Complexity and Technology
Science & Technology Dynamics
University of Amsterdam, The Netherlands
A shorter version of this paper is forthcoming in the Encyclopedia of Life Support Systems - Theme section KM 1.29- Knowledge Management, Organizational Intelligence and Learning, and Complexity, UNESCO: Paris, 2000.
The study of technology is itself a complex issue. First, contrary to biological species technologies are not given in nature, but man-made constructs; they are the products of cultural evolution. The various actors involved may use different definitions of technology. Furthermore, technologies are continuously evolving. With the further development of technologies, their definitions and relevant perspectives may also have to change. These definitions and perspectives, however, are basic to the discursive traditions studying technology and its relevant contexts.
I distinguish three main perspectives in the study of technology: (1) economics, (2) the history of technology and the sociology of science, and (3) science and technology (S&T) policy analysis and R&D management. I shall argue that the combination of these three perspectives challenges us to consider technology as a subject of complexity studies. From this combined perspective, I then proceed to specify the role of technological change in reshaping the relations between nature, culture, and society in a knowledge-based economy.
2. Prevailing perspectives
2.1. Technology as a subject of economics
The primary function of markets is to resolve imbalances: differences between supply and demand can be equalized by the price mechanism. Market clearing operates as a dynamics, yet at specific moments in time. Accordingly, economic theorizing focused on technological developments initially in terms of (so-called) comparative static analysis: how can the a posteriori configuration be compared with one or more prior configurations?
One convenient way to model the operation of the market in the case of a choice among (available) technologies has been provided by the production function. The production function represents the relation of output to input factors (e.g., labour, capital, raw materials). When prices change, an entrepreneur is able to substitute among input factors using different techniques. Thus, the model assumes a single optimum given factor prices. The production function is then a hyperbole (e.g., Output = c * Capital * Labour) as depicted in Figure 1.
The optimal technology at a certain moment in time can be found by drawing the tangent of this hyperbole with the straight line representing the current ratio between factor prices.1 Schumpeter (1939, at p. 87) introduced the analytical distinction between shifts along the production function indicating factor substitution, and shifts of the production function towards the origin, as an indicator of ongoing technological development. In general, technological innovation and development allow for the same output using less inputs.
Although the later Schumpeter (1966) emphasized that technological developments are also driven by economic factors (e.g., factor prices), the distinction has been fruitful both for empirical research (e.g., Solow 1957; Salter 1960) and for theoretical developments in this domain (e.g., Nelson and Winter 1977; Sahal 1981). The shifts of the production function, however, could not be explained in neo-classical economics. From this perspective, technological progress had to be considered as an external given or a residual factor (Abramowitz 1956; Solow 1957; cf. OECD 1964).
In a seminal study, Nelson and Winter (1982) proposed to make technological developments endogenous to economic theorizing by using Markov chain models.2 These authors, however, also changed the unit of analysis. While the (neo-classical) price mechanism and (Schumpeter's) innovative dynamics had analyzed the economy as a system (Alchian 1950), Nelson and Winter (1982) attempted to explain technological developments in terms of longitudinal developments in firm behaviour (Andersen 1994). From this perspective, firms become the carriers of technologies and the driving force behind innovation.
Evolutionary economists have developed the discipline during the 1980s using concepts like trajectories and regimes (Dosi 1982). Models from evolutionary theorizing like predator/prey models (Lotka-Volterra) have been applied to economic phenomena (e.g., Saviotti and Metcalfe 1984; Sahal 1985). More recently, evolutionary economists have focused on the network level of the adopters of a technology, rather than on individual firms. The utility of using a particular technology increases with the number of adopters. Therefore, standardization can emerge spontaneously, leading to the "lock-in" of one technology. This phenomenon has been related to the emergence of dominant designs in the history of various industries.
One well-known example of a "lock-in" is the QWERTY keyboard (David 1985). This keyboard was engineered in order to optimize typing speed in the case of mechanical typewriting. It had been designed so that the typebars have a minimum chance of jamming given the character frequency distribution in the English language. Since mechanical typewriting is out of use, the QWERTY keyboard has become suboptimal. However, one is no longer able to break out of the lock-in given learning curves and network externalities (Arthur 1988; Leydesdorff & Van den Besselaar 1998).
2.2. History of technology and the sociology of science
Historians, scientists, and engineers with a common interest in the social contexts of the development of science and technology have provided us with a wealth of source materials about technologies and innovations. How are technologies shaped historically? How exogenous have technologies been to the development of the sciences? Or, from a very different perspective: How are user interests represented in the development of technological artifacts?
In his study America by Design (1977), the historian David Noble focused on the period 1870-1910 as the "scientific-technological revolution" in industrial production. He characterized this period as "the wedding of the sciences with the useful arts." While technologies before this time were mainly embodied in artifacts (e.g., the steam engine), corporations from then onwards were generated on the basis of engineering activities (e.g., Thomas Edison; cf. Hughes 1987); and, vice versa, large corporations began to develop their own R&D facilities (e.g., in the German chemical industry).
