Co-Designing Social Systems by Designing Technical Artifacts
A Conceptual Approach
Ulrich Krohs
Abstract Technical artifacts are embedded in social systems and, to some extent, even shape them. This chapter inquires, then, whether designing artifacts may be regarded as a contribution to social design. I explicate a concept of general design that conceives design as the type fixation of a complex entity. This allows for an analysis of different contributions to the design of social systems without favoring the intended effects of artifacts on a system over those effects that actually show up. First, the clear-cut case of socio-technical systems is considered. Here, functions of artifacts can be planned fairly precise. In societies, in contrast, the actual functions of an artifact can hardly be predicted, which is due to strong self-organizing processes. Nevertheless artifact design can be shown to contribute to the design of the system also in this case.
Different bodies attempt to design social systems. Among them are governments, political parties, media, and economic enterprises, and at the level of individuals: politicians, journalists and businessmen, and also proponents and followers of theories of Social Systems Design (SSD). Besides being formed by such intentional influences, society shapes itself to a large extent via non-intended, self-organizing processes. So the design of social systems, as far as it exists, is probably best described as a hybrid, resulting in part from intentional and in part from non-intentional processes. The dichotomy of intentional and non-intentional design is well known from other areas, paradigmatically from the design of technical artifacts on the one hand, and from the design of biological organisms on the other. With respect to technical artifacts, the design process is an intentional one in which goals are followed. In contrast, there is no intentionality involved in the processes that shape the design of organisms: biological evolution is non-intentional. As outcomes of the different kinds of design processes, there are at least two different kinds of design: one of the kinds is intentional design, as the design of an artifact, which may be laid down in a construction plan,
U. Krohs, Konrad Lorenz Institute for Evolution & Cognition Research, Altenberg/University of Hamburg provided that conventions exist about how to interpret and to realize the plan, which again is an intentional process. Biological or natural design forms a second kind and should clearly not be understood as referring to intentions. According to neo-Darwinian biological theories, the design of an organism is laid down mainly in its DNA.98 I take it that the term “design” is used correctly in both cases, despite the lack of intentional-ity on the side of organismic design.99 This means that the different cases are assumed to have some important commonality. We seem to refer to a core meaning of “design” that is conserved in both uses of the term. To capture this core meaning, I will develop a concept of general design that includes both intentional and natural design. This will be done in the second section of my chapter.
The concept of general design shall be applied to social systems. It seems most workable to start with well-defined systems. In the third section of my chapter, I will therefore take a look at the design of socio-technical systems. These are systems like factories and similar enterprises that clearly have a prominent technological component. The paradigmatic example of such a system is a coalmine, which was investigated by members of the Tavistock Institute when they first introduced the concept of a socio-technical system. Such a system is made up of the machines, the workers, the administration, and their more or less institutionalized interactions (Trist and Bamford, 1951; Emery and Trist, 1960). The machines may serve functions in the system that would hardly be realizable without them; but the functions alone do not make up the system. Though many contemporary sociological approaches neglect the significance of the materiality of a system,100 functions crucially depend on a bearer. To make my point, I must refer to early functionalists like Malinowski, Merton, and Parsons, who emphasized the role of the material components of social systems: “no organized system of activities is possible without a physical basis and without the equipment of artifacts” (Malinowski, 1941, 68).101 However, talking about the functions of the components of a system requires an explication of the concept of function. Usually, the function of an artifact is regarded as being grounded in, or elsewhere linked to the goals of the designer. This seems to be too strong a requirement, since one also talks about functions with respect to components of biological organisms, where no reference is made to any intended goal. The concept of biological function is often based on that of design (e.g., Kitcher, 1993), and the non-intentional concept of general design allows therefore for a definition of functions that can be applied to the intentional case of technical artifacts as well as to possible non-intentional cases of functions in societies.
The structure of a socio-technical system and the functions of its components may come quite close to what was intended by those who had designed it. Therefore, a socio-technical system may be regarded as a designed one without much deduction. The situation may be different for larger social systems, like societies, to which I will proceed in the fourth section. Societies are planned to a much lesser extent than socio-technical systems. Nevertheless, the structure of a society will rely to a considerable extent on planned factors, since it is influenced by the constitution of the society, by laws, institutions, etc. Moreover, the structure of a society will be influenced by the design of the machines used by its members and by the design of the socio-technical systems that are embedded in it. As Merton states, “[n]ew applications of science to production by the engineer ... are inescapably social decisions affecting the routines and satisfactions of men at work on the machine and, in their larger reaches, shaping the very organization of the economy and society” (1947, 567). Some of these influences of artifact design on society and some functions of artifacts in society may be intended. Nevertheless, additional, non-intended effects will occur in many cases. Therefore, if such larger social systems are at least in part designed systems, which will be shown in section four, we are confronted again with non-intentional - or at least partly non-intentional - design.
