concrete situations as the actual environment is not considered in the theoretical model. A designer should also consider any unintended consequences. Theoretical models can also play a role during a localized design process in the sense that a localized design problem is 'taken from' the specific situation and generalised. This is the transition area between 'research by design'; and 'research driven research'. A conscious simplification facilitates the studying of the (hypothesized) essence of the problem and establishes a relationship with generic urban & regional design knowledge. For several years now a debate has been going on at the Delft University of Technology whether a design can be regarded as scientific output. This debate seems to deal with the question of whether a design, a spatial model, is an acceptable means of communication. The debate in fact deals, or should deal, with the question of whether urban & regional design is indeed a (practical) scientific discipline. As there are limited possibilities for proving hypotheses this question ought to be taken seriously. The research approach described here together with the examples in this book of concrete research projects, allows this question to be answered affirmatively. This is supported by examples of practical scientific knowledge, dating mainly from the nineteen sixties and seventies, in the form of organizational principles and theoretical models (or initiatives in that direction). The research projects by academic staff and students described in this book also show that the research is bearing fruit, that in the terms of Lakatos there is a 'positive heuristic'. It cannot be concluded however, that this research programme is the only possible one. An interpretive-theoretical model study does not produce research results that exclude any other result. This makes this type of empirical research similar to design, in that it is a process that always has more than one possible outcome. More and better knowledge regarding urban & regional design, however useful in light of the spatial problems and consequently with regard to the functioning of people and institutions, does not in itself improve the (future) spatial situation. In a democratic context it is not the experts who determine what happens, but the elected administrators whom these experts supply with knowledge and insights. More and better knowledge does not necessarily mean that better decisions will be made. My findings lead to recommendations for university research and education. The most important of these is that a change in culture is necessary if urban & regional design is to be scientifically approached. The guild-like culture that characterises urban & regional design education should therefore be transformed into a culture in which general scientific rules are applied and taught, in which lecturers possess knowledge about the various views regarding science, in particular in relation to urban & regional design, and in which critical debates are encouraged. From a substantive viewpoint, not only spatial-ecological and economic-technological aspects but socio-cultural aspects of urban & regional design should receive attention. This attention should concern linear as well as cyclic processes including changes in these processes. Organizational principles and theoretical models should be further developed. Knowledge from other fields of science should be 'translated' into forms that are suitable for research by design. In view of the cross fertilisation between education and research - as shown in this book - the two have to be considered in close connection. Special attention should be given to those students who show interest in a scientific approach to the discipline. It is after all these students in particular who will help feed the body of knowledge of urban & regional design.