Patent legislation was a necessary complement to this development (Van den Belt & Rip 1987). The further development of the interface of science, technology, and markets has led to the emergence of a so-called "technostructure" both within large corporations and in state apparatuses during the first half of the 20th century (Galbraith 1967). "Scientific management" (Taylor 1911) feeds into the division between white-collar work concentrated within the technostructure and blue-collar work on the shop floor. In this configuration, technological developments remained relatively shielded from immediate consumer interests and oriented towards the longer-term planning of investments.
Sustained and institutionalized interaction between markets and sciences changes the codes in these communications mutually. Braverman (1974, at p. 167), for example, noted that "the key innovation is not to be found in chemistry, electronics, automatic machinery, aeronautics, atomic physics, or any of the products of these science-technologies, but rather in the transformation of science itself into capital." Conversely, the absorption of science by capital has gradually transformed the latter: the productive forces are no longer necessarily linked to the managerial decisions and instrumental actions of real people engaged in a labour process (Habermas 1968). One has to account increasingly for the interaction terms.
2.3. Technology policy analysis
Mission orientation in the development of science-based technologies has become a mode of operation since the Manhattan project in World War II. Under peaceful conditions it took the Sputnik shock (1957-1960) to generate science and technology policy analysis as a separate field of scholarly work. Policy analysts share with social historians of S&T their orientation towards the dynamics of the systems under study, but they share with economists a focus on choices in terms of present strengths and weaknesses.
The Organization for Economic Cooperation and Development (OECD) in Paris has had a leading role in shaping S&T policies since the 1960s. One of the main results of these efforts has been that one has for analytical reasons to differentiate among various sectors in society, disciplines of science, national cultures, etc. (e.g., OECD 1988). The issue is complex in terms of the subject matter (Nelson 1982 and 1993).
The differentiation between strategy and operational structure has been appreciated as constitutive of technology management (Chandler 1962; Galbraith and Nathanson 1978). Knowledge-intensive production processes cannot be managed using a single criterion (like for example, prices or growth figures) since optimization in different dimensions has to be traded off. For example, technological alternatives can be compared in terms of price/performance ratios. Economic selections can be discussed in terms of representations (e.g., utility functions); and scientific reflections and technological constraints can no longer be considered as exogenous to the economic system.
3. "Technology" as a complex phenomenon
This short summary of technology studies illustrates that "technology" is not a given nor can it be considered as a common denominator. Technologies are complex constructs that can be appreciated differently from various perspectives. In general, three main subdynamics have been distinguished: (i) the selection mechanism initially associated with the market, (ii) the historical generation of variation along the time axis, and (iii) the wish to control the adjustment mechanism (that is, the retention) in both the private and the public domain. The various forms of theoretical reflection open windows of appreciation onto this complex dynamics. The definitions have different meanings in the context of the relevant theoretical traditions (Leydesdorff 1997a).
Using an evolutionary metaphor, I propose to proceed by distinguishing between (i) the phenotypical complexity of technological developments and innovations and (ii) reflections on these developments. The theoretical reflections reduce the complexity by adopting a perspective. One may wish to consider the perspectives as "genotypical," while the "phenotype" is interactively constructed among engineers, economists, managers, users, and policy makers (cf. Langton 1989).
The initial reflections of, among others, engineers and entrepreneurs, can further be extended into specialized discourses. These technology studies are no longer directly coupled to the efficient use of technologies, but indirectly they study the effectiveness of technologies for the sustainability of firms, industries, and society. Note the importance of the specification of the systems of reference when studying this knowledge infrastructure. The distinction among the relevant systems of reference has itself been subject of literature. At which level, and in relation to which units of analysis, should one define technological developments (Lundvall 1998; Leydesdorff & Etzkowitz 1998a)?
Whereas economists have tended to consider firms and consumers as basic units of analysis, policy analysts are inclined to focus on nation states. Recently, however, it has become clear that regional innovation systems, or perhaps multi-national systems like the European Union, can be important players as well. A third unit of analysis is the specific technological development itself, for example, biotechnology or communication technologies. The development of new (generic) technologies may be intertwined with the emergence of new sectors of social production and distribution (e.g., McKelvey 1997).
In other words, technological developments are pervasive in modern cultures. While previous cultures (like high-cultures) were mainly static, capitalist economies are innovative (Marx 1848). Each innovation is historically based on incidental inventions, but the principle of innovation has become a major characteristic of capitalist societies. As noted, in order to be sustained as a mode of production, the various subdynamics have to reinforce one another. In addition to price competition at each moment in time, the dynamic principle of innovation is "continuously upsetting the movement towards equilibrium" (Nelson & Winter 1977). Along the time axis, the "creative destruction" of previous investments may set the system free for new waves of technological advancement (Schumpeter 1943).
