Emergence of a Triple Helix of

University‑Industry‑Government Relations

Science and Public Policy (forthcoming)


Loet Leydesdorff * & Henry Etzkowitz #

* Science and Technology Dynamics, Nwe Achtergracht 166, 1018 WV Amsterdam, The Netherlands; <>;

# Science Policy Institute, State University of New York, Purchase, NY 10577, USA; <>



In the 1990s, newly industrializing, de‑industrializing and re‑industrializing nations, somewhat to their surprise, find that they share a mutual interest in fostering knowledge-based economic and social developments requiring the creation of boundary-spanning mechanisms.  Despite their quite different developmental histories, a broad spectrum of societies, formerly conceptualized under the divergent rubrics of the first, second, and third worlds, have formulated innovation strategies based upon the deliberate elaboration of academic‑industry relations through reflexive S&T policies.


How should one model the knowledge infrastructure of such a global system?  Evolutionary economics has focused hitherto on the co-evolution of firms and technologies given a knowledge infrastructure (e.g., Nelson 1993 and 1994); and sociological studies have highlighted the institutional dynamics in academic-industry relations (e.g., Gibbons et al. 1994; Etzkowitz 1994a).  In our opinion, three dynamics can be distinguished: the economic dynamics of the market, the internal dynamics of knowledge production, and governance of the interface at different levels.


We propose to model this complex system as a Triple Helix of University-Industry-Government Relations (Etzkowitz & Leydesdorff 1995).  At a workshop in Amsterdam (3‑6 January 1996) the Triple Helix was discussed in various national and regional contexts.  Ninety participants took part, representing Latin America, Eastern Europe, Western Europe, North America, Australia, and Southeast Asia.  The discussion was focused on the future of university research in the emerging regime of knowledge production and dissemination.  In the first part of this report, we elaborate the Triple Helix model.  In the second, we summarize the contributions to the workshop, and in a final part some theoretical and policy implications will be noted.




The Triple Helix model takes the traditional forms of institutional differentiation among universities, industries, and government as its starting point.  The evolutionary perspective adds to this historical configuration the notion that human carriers reflexively reshape these institutions.  The model thus takes account of the expanding role of the knowledge sector in relation to the political and economic infrastructure of the larger society.


Historically, the systematic interaction between markets and sciences can be traced back to the second half of the 19th century.  Braverman (1974), for example, has characterized the "scientific‑technical revolution" (which he located between 1870 and 1890) as "the transformation of science itself into capital" (cf. Noble 1977).  Indeed, Marx himself took note of this nascent dynamic, viewing it as a future source of industrial growth, transcending simple combinations of land, labor and capital (Marx 1953).


Institutional differentiation between the nation state and the economy preceded this transformation.  Thus, the analysis should account for two differentiations: the functional differentiation between sciences and markets, and the institutional difference between private and public control.  Over time the cross-tabulation of these categories leads to a model of technological developments in terms of university-industry-government relations.


A dynamic analysis of the infrastructure of knowledge has become urgent in the light of proposals for reindustrialization, typically involving the development of closer connections between the state and industry, and academia and industry (e.g. OECD 1980).  Since the late 1970s, national coordination beteen industrial policies and science & technology policies has taken hold in Western countries that had earlier achieved the highest degree of separation between institutional spheres.  One implication of the Triple Helix model is the analysis of the binding forces among autonomous, yet tightly connected, institutional arenas.


The increase of interactions among the institutions has had the effect of generating new structures within each of them, such as centers in universities or strategic alliances among companies.  These interactions have also led to the creation of integrating mechanisms among the spheres in the form of networks, e.g., of academic, industrial and governmental researchers, and hybrid organisations such as incubator facilities.


Recently, universities in various countries have suffered budget cuts of sometimes more than twenty percent.  However, this does not necessarily indicate decline.  The system is in transition: growth has been concentrated in sciences at the interfaces like biotechnology, artificial intelligence, advanced materials, etc. (Gibbons et al. 1994).  Cooperative research centers are in many countries the fastest growing institutions of university research (Etzkowitz & Kemelgor, forthcoming).  Thus, we are in need of a model that focuses on the transition mechanisms in the complex set of interactions.




The noted "transformation of science itself into capital" has not been smoothly and equally distributed over the range of possibilities.  Leading technologies like chemistry and electro‑technical engineering induced long waves in the economy which have provided corporations with opportunities to develop sectoral patterns of innovation (Noble 1977).  Since Keynes, economic downturns in the innovative dynamics have been controlled by macro‑economic government interventions (cf. Freeman & Perez 1988). 