There is no canonical conceptual framework that allows us to deal equally well with the different sorts of design that are related to different classes of functionally organized entities. I aim for a unified rather than a separating view: it seems to be plausible that, if we have three or four classes in which function and design go together in a similar way, then a commonality on the conceptual level can be expected. If we do not rely on such commonalities, we forego the chance to learn from one field with respect to the other.
Non-intentional design, being the more general case, can be found in biological systems. Most concepts of biological design focus on the design process (Allen and Bekoff, 1995; Buller, 2002). That reference to the design history is essential is often taken for granted in the case of artifacts as well (e.g., Lewens, 2004, 51-52).102 At first view it seems obvious to refer to the design process: all important decisions with respect to the final product are made within this process, and here is the place where goals are considered that have to be met by the product. Consequently I had to refer to the design process in the last section. However, any account that was to identify design with the process of designing would have insurmountable shortcomings. First, two convergent design processes may yield the same result. There might be many different ways to come up with the identical design of a technical artifact, like a chair or a combustion engine. The order of many steps in the process may be inverted, processes may branch or some process may bypass another. As long as the processes converge, the result will be identical, and the result matters with respect to the designed entity, not the way by which it was reached. Only the distinction between design and design process allows us to speak about identical results being reached in different ways. Second, we say that the design of, e.g., a car may be modified. This does not mean that the process of designing may be modified in a retrospective manner; even a Huxleyan ministry of truth can only mock a changed past rather than really change it. What we mean when we talk about a modification of a design is that a new design process starts from the results of a previous one, resulting in a different design. So, again, the design of an entity should not be identified with the process of designing. Instead, it has to be conceived as the outcome of the design process (Davies, 2001, 61-62; Krohs, 2004, chap. 4; Krohs, 2007). But what is the outcome? Sometimes, it is assumed to be the structure or internal organization of a complex entity (e.g., Lauder, 1982), but if the design really was the internal organization of the entity, we would also have to talk about the design of the solar system and other organized purely physical entities, because the organization of a non-designed entity does not necessarily differ very much from the organization of a designed entity. Consider cloud streets or sand ripples in the sea as highly organized but non-designed structures, or compare the organization of the solar system with that of a (perhaps very particular) carousel. So design should neither be identified with the process of designing, nor conceived as the structure or organization of a designed entity. Design rather seems to be something that mediates between these two.
98
The neo-Darwinian research program relies on genetic determinism. The perspective had to be broadened by reference to epigenetic contributions to inheritance (cf., e.g., Jablonka and Lamb, 2005). In current biological research programs that integrate developmental with evolutionary processes, the focus is shifted from inherited design to developmental processes, which are now conceived as being at the center of the generation of biological form (Muller and Newman, 2003).
99
Since biological design is to be conceived as non-intentional, the concept of design discussed here has no affinity at all to the notion of “intelligent design”, which has been made the topic of many unfortunate political debates.
100
Functionalist accounts of social systems that follow Luhmann consider systems as being constituted of communicative interactions only, not of material components (Ropohl (1999) develops a formalized version of an act-focused sociological approach). Likewise, Searle, in his intentionalist conception of society, does not count artifacts as components of societies, though speaking about the assignment of functions to them (1995, 13-23). His ontology of social reality embraces only the following three “elements”, as he calls it: the assignment of function, of collective intentionality, and of constitutive rules (1995, 13, 29).
101
The importance of function bearers is reconsidered in some recent approaches. Callon and Latour’s Actor-Network-Theory and Pickering have a strong focus on material agency (e.g., Callon, 1986; Latour, 1988; Pickering, 1995), but their frameworks are hardly suitable for looking for similarities between social and other systems.
102
A different view is put forward by Houkes et al. (2002) but since this approach is applicable in the realm of intentional design only, it is too restricted to account for the partly non-intentional design of social systems.