This heartbeat of capitalism is not expected to tick with a single frequency. The time frames of economic activities, political programs, and scientific-technological developments are of different orders. Freeman & Perez (1988) associated the "long waves" in the economy with the emergence of new technologies. For example, the steam engine induced the industrial revolution, while the scientific-technological revolution of the late 19th century was mainly dominated by the chemical industry, electricity, and steel production (Rosenberg 1976). Currently, one is witnessing a revolution in information and communication technologies (e.g., Castells 1996).
4. The Endless Transition
Technological revolutions generate adjustment crises in the system of social reproduction, since new dimensions of wealth generation can be added to the system. When a new dimension is added to a complex system, all bets on the basis of its history are off. The system is entrained in a drift towards a new stabilization that, if successful, may become a global state (Kampmann et al. 1994).
Today, many analysts perceive themselves as in the middle of such a historical transition. The metaphor of a transition, however, suggests a bridge to a new state or a position at the forefront (Vernon 1972). The so-called "transition economies" of Eastern Europe and the FSU have taught the painful lesson that transition is a permanent condition, one that has become a hallmark of the knowledge-intensive innovation system (Etzkowitz & Leydesdorff 1998). Such a system cannot be equilibrated or hierarchically controled, although these subdynamics remain necessary for its progressive development.
For example, while patenting has been a social solution to the problem of protecting industrial property rights during the previous scientific-technological revolution, the Internet requires a new definition of the problem of intellectual property rights (David 1993). First, the system of legislation may try to adapt by incorporating forms of trade-marks or copyrights into the existing systems. In a further stage, the uncertainties that have not been sufficiently resolved may require the legislative system to undertake a more drastic reform. If intellectual property rights legislation remains locally behind, a subsystem may profit temporarily. The Netherlands, for example, postponed introducing patent legislation around the turn of the century in order to maintain competitive advantages based on colonial trade. However, in the longer term each system is under the threat of decay if it does not adapt to changing environments.
Thus, the cycles of production, technological innovation, and policy-making drive each other as in a triple helix. At the interfaces, mechanisms have to be developed in order to make the various subsystems sensitive to the need of a next update. When all the subsystems are in place and tendentially solving the problems of harmonization, a hypercycle that drives the knowledge-intesive economies can be sustained (Leydesdorff 1997b). If this fails, the system is endangered with crises of hyper-inflation, hyper-uncertainty, or desperate attempts to synchronize the different subdynamics from a hierarchically defined vantage point. The hierarchical solution (for example, by a strong state) may entail large social costs, since it can be expected to retard the experimental dynamics of evolutionary change.
5. The coevolution of technology, society, and culture
To better understand the dynamics of the social system and the role of technological developments therein, it is useful to distinguish between the macro- and the micro-perspective. Note that these levels are nested operations which should not be considered as having a hierarchical order. The macro-system is not a pre-given "natural selector" as in biology; it is a historical reconstruction from which micro-level agencies select the resources they need to improve their niches for competitive survival. Each selection generates new variations. What can be varied and what is operating as a selector may also vary over time, as in a co-evolution.
5.1 Macro-evolutionary change
What is evolving and what is co-evolving? Sociologists and economists have been fascinated with the transition from mediaeval culture to the modern social system. What has happened to the social system in evolutionary terms? How might this transition be explained? Marx, for example, discussed the advent of modernity in terms of changing class relations; Weber emphasized the function of Protestant values in the emerging social relations of modern capitalism. Others have pointed to tensions internal to the mediaeval system like the struggle over investiture (Luhmann, e.g., 1989, at pp. 262 ff.) or the revolution in communication brought about by the development of printing (Eisenstein, 1979; Kaufer & Carley, 1993).
All these different explanations assume that a transition at the system level should be analyzed historically, that is, in terms of ex ante causes. From an evolutionary perspective, however, the emerging new order is dependent on the specificity of the ex post selections. Selection operates in terms of functions, independently of underlying structural arrangements. Thus, the variations leading to a transition are less important in understanding the dynamics of a system than the selection mechanisms which emerge (Rosenberg, 1994). Furthermore, as the optimization proceeds, the underlying structures are under pressure to be adjusted, or redefined.
Functional differentiation was constructed in Europe on the basis of a century of (primarily religious) wars between 1550 and 1650. The episteme of the modern sciences (Foucault, 1972) can be distinguished from those of pre-modern times in terms of the functional differentiation and the universalistic orientation of communications. Note that functional differentiation requires two levels of communication: scientific communication, for example, has both a substantive value in itself and a function for the emerging higher-order system of "universal" theorizing. In principle, both substantive novelty (the "context of discovery") and methodological warrant (the "context of justification") can be distinguished analytically.
Institutional differentiation became possible when the order of modernity was fully established during the 18th century. The adjustment of underlying structures to functions under selection pressure can be considered as a process of adaptation. Adaptation is a gradual process taking place through selection from among possible formations. Therefore, institutional codification can be expected to lag behind the processes of functional differentiation of communication.