The sustained interaction processes between vertically integrated institutions like industry and academia have generated internal differentiation processes within each of these institutions.  For example, R&D has provided the economy with innovations that upset the movement towards equilibrium in the market (Schumpeter 1939).  Accordingly, the modern corporation has had to develop an interface between business and R&D.  Galbraith (1967) coined the term "technostructure" for the management of the optimization between price maximalization and growth maximalization.  Price competition has to be offset against (knowledge intensive) product competition.


The recursive interaction between knowledge-based innovations ("search processes") and market forces ("selection processes") tends to lock firms into technological trajectories (Nelson & Winter 1982).  A prime example has been the succession of airplanes from the DC-3 to the DC-10.  Adjustment of the knowledge infrastructure itself is required when one has to choose between trajectories as in the case of supersonic airplanes versus wide-body airplanes in the late sixties.  Usually, the state then becomes involved.


While evolutionary economics has focused on the co-evolution of technological trajectories and selection environments, the Triple Helix model endogenizes the knowledge infrastructure of society as a next-order regime (cf. Dosi 1982).  "National systems of innovation" can be considered as competing in a global economy in terms of their knowledge infrastructure (Porter 1990).  Various studies of these systems have emphasized the importance of values shared between users and producers (Lundvall 1992; Nelson 1993).  However, similar mechanisms of mutual trust-building relations have been described within sectors of the economy (e.g., medicine; cf. Von Hippel 1988) and at the level of regions (e.g., Blume & Leydesdorff 1984).  Thus, the notion of governance can be generalized to the concept of a nested structure of reflexive controls.  National governments, regional development agencies, academic hospitals, cooperative research centres, etc., develop policies with different objectives and at different levels.


While a co-evolution or a double helix can be stabilized relatively easily, a complex and potentially unstable system emerges when three dynamics interact.  In certain circumstances, such as the exigencies of war-time, governments may be able to harmonize and lead the interactions.  In a liberal economy, however, integration occurs ex post, i.e., on the basis of asynchronous selections by the (sub‑)systems upon each other.  Irreversibilities, lock-ins in inferior technologies (like VHS for the VCR), crises, and phase transitions are well-known phenomena in complex systems (e.g., David & Foray 1994).  Nexuses at the interfaces between the helices may unexpectedly begin to absorb energy by allowing flows in specific combinations.  The Triple Helix model is sufficiently complex to accomodate various forms of chaotic behaviour in the resulting system.


New developments at the network level are generated and stabilized in co-evolutions (or processes of "mutual shaping") between two helices.  The perspective of the third helix provides each of the partners with a reflexive angle that can lead to other selections than those which seem to have occurred "naturally".  Thus, the helices communicate by selecting upon the variations in and the interactions among the other ones (cf. Leydesdorff 1994).  This recursive selection can be considered as a process of creative destruction (cf. Schumpeter 1939).  The complex system recovers from its tendency to disintegrate by organizing new combinations. 


Both market mechanisms and interactions among reflexive actors in negotiations shape the emerging regime by generating a variety of niches with their own trajectories.  For example, a new regional innovation environment, such as an industrial district cooperating with universities or other knowledge-producing organizations with local government, can be generated with an impact at the network level.  The regime at the network level, however, is no longer localizable; it is distributed, and it acts by reproducing distributions.  Whereever and whenever the various dynamics begin to resonate, flows of energy, resources, and human capital can be absorbed.  While the regime is expected to develop "on the edge of chaos," niches may crystallize at unexpected places for prolonged periods.  Within the niches a reproduction of the differentiation is likely, but in potentially a different order.


Sandra Negraes Brisolla (University of Campinas, Brazil), for example, addressed the question of why by 1985 Brazil had fallen behind Korea in "technological capacity."  In the early 1980s, at the time of the transition from military dictatorship to democracy, the roles were still reversed.  The analysis concluded that Korea's successful management of the financial sector seems to have been a crucial factor.  While Brazil had maintained an open market during the "debt crisis", the Korean government managed to shelter its banking system protectively, and thereby created financial conditions for the development of a knowledge-based economy.