During the 18th century —with the possible exception of the United Kingdom— national state formation did lag: the organization of society remained in some places entrenched in the mediaeval differentiation of nobility (France, Austria, Spain, and Italy), while in other places regional differences opposed centralization into a nation state (Germany, The Netherlands). At the end of this century, however, the American and French revolutions established two nation states based on the semantics of an institutional differentiation between civic society ("the pursuit of happiness") and the national state (cf. Montesquieu, 1748). Although Napoleon tried to export the new ideas about political codification as "universal," development during the period 1815-1870 can be characterized with hindsight as the social elaboration of the differentiation between civic society and the various national states (e.g., Gouldner, 1976).
This institutional development went side by side with the transition from mercantile capitalism (oriented towards global markets) into industrial capitalism (locally organized). Marx noted this transition on the occasion of the revolutions of 1848. When he published Capital I in 1867, the new system was nearing completion. After the American Civil War (1861-65), the Meji-restoration in Japan (1867), the Paris Commune (1870), and the subsequent German and Italian processes of unification, the major nation systems were in place with their respective capitalist economies.
Thus, a dually differentiated system was established: the nation states contain institutional mechanisms which are able in principle to reinforce specific selections from the functional differentiations, for example, by maintaining national boundaries. While in the functionally differentiated system, the control mechanism is firmly based in universalistic assumptions, the complex system may periodically shift control to institutionally warranted structural elements (e.g., the bureaucracy). Given the dual-layeredness of these dynamics, control may become a focus in drift (Elzinga 1985). The system can also alternate between phases of contraction and expansive modernization (cf. Freeman & Perez, 1988).
The two dynamics can be combined in different ways. With hindsight, Marxism can be considered as an ideology that tends to take "real existing human needs" as its frame of reference, and therefore emphasizes the institutional realization of power in society. From a liberal perspective, "power" and state organization are considered as codified embodiments of a system of checks and balances with the aim of promoting the further ("free") development of the subsystems defined in terms of functions (economy, science, etc.).
When the organization of society has become so complex that two mechanisms are available for the integration, one should expect the possibility of a range of possible interactions. At specific sites the two selections may begin to reinforce each other, as in a resonance. Such "lock-ins" occur locally and discretely, and if stabilized, the resulting patterns can be expected to follow relatively independent trajectories. In other words, selections can be made at some places for stabilization.
Selection is a recursive operation. Selections at each moment in time (like in the case of a market) can be selected for stabilization over time. When a next selection operates on a distribution of historical stabilizations (which follow their trajectories), a "globalization" into a regime can be developed (Leydesdorff 1994). A regime is not directly observable: it remains pending in the form of selection pressure on observable realizations.
In other words, the codification of two selection mechanisms at the level of society makes it likely that a pattern of local interferences will emerge in a co-evolution between these codifications. The resulting patterns of communication can be considered as locally distributed "lock-ins" or niches. This distribution, however, contains the expectation of a next-order regime at the global level. The theoretical question remains whether one is able to specify the relevant co-evolutions, and then to articulate an expectation concerning the further development of this complex system.
5.2 The micro-system
The micro-system uses recombination for the generation of new dimensions within the existing relations. The institutional structures can then be considered as the epistatic retention of the social system. As noted, the institutional differentiation no longer coincides with the functional differentiation, while the former are under pressure from the latter to adjust. One can expect the functions to be communicated within the institutional subsystems more densely than among them (cf. Simon 1973; Kauffman 1993).
Although communication across institutional borders may initially be sparse and erratic, Arthur's (1988) model of the "lock-in" demonstrates how this (white) noise may lead to the development of a trajectory under the condition of increasing marginal returns. For example, in the case of a specific brand of a technology, network externalities and learning curves may lead to the codification of a dominant design. Arthur (1988) used the example of the VHS in the case of VCRs. After "lock-in," potentially superior technologies can no longer compete given the infrastructure in the market.
The knowledge infrastructure inherently contains mechanisms of increasing returns because of the specialization involved. Learning effects are pervasive when heuristics are also "locked-in." Trajectories are expected to emerge in technology/market combinations on the basis of the local and distributed interaction patterns between the functions and institutions noted above. Historically, the process may take a while. For example, after the founding period of airplane design by the Wright brothers (1903) a dominant design, namely the Douglas DC3, emerged only in 1936. But once it emerged, development was swift: steep learning curves could be based on more than 10,000 copies sold. McDonald Douglas's main competitor, Boeing, quickly adapted by bringing the 307 Stratoliner (1938) and the 377 Stratocruiser (1944) to the market at the expense of destroying in-house competences (Frenken & Leydesdorff, forthcoming).
As markets become more knowledge-intensive, as is historically the case during the 20th century, co-evolutions between technologies and institutions can increasingly be observed (Nelson 1994). Similarly, sectoral policies by national governments tend to shape technological trajectories, as in an energy household or, more recently, in biotechnology. In the latter case, market forces have been constitutive of the development as well. While a double helix is expected to stabilize a trajectory (as in a co-evolution), a triple helix of university-industry-government relations is expected to contain the species of chaotic behaviour (Leydesdorff & Etzkowitz 1998b; cf. Etzkowitz & Leydesdorff 1997). Thus, an endless transition can be generated endogenously.
The disturbance terms are crucial to the model (Foray 1998). First, the drift generated in the interactions may "lock-in" because what originated as a drift may (above a certain threshold) become recognizable as a signal from another perspective. Figure 2 shows this Arthur-model. The winner — which does not have to be the best technology when assessed from a longer term perspective — takes all.
It can be shown that "lock-ins" are extremely robust against major improvements in the other technologies or in terms of network effects. A "lock-out" can notably be forced by an innovation that makes the network effects of the prevailing technology obsolete. For example, when one is able to download movies directly from the Internet, the advantage of storing the information on VHS tapes disappears. Return to an equilibrium or capturing the market by a "lock-in" into another technology is possible, but under specific conditions (Leydesdorff & Van den Besselaar 1998).
If one extends these models with two or three sources of variation and for the case of more complex institutional arrangements, one is able to simulate double and triple helices (Leydesdorff, forthcoming). For example, one can assume that the two technologies do not compete in a single market, but in different submarkets or in different nation states (or perhaps, regions). The variations in these selection environments can be shown to speed up the finding of a match in the case of two weakly coupled dynamics (as exibited in Figure 3). If there is no coupling, the "lock-ins" follow consecutively (as in Figure 4). Thus, a double helix tends to enhance "lock-in."
If this configuration is extended to three sources of variation, a weak coupling (that is, by allowing for some hierarchical conditions between the co-evolutions using additional "if then"-statements) may disturb the consecutive patterns of "lock-in." Figure 5 exhibits a run of the simulation model in which an initially loosing technology "locks-in" at a later stage. Note that this implies that an initially winning technology can also lose the race for eventual "lock-in."
Figure 6 exhibits a similar run in which the competition continues to exist. Given that there are three sources of variation (N = 3) and all mutual relations are in place (K = 2), this corresponds to the envisaged possibility of two sub-optima in the analytically elaborated example of Kauffman (1993, at pp. 42f).3 Thus, the analytical model of epistatic relations of Kauffman (1993) can be related to the simulation models of Arthur (1988) for "lock-in" in the case of a triple helix (that is, N = 3 and K = 2). One condition for the complex behaviour of this system is that the mutual relations disturb one another. In other words, the three helices have to operate without prior synchronization so that a specific condition can be fulfilled on the basis of stochastic variations in their mutual interactions.
Synchronization occurs ex post, like a harmonization (Kampmann et al. 1994). Attempts to control ex ante, for example, in terms of a strong state intervention into the evolutionary dynamics of the system, are expected to lead to unintended consequences (Giddens 1984; Beck 1986). Reflexive policies are needed in which one signals and analyzes the potentials for further development versus possible bottlenecks. For example, one may have to adjust one's definition of the region relevant for technological development in a next update. One should appreciate the multi-layeredness and the complexity of the system as a potential resource for new recombinations. Niche management and human capital management are crucial instruments for creating technologies that sustain community formation in such configurations.
6. Technological innovation in complex systems
The helices select upon each other in asymmetrical relations. Selection, as noted, is a recursive operator: some selections are selected for stabilization; some stabilizations ("trajectories") can be selected for globalization ("regimes"). For example, the oil crises of the 1970s induced a search for more efficient use of fuel in combustion engines, but these crises did not essentially affect the regime of the car as a dominant transportation system. Catalysts for further combustion are defined as a problem along the trajectory of car engine development. High-speed trains, on the other hand, may help to revive the regime of public transportation.
In other words, in knowledge-based economies one needs both types of innovations. Theoretical specification of the systems of reference informs our expectations, although the systems are expected to change and thus to exhibit unexpected and unintended consequences. Still, the quality of the update of our expectations can be made the subject of study. Here, the knowledge-driven system closes in upon itself. The various theoretical perspectives disturb each other and may drive each other into a next update so long as (e.g., interdisciplinary) communication can be sustained. When a single perspective prevails (for example, that of a central state or short-term profit taking), the evolutionary dynamics can be blocked temporarily because communication is no longer encouraged among the relevant dimensions.
From this perspective, the complexity is endogenous to the system as an interactive result among various forms of communication. Each of the perspectives reduces the complexity by taking a specific angle. The perspective of complexity theory translates these substantive perspectives into an attempt to position and weigh them in order to achieve formalization into a model. The positive theories study cases which happened, that is, the deselected cases. The algorithmic model, however, has to be specified in terms of these negative selections.