Brisolla proposed to distinguish (for example, national) systems in terms of their "technological capacity" to transform the economy.  Such a capacity includes relevant factors like banking, the educational system, and the flexibilities for transformations within the relevant markets.  From an evolutionary perspective, the specificities of such a system should be analyzed in terms of niche creation and niche management within a global economy.  Jian Tong (Chinese Academy of Sciences) contributed by specifying the categories for varation, selection, and retention in this analysis:


    "In Nelson & Winter's theory, the game is played within markets and industrial firms are the only players, whereas to address the triple helix analytically, academic institutions, industrial firms and government agents apparently all should be players in a game played beyond as well as within markets."


Tong proposed to consider "human capital" as the main factor generating variation in knowledge‑based economies, while "niche selection" provides feedback and adaptation processes make institutional reproduction possible.  This design highlights the crucial position of cognitive reflexivity in the processes that transform institutions and markets.




The conference focused on the implications of knowledge-based economic development for the university.  Universities seem to be going through a second academic revolution: the economic function of the university is increasingly institutionalized in addition to the differentiation between higher education and research (cf. Etzkowitz 1994a).  As Judith Sutz (University of Montevideo, Uruguay) expressed it forcefully in her contribution to the conference:


    "The increasing demand for funds from universities and research institutes gets a similar response worldwide: support yourselves!  That is to say, connect yourselves with industries and the government, offer your knowledge and your capacity to generate new knowledge and charge for it.  Only in this way will you be able to extend your laboratories, hire young people, and increase your salaries."


As noted, the budget cuts do not necessarily indicate decline, since the system is in transition.  Tim Turpin and his colleagues from Wollongong (Australia) concluded from a comparison of developments in China and Australia that ten to fifteen years of transformation have thoroughly "disorganized" the previously existing institutional boundaries.


The study of the Triple Helix requires a model that complements the institutional perspective with a focus on interactive operations at the system level.  The institutional frameworks can be considered as the fingerprints of communication structures that have been functional hitherto.  All the actors and agencies involved are reflexive, i.e., recurrently adjusting their positions given institutional constraints and opportunities.  Communicative competences (Habermas 1981) become as important as achievements within each of the helices.  Quality can then be defined in terms of specific combinations.  A further specification of qualification in terms of communicative competence follows naturally from these insights.  (Tobias et al. (1995) report that those who are able to translate among specialist "languages" are sometimes considered as "gold collar" workers in the American economy.)  In cross-institutional interactions, the disciplinary frameworks function as heuristics that guide the researchers in improving the substantive quality of their communications.


Academic, industrial, and governmental institutions each contain communication structures and culturally encoded messages that are sometimes difficult for outsiders to interpret.  Science journalists, venture capitalists, technology transfer officers, and others who have often passed through several institutional spheres in the course of their careers, have become adept at the translation process.  Translations between institutional codes, such as from scientific findings into stock market offerings, are highly specific.  Some academic scientists have learned to speak the language of business, while at the same time learning not to divulge too much of the particulars of their science to their industrial interlocutors.  Similarly, intellectual property attorneys, even those without a science degree, have acquired a credible facility with scientific jargon in their particular fields.  Although many of these people remain in their respective spheres, they have mastered the art of cross‑institutional conversations.


The Triple Helix operates in terms of translations among specific (i.e., "high quality") communications and therefore highly selective transformations of institutions.  The helices drive each other in terms of specificity and selectivity.  Given the development of academic-industry relations, new transformation and differentiation processes are increasingly necessary also in the helix of government: a non‑zero sum has to be organized by governing agencies internally to legitimate the relations among neo‑corporatist arrangements, public funding, and democratic control.  The debate has focused hitherto on "intellectual property rights", but in practice one seeks arrangements that sustain the transitions by resolving legal tensions.  Within academia, particularly, the variety among disciplines can be considered as a rich source of recombinations.  However, this presupposes a functional instead of a bureaucratic orientation on the part of the administration, and scholars who are able to appreciate each other's perspectives despite disciplinary "incommensur­abilities".




While the new regime is globally constructed, the nested subsystems develop over a distributed range of local trajectories.  Several papers (e.g., by Philippe Laredo and Philippe Mustar from the École des Mines, Paris) focused on the deliberate attempts of the European Commission to develop transnational networks of scientists and industrialists in order to shape the European Union.  What effects has this effort had in terms of research and the formation of SMEs?  Europe's supranational networks came on top of national programs that tended to focus more on strategic priorities (like biotechnology) than on university‑industry relations.  In the US, on the other hand, university‑industry relations have been a priority for lower‑level (i.e. state) governments with a focus on regional development (Berglund and Coburn 1995).  Thus, different dynamics can be discerned on each side of the Atlantic Ocean (OECD 1988).