Selection as a negative operation has the status of a hypothesis. Recursively, selections upon selections may generate positive stabilizations. The formal model thus abstracts from the substantive models on the basis of which it has to be built. The substantive theories provide a positive appreciation, while the formal model specifies in terms of assumed selections. The systems under study remain theoretical constructs necessary for an analytical understanding. Crucial for the development of the analytical perspective are questions like: What is expected to evolve? How is this done (that is, the modus operandi)?, and why (what is the evolutionary function)?
In an early attempt to specify selections in the case of technological trajectories and technological paradigms, Dosi (1982, at p. 152) specified the relevant selections as follows:
In broad analogy with the Kuhnian definition of a "scientific paradigm", we shall define a "technological paradigm" as a "model" and a "pattern" of solution of selected technological problems, based on selected principles derived from natural sciences and on selected material technologies.
I have argued on macro-sociological ground that markets, R&D, and reflexive control can be considered as the relevant systems of reference. This accords with Dosi's specification of the micro-operation: technological problems are generated in the market, the principles for their solution require R&D, and resources have to be organized at the relevant interfaces. The main analytical perspectives discussed above correspond to these three dynamics: the market is a clearing mechanism at each moment in time, while technological innovations upset the movement towards equilibrium, and the reflexive axis rules to provide control. This last factor is nowadays further differentiated in terms of the public/private interface in civic societies.
Both governments and entrepreneurs play a role in changing the knowledge infrastructure. Furthermore, the historical conditions of the system constrain and enable the complex dynamics at the phenotypical level. The model corresponds indeed with Marx's dialectics between production forces and production relations, but the perspective of an eventual resolution has to be given up (Schumpeter 1943). While continuously disturbing one another, the systems are entrained into resonances which fail to occur. This lack of closure challenges our inventiveness and thereby it liberates the forces of technological innovation for the creative destruction of previous production relations.
7. Normative implications
The institutional retention mechanisms can be expected to select upon the wealth of options that are set free by the various subdynamics and their interactions. Each selection changes the settings, thus driving the system potentially into a next update. The selective operations close in upon themselves in driving the self-organization of the complex system, that is, by replacing the given situation ("nature") with cultural constructs that are technologically based.
The normative implications become dependent on the (reasoned) choice of an analytical perspective. Is the (hypothesized) system in a trajectory state, in a regime state, or still "in the making"? During a period of gestation, a dominant design has yet to be developed. "Lock-in" can be expected to occur locally, that is, at the firm level. Then it may spread throughout the industry in a process of diffusion. Consequentially, the nature of the competition changes. One expects standardization and codification processes to drive the system thereafter into a regime state.
The various perspectives do not exclude one another. Different dynamics can be studied. As noted, the various dynamics are also expected to operate with different time horizons. Long waves in the economy are propelled by fundamental changes in generic technologies (Freeman & Perez 1988), while business cycles are superpositioned over them, and markets are expected to clear at each moment in time (Schumpeter 1939).
New combinations of selected technological problems, selected principles derived from natural sciences, and selected material technologies are sometimes able to change the epistatic relations more fundamentally than in terms of gradual improvements. The mature system can be dissolved only by radical innovations, which cannot be produced by focusing on existing research practices and network developments, since these subdynamics are integrated within the current system. Remember the example of downloading one's videomovies from the Internet, which makes the hitherto locked-in VHS system obsolete. The old paradigm will be active in resisting its replacement, since it is embedded in prevailing markets and social practices. However, technological unemployment is a consequence of insufficient awareness of the technological changes of the tide. Both the infrastructure and the skills have continuously to be updated and sustained (Pasinetti 1981).
From a complex systems perspective, the intensity of these technological and cultural transformations can be expected to follow a power-law distribution as in the case of, for example, earthquakes. One is not able to know in advance when and where the earth will shift, but one is able to specify the reflexive expectations and the uncertainties involved. If the system is assessed as in its upswing towards a trajectory, hierarchical controls and sharp choices can be rather effective. If the system is in a downswing, R&D input which focuses on unresolved problems and the creative destruction of oligopolistic arrangements may be most useful. Furthermore, the creation of layers of communication that allow existing systems to explore new dimensions of communications may be a wise policy. For example, the European focus on the "subsidiarity" of the EU network systems enables the partners to profit both from national advantages and from new options in emerging configurations.
In general, a technological culture is no longer able to return to a natural state for its integration. Integration has to be achieved through communication, while at the same time different perspectives are needed to sustain the complex dynamics. Each perspective tends to claim the lead of the system by black-boxing the competing ones. To the extent that integration across perspectives can be sustained, niches may enable us to develop our communicative skills in order to prepare for a next update (Leydesdorff 1999). Whereas one is historically locked-in into technological trajectories, regimes and paradigms guide the heuristics at a next-order systems level. These trajectories can be considered as the recursive terms in the non-linear model, while the regimes include also the interaction terms. Wealth generation, job creation, community formation, and technological change, however, are intertwined at the latest level. These expectations can only be specified theoretically.