Variations abound: comparisons between European countries reveal different patterns of university‑industry relations, different sectors, and different technologies (e.g., Faulkner & Senker 1994; Etzkowitz & Leydesdorff 1995).  But the emerging system has characteristics that transcend local specificity.  New modes of the production of scientific knowledge, as biotechnology, have generated co‑evolutions along axes of previously unknown interactions (Maureen McKelvey at Linköping, Sweden).  Petra Ahrweiler (University of Bielefeld, Germany), in her contribution to the conference, spoke of how the artificial intelligence community in the Federal Republic has consciously been developed as a "scientific‑politico-economic" network for discussions and decisions about the various options available for developing the field.  Terry Shinn (GEMAS, Paris) showed that communities that discuss "research technologies" and scientific instruments across disciplinary and institutional boundaries have existed since the late 19th century.  Problem-solving requires the use of theories as heuristics, and an orientation towards interaction with other domains.  Such transformations affect the nature of scientific theorizing by developing other dimensions of relevance (cf. Etzkowitz & Peters 1991).


These transformations also challenge the political system to develop new codifications to enable and legitimate the use of public resources for further development of the Triple Helix.  The structuring of the exchange, rather than its direction, has become the main function of government agencies: processes of exchange have to be organized with a focus on mutual learning.  The further specification and codification of mutual interests and expectations is needed.  However, the system is vulnerable to corruption, since it must be institutionally managed, and one has to be highly selective.  Quality control is shifted from institutional codification to communicative precision and reliability.


Traditionally, the patent system has been considered as an example of functional legislation at a nexus (Van den Belt & Rip 1987).  The patent system itself has evolved to define mathematical algorithms and new forms of life as inventions in addition to more conventionally innovative artifacts.  Andrew Webster (University of East Anglia, UK), however, reported on a survey among British academics about the shortcomings of the patenting system to cover the new arrangements.  Bill Kaghan (University of Washington, US) noted that the patent system is based on "a laboratory model of science", while many sciences are increasingly generating innovations using a "desktop" model.  Some elements can be patented, while others need copyright protection.  The institutional models used by policy‑makers and transfer agents are in need of an update.


Pierre Benoît‑Joly (Grenoble, France) argued on the basis of case materials from the Rhône‑Alpes region that the processes of knowledge transfer require phases in which standardization is achieved so that the discussion can be closed, and phases in which the restricted discourse of a university department and its spinn‑off company has to be elaborated at the interface.  In different phases of the incubator process, other kind of support structures are needed (Anne‑Marie Maculan, Rio de Janeiro, Brazil).  Cooper Langford (University of Calgary, Canada) showed with extensive case materials from the Canadian system that complex networks experience phase transitions.  At certain stages each next link may trigger a crystallization process of a next‑order dynamics (cf. Kauffman 1993). 


Niche formation requires a fine mix of protection from and exposure to market forces.  New systems have to be constructed carefully, and adjustment mechanisms cannot be generalized as recipes.  The system must be analyzed in terms of latent functions that have to be fulfilled, and not only in terms of institutional requirements.  While it is shortsighted to require from universities that they become profit centers overnight, academic institutions have themselves taken the initiative to restructure, often against opposition from disciplinary interests.  But what should one do if these quasi‑entrepreneurs lose money?  Indeed, what constitutes a loss, or a profit for that matter, and how should it be accounted?  One frequently quoted (although methodologically debatable) study has credited the Massachussetts Institute of Technology with being the source of 1/10th of the wealth of the Massachussetts economy.  Frances Anderson of the NRC (Canada) discussed the tensions between funding and public accountability in the case of the need for long-term investments in the knowledge infrastructure.


Public institutions have not traditionally been shaped to produce measurable results, although they are increasingly being steered in that direction, whether bluntly, as in the U.K. or more indirectly, as in the U.S.  Although the dimensions--public/ private/ knowledge--are initially orthogonal (J.‑C. Spender, Department of Commerce, US), some universities have been quick to redefine their research and teaching strengths as "profit centers," or in more ethereal teminology as "steeples of excellence."  The direction of this development can be assessed.  For example, it is becoming increasingly common for a university to retain a consulting firm to carry out market surveys to position new programs within the tuition and quality levels of its academic competitors.