Abramowitz, M. (1956). Research and Output Trends in the United States since 1870. American Economic Review 46, 5-23.
Alchian, A. A. (1950). Uncertainty, Evolution, and Economic Theory. Journal of Political Economy 58, 211-222.
Andersen, Esben Slot (1994). Evolutionary Economics: Post-Schumpeterian Contributions. London: Pinter.
Arthur, W. Brian (1988). Competing technologies. Pp. 590-607 in: Dosi et al. (1988).
Arthur, W. Brian (1989). Competing Technologies, Increasing Returns, and Lock-In by Historical Events, Economic Journal 99, 116-31.
Beck, Ulrich (1986). Risikogesellschaft. Auf dem Weg in eine andere Moderne. Frankfurt a.M.: Suhrkamp.
Bijker, Wiebe, Thomas P. Hughes, and Trevor Pinch (Eds.), (1987). The Social Construction of Technological Systems. Cambridge, MA: MIT.
Braverman, Harry (1974). Labor and Monopoly Capital. New York/London: Monthly Review Press.
Castells, Manuel (1996). The Rise of the Network Society. Melden, MA: Blackwell.
Chandler, Alfred D. (1962). Strategy and Structure. Cambridge, MA: MIT Press.
David, Paul A. (1985). Clio and the Economics of QWERTY, American Economic Review 75, 332-37.
David, Paul (1993). Intellectual property institutions and the panda's thumb: patents, copyrights and trade secrets in economic theory and history. In: M. B. Wallerstein et al. (Eds.), Global dimensions of intellectual property rights in science and technology. Washington, DC: National Academy of Sciences Press.
Dosi, Giovanni (1982). Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technical Change, Research Policy 11, 147-62.
Dosi, Giovanni, Christopher Freeman, Richard Nelson, Gerald Silverberg, and Luc Soete (Eds.), Technical Change and Economic Theory. London: Pinter.
Eisenstein, Elizabeth L. (1979). The printing press as an agent of change, communications, and cultural transformation in early modern Europe, Vols. 1-2. London: Cambridge University Press.
Elzinga, Aant (1985). Research, bureaucracy and the drift of epistemic criteria. Pp. 191-217 in: Björn Wittrock and Aant Elzinga (Eds.), The University Research System. Stockholm: Almqvist & Wicksell.
Etzkowitz, Henry & Loet Leydesdorff (Eds.) (1997). Universities and the Global Knowledge Economy: A Triple Helix of University-Industry-Government Relations. London: Cassell Academic.
Foucault, Michel (1972 ). The Archaeology of Knowledge. New York: Pantheon.
Foray, Dominique (1998). Errors and Mistakes in Technological Systems: From Potential Regret to Path Dependent Inefficiency. Pp. 217-39 in: Jacques Lesourne and André Orléan, Advances in Self-Organization and Evolutionary Economics. London, etc.: Economica.
Freeman, Chris and Carlota Perez (1988). Structural crises of adjustment, business cycles and investment behaviour. Pp. 38-66 in: Dosi (1988).
Frenken, Koen and Loet Leydesdorff (1999). Scaling Trajectories in Civil Aircraft, Research Policy (forthcoming).
Galbraith, John K. (1967). The New Industrial State. Harmondsworth: Penguin.
Galbraith, J. R. and D. N. Nathanson (1978). Strategy Implementation: The Role of Structure and Process. St. Paul, MN: West Publishing.
Giddens, Anthony (1984). The Constitution of Society. Cambridge: Polity Press.
Gouldner, Alvin W. (1970). The Coming Crisis of Western Sociology. New York: Basic Books.
Habermas, Jürgen (1968). Technik und Wissenschaft als "Ideologie". Pp. 48-103 in: Technik und Wissenschaft als "Ideologie". Frankfurt a.M.: Suhrkamp.
Hughes, Thomas P. (1987). The evolution of large technical systems. Pp. 51-82 in: Bijker et al. (1987).
Kampmann, Christian, Christian Haxholdt, Erik Mosekilde, and John D. Sterman (1994). "Entrainment in a Disaggregated Long-Wave Model." Pp. 109-24 in: Leydesdorff and Van den Besselaar (1994).
Kaufer, David S., and Kathleen M. Carley (1993). Communication at a Distance: The Influence of Print on Sociocultural Organization and Change. Hillsdale, NJ: Erlbaum.
Kauffman, Stuart A. (1993). The Origins of Order: Self-Organization and Selection in Evolution. New York/Oxford: Oxford University Press.
Kauffman, Stuart A. (1995). At Home in the Universe. New York/Oxford: Oxford University Press.
Langton, Christopher G. (Ed.) (1989). Artificial Life. Redwood City, CA: Addison Wesley.