Is there so much differentiation that nothing can be said in general?  We don't think so: the researcher should not study only the observable variations, but also try to specify selecting mechanisms.  The distributed events can be considered as the outcomes of interactions between underlying dynamics (cf. Leydesdorff 1995).  The understanding of these different dynamics is important for one's orientation, and consequently for the specification of a research agenda.


Exchange is an asymmetrical process in principle.  The analysis of the asymmetries provides the analyst with access to the underlying dynamics in each concrete case.  What is expected by whom, and why?  Which systems are evolving?  What is being transformed, and what is communicated in the processes of interaction?  Thus, the focus on the "symmetrical" socio-cognitive interaction which has hitherto been dominant in the sociology of science should be extended to include an analysis of the potentially asymmetrical meaning of such interaction for the internal development of the discipline and its various social relations.  All interacting systems have to reintegrate their operations, but along different axes and potentially different time horizons.


As noted, the hidden assumption of evolutionary economics has been the theory of the firm.  This focus is sometimes extended to the group of companies in an industry, or a heterogeneous network of companies in an industrial district.  These delineations, however, have remained largely institutional, while we are pleading for a focus on communicative interactions and mechanisms as well.  This extension enables us to extend the analytical framework so that the study of the knowledge infrastructure of society can be endogenized.


For example, Magnus Gulbrandsen (Oslo, Norway) showed how new developments at the interfaces among Scandinavian universities can be assessed in terms of successful instances of knowledge-based economic developments as in the regions of Grenoble and Cambridge.  Additionally, the strong Latin American participation in the conference provided opportunities for comparisons of these results with Mexican (Rosalba Casas and Mathilde Luna) and Brazilian (Ary Plonski) experiences.




The Triple Helix model leads us to view the institutional actors on an equal level in the network.  However, each is positioned differently with reference to the infrastructure that they collectively reproduce.  Therefore, the focus on observable interests and agency should be complemented with attention to expectations and orientations in communication systems.


Events that have occurred can be considered as realizations of sub‑optima; they are merely the beginnings of future formats that can be attained through processes of mutual adjustment.  The actual histories are embedded in a phase space, typically a very early stage of possible trajectories of development.  This range can be simulated algorithmically on the basis of an analytical reconstruction of the observable phenomena (cf. Allen 1994; Bruckner et al. 1994).  The simulation results make us sensitive to possibilities that may appear counter‑intuitive from the present state.


For example, the simulations may help us to understand potential bottlenecks (e.g., "lock ins" such as the premature adoption of a particular technological standard).  There are bottlenecks in the system that perhaps must be cleared before further advance can occur.  Unintended consequences of various options, both positive and negative, can be identified.  When one understands the underlying dynamics, it is sometimes possible to foresee future developments, much as Marx did when he envisioned a science-based dye industry emerging from Perkins' 1856 discovery (Beer 1959).


In our opinion, the knowledge-based economic regime has made the distinction between laissez faire and active-state intervention obsolete: governance nowadays means codifying high-quality selections that set free new areas of activity as zones of recombination (cf. Etzkowitz 1994b).  One expects the new opportunities (and the new jobs!) to emerge not in the existing institutions, but in careful recombinations that are based on knowledgeable reconstructions.  An economic and science policy analysis that fails to consider these potentials for recombination of elements among the helices will miss the lessons of several decades of experience in knowledge-based economic developments.


*An edited selection of papers from the conference will be published by Cassell in a volume provisionally entitled Universities in the Global Knowledge Economy, edited by Henry Etzkowitz and Loet Leydesdorff.






Allen, Peter M. (1994).  "Evolutionary Complex Systems: Models of Technology Change," in: Leydesdorff & Van den Besselaar (1994): 1-17.

Beer, Joseph (1959). The Emergence of the German Dye Industry (Urbana: University of Illinois).

Berglund, Dan & Christopher Coburn (1995). Partnerships: A compendium of state and federal cooperative technology programs (Columbus, Ohio: Batelle).

Blume, Stuart S. & Loet Leydesdorff (1984).  "The Role of the University in Today's Economy," Special Issue of the Intern. J. of Institutional Management in Higher Education 8, 97‑ 181.

Braverman, Harry (1974). Labor and Monopoly Capital. New York/London: Monthly Review Press.

Bruckner, Eberhard, Werner Ebeling, Miguel A. Jiménez Montaño, and Andrea Scharnhorst, "Hyperselection and Innovation Described by a Stochastic Model of Technological Evolution," in: Leydesdorff & Van den Besselaar (1994): 79-90.