Leydesdorff, Loet (1997a). The Non-linear Dynamics of Sociological Reflections, International Sociology 12, 25-45.
Leydesdorff, Loet (1997b). The New Communication Regime of University-Industry-Government Relations. Pp. 106-117 in: Etzkowitz and Leydesdorff (1997).
Leydesdorff, Loet (1999). Luhamnn, Habermas, and the Theory of Communications, Systems Research & Behavioral Science (forthcoming).
Leydesdorff, Loet (in preparation). The Triple Helix as an Evolutionary Model of Innovations.
Leydesdorff, Loet, and Henry Etzkowitz (1998a). Triple Helix of innovation: introduction, Science and Public Policy 25, 358-64.
Leydesdorff, Loet, and Henry Etzkowitz (1998b). The Triple Helix as a model for innovation studies, Science and Public Policy 25, 195-203.
Leydesdorff, Loet, and Peter Van den Besselaar (Eds.) (1994). Evolutionary Economies and Chaos Theory: New directions in technology studies. London: Pinter.
Leydesdorff, Loet, and Peter Van den Besselaar (1998). "Competing Technologies: Lock-ins and Lock-outs." Pp. 309-23 in: Daniel M. Dubois (Ed.), Computing Anticipatory Systems, Proceedings of the American Institute of Physics 437. Woodbury, NY: American Institute of Physics.
Luhmann, Niklas (1989). Gesellschaftsstruktur und Semantik. Vol. III. Frankfurt a.M.: Suhrkamp.
Lundvall, Bengt-Åke (1988). Innovation as an interactive process: from user-producer interaction to the national system of innovation. Pp. 349-69 in: Dosi et al. (1988).
Marx, Karl (, 1967). The Communist Manifesto, translated by Samuel Moore in 1888. Harmondsworth: Penguin.
Marx, Karl (1867). Das Kapital I. Hamburg, Meisner.
McKelvey, Maureen D. (1997). "Emerging Environments in Biotechnology." Pp. 60-70 in: Etzkowitz and Leydesdorff (1997).
Montesquieu, Charles de Sécondat (1748). De l'esprit des lois. Paris.
Nelson, Richard R. (Ed.), (1982). Government and Technical Progress: a cross-industry analysis. New York: Pergamon.
Nelson, Richard R. (Ed.), (1993). National Innovation Systems: A comparative study. New York/Oxford: Oxford University Press.
Nelson, Richard R., and Sidney G. Winter (1977). In search of useful theory of innovation, Research Policy 6, 36-76.
Nelson, Richard R., and Sidney G. Winter (1982). An Evolutionary Theory of Economic Change. Cambridge, MA: Belknap Press.
Noble, David (1977). America by Design. New York: Knopf.
OECD (1964). The Residual Factor and Economic Growth. Paris: OECD.
OECD (1988). Biotechnology and the Changing Role of Government. Paris: OECD.
Pasinetti, L. L. (1981). Structural Change and Economic Growth: A Theoretical Essay on the Dynamics of the Wealth of Nations. Cambridge, etc.: Cambridge University Press.
Rosenberg, Alexander (1994). Instrumental Biology, or, the Disunity of Science. Chicago: Chicago University Press.
Rosenberg, Nathan (1976). Perspectives on Technology. Cambridge, etc.: Cambridge University Press.
Sahal, Devendra (1981). Patterns of Technological Innovation. Reading, MA: Addison Wesley.
Sahal, Devendra (1985). Technological guideposts and innovation avenues. Research Policy 14, 61-82.
Salter, W. E. G. (1960). Productivity and Technical Change. Cambridge, etc.: Cambridge University Press.
Saviotti, Paolo P., and J. S. Metcalfe (1984). A theoretical approach to the construction of technological output indicators, Research Policy 13, 141-51.
Schumpeter, Joseph (1943). Socialism, Capitalism and Democracy. London: Allen & Unwin.
Schumpeter, Joseph  (1964). Business Cycles: A Theoretical, Historical and Statistical Analysis of Capitalist Process. New York: McGraw-Hill.
Schumpeter, Joseph (1966). Invention and Economic Growth. Cambridge, MA: Harvard University Press.
Simon, Herbert A. (1973). The Organization of Complex Systems. Pp. 1-27 in: Howard H. Pattee (Ed.), Hierarchy Theory: The Challenge of Complex Systems. New York: George Braziller Inc.
Solow, R. M. (1957). Technical Change and the Aggregate Production Function, Review of Economics and Statistics 39, 312-20.
Taylor, Frederick W. (1911). Principles of Scientific Management. In: Scientific Management. New York: Harper, 1947.
Van den Belt, Henk, and Arie Rip (1987). The Nelson-Winter-Dosi model and synthetic dye chemistry. Pp. 135-58 in: Bijker et al. 1987.
Vernon, R. (1972). The Economic and Political Consequences of Multinational Enterprises. An Anthology. Cambridge, MA: Harvard University Press.