David, Paul A. & Dominique Foray (1994). "Dynamics of Competitive Technology Diffusion Through Local Network Structures: The Case of EDI Document Standards," in: Leydesdorff & Van den Besselaar (1994): 63-78.

Dosi, Giovanni (1982).  "Technological Paradigms and Technological Trajectories," Research Policy 11, 147-62.

Etzkowitz, Henry (1994a). "Academic-Industry Relations: A Sociological Paradigm for Economic Development," in: Leydesdorff & Van den Besselaar (1994): pp. 139-51.

Etzkowitz, Henry (1994b). "Technology Centers and Industrial Policy: the Emergence of the Interventionist State in the USA," Science and Public Policy 21, 79-87.

Etzkowitz, Henry & Lois Peters (1991). "Profit from Knowledge: Organizational Innovations and Normative Change in American Universities," Minerva 29, no. 2, pp. 133-66.

Etzkowitz, Henry & Loet Leydesdorff (1995). "The Triple Helix of University-Industry-Government Relations: A Laboratory for Knowledge Based Economic Development," EASST Review 14, no. 1, pp. 11-19.

Etzkowitz, Henry & Carol Kemelgor (In Press). "Research Centres: The Collectivisation of Academic Science," Minerva, forthcoming.

Faulkner, Wendy, & Jacqueline Senker (1994). Knowledge frontiers: public sector research and industrial innovation in biotechnology, engineering ceramics, and parallel computing (Oxford/ New York: Clarendon Press).

Galbraith, John K. (1967). The New Industrial State (Harmondsworth: Penguin).

Gibbons, Michael, Camille Limoges, Helga Nowotny, Simon Schwartzman, Peter Scott, & Martin Trow (1994). The New Production of Knowledge: The dynamics of science and research in contemporary societies (London, etc.: Sage).

Habermas, Jürgen (1981).  Theorie des kommunikativen Handelns (Frankfurt a.M.: Suhrkamp).

Kauffman, Stuart A. (1993).  The Origins of Order: Self-Organization and Selection in Evolution.  New York: Oxford University Press.

Leydesdorff, Loet (1994).  "The Evolution of Communication Systems," Systems Research and Information Science 6 (1994) 219-30.

Leydesdorff, Loet (1995).  The Challenge of Scientometrics: The development, measurement, and self-organization of scientific communications (Leiden: DSWO Press, Leiden University).

Leydesdorff, Loet & Peter Van den Besselaar (Eds.), (1994). Evolutionary Economics and Chaos Theory: New Directions in Technology Studies (London: Pinter).

Lundvall, Bengt-Åke (Ed.), (1992). National Systems of Innovation (London: Pinter).

Marx, Karl (1953 [1857]).  Grundrisse der Kritik der politischen Oekonomie (Berlin: Dietz).

Nelson, Richard R. (Ed.), (1993). National Innovation Systems: A comparative study (New York: Oxford University Press).

Nelson, Richard R. (1994). "Economic Growth via the Coevolution of Technology and Institutions." In: Leydesdorff and Van den Besselaar (1994): pp. 21-32.

Nelson, Richard R. and Sidney G. Winter (1982). An Evolutionary Theory of Economic Change (Cambridge, Mass.: Belknap Press of Harvard University Press).

Noble, David (1977). America by Design (New York: Knopf).

OECD (1980).  Technical Change and Economic Policy (Paris: OECD).

OECD (1988).  Biotechnology and the Changing Role of Government (Paris: OECD).

Porter, Michael E. 1990. The Competitive Advantage of Nations (London, etc.: Macmillan).

Schumpeter, Joseph A. (1939). Business Cycles: A Theoretical, Historical and Statistical Analysis of Capitalist Process (New York: McGraw-Hill).

Tobias, Sheila, Daryl E. Chubin, & Kevin Aylesworth (1995). Rethinking Science as a Career: Perceptions and Realities in the Physical Sciences (Tucson, AZ: Research Corporation).

Van den Belt, Henk & Arie Rip (1987). `The Nelson-Winter-Dosi model and synthetic dye industry,' in: Wiebe Bijker, Thomas P. Hughes and Trevor Pinch (Eds.), The Social Construction of Technological Systems (Cambridge, Mass.: MIT), pp. 135-58.

Von Hippel, Eric (1988). The Sources of Innovation (New York/Oxford: Oxford University Press).