Article "Technological paradigms and technological trajectories" by Giovanni Dosi, research policy 11 (1982) 147-162.
review by R.Molenaar (robmol@dds.nl)
The author of the article Govianni Dosi is the one who translated the notion of scientific paradigm from Kuhn to a technological analogy: a technological paradigm. He defines it as: "An outlook, a set of procedures, a definition of the relevant problems, and of the specific knowledge related to their solution". Very important is that within the paradigm itself the direction of advance is programmed. Dosi calls it a: ‘technological trajectory’.
The theory of a technological paradigm. does not explain a general theory of technical change rather it tries to explain why certain technological developments emerge instead of others.
The ‘old ‘ economic theories of technological growth are rather crude instruments to explain a technological trajectory according to Dosi. A brief summery follows:
we can distinguish two approaches:
1. demand-pull theories
2. Technology push theories
ad.1 This is a benthamite theory of utility. The producer distinguishes a clear need or demand within a consumer group. he will try to fulfill it if he sees profit-potential in it. There are numerous problems with this theory like: what is a real need for consumers for a producer? etc. Dosi summarizes the problems as follows:
1. The concept of passive and mechanical reactiveness of technological changes vis a vis market conditions is flawed.
2. There is an incapability of defining the why and when of certain technological developments instead of others.
3. The neglect of changes overtime in the inventive capability, which do not bear any direct relationship with changing market conditions.
The most important notion is that a demand-pull approach fails to produce sufficient evidence that "needs expressed through market signaling" are the prime movers innovative activity.
ad.2
The technology-push theories are no neat theories like the demand pull theories. Dosi suffices with summing up some important features:
1. The increasing role of scientific inputs in the innovation process.
2. The increased complexity and fixed nature of R&D expenditures which makes a quick response to a change in demand rather difficult.
3.A significant correlation between R&D efforts and innovative output. The lack of clear correlation’s between market and demand patterns on the one hand and innovative output on the other hand in cross-country comparisons.
4. The intrinsically uncertain nature of the outcome of R&D efforts in response to a clear set of consumer preferences.
The problem with the theory-push theories is that they fail to take into account the intuitive importance of economic factors in shaping the directions of economic change (in effect they leave the demand-pull side of the story completely out of the picture).
For this class especially it is important to note that Dosi would like to find a theory which tries to explain "the complex structure of feed-back mechanisms between the economic environment and the directions of economic change".
In section 3 Dosi starts of with defining more clearly the analogy between a scientific paradigm and a technological paradigm. A technological paradigm is a ‘perception’ of a limited set of possible technological alternatives and of a notional future development. Especially the last part is important in Dosi’s view. There is a selected technological problem based on selected scientific principles derived from natural sciences on selected material technologies. Within this paradigm the search path to incrementally solve this ‘puzzle’ is programmed. This means that there is a clear distinction between positive and negative heuristics to solve the puzzle. Dosi defines this as a technological trajectory. In a more baroque phrasing: both the scientific and the technological paradigm represent strongly selective gestalten embodying powerful heuristics.
Now we come to a crucial question in Dosi’s article: how did an established paradigm emerge in the first place, how was it preferred to other possible ones.
Here Dosi introduces a bridge between the demand-pull and technology push theories. ‘Economic forces’ together with institutional and social factors operate as a selective device. "Within a large set of technological possibilities (or paradigms, rm) of directions of development, notionally allowed by science, a first level of selection operates on the basis of rather general questions like: Is any practical application conceivable?" or in my words is there an economic rationale to choose this technological paradigm? "The economic criteria acting as selectors define more and more precisely the actual path followed inside a much bigger set of possible ones".
After this Dosi returns to a more technological-push approach; "once a path has been selected and established, it shows a momentum of its own. This enforces the direction towards which the "problem solving activity" moves. This is the notion of two other important authors in the field: Nelson and Winter. They call this a natural trajectory of technological progress.
Once ‘a lock-in’ in the paradigm has been effectuated the trajectory can be seen as a ‘cylinder’ in the multidimensional space defined by these technological and economic variables. The outer boundaries of the possibilities of the paradigm are then fixed. Dosi then continues with a few observations of which I shall only name the most important (given the time restraint):
- "progress’ on a technological trajectory is cumulative.
- If the technological paradigm changes one has to start from scratch in the beginning of problem solving activity (i.e. to select a new paradigm).
3 Given information problems it is probably impossible to select and asses the superiority of one technological paradigm over another. Objective criteria to select can only be formulated ex post.
Dosi continues with the reviewing of the economic, institutional and social factors in greater detail:
in fact he doesn’t consider these factors in great detail. He stipulates the importance of factors such as:
1. the economic interests of organizations involved in R&D in these new technologies.
2. their technological history.
3. institutional variables like public agents, the military etc.
He stresses in particular the role played in the establishment of a particular technological trajectory by public or political forces.
In fact after this Dosi distinguishes two selection layers: first the institutional one and second the market one. The first one is ex ante the second one is ex post. "The Economic and social environment affects technological development in two ways: first selection of the direction of mutation (i.e. the selection of the technological paradigm). Second: selecting among mutations in a more Darwinian matter (i.e. the ex post selection among Schumpeterian trails and errors). So after the establishment of more technological paradigms there is a final test: does the market want to buy it? Or in a more sociological context: can the market be convinced to buy it. Dosi formulates this in a difficult phrase ( at least for me): "The supply-side determines, so to speak, the ‘universe’ of possible modalities through which generic ‘needs’ or productive requirements (which as such do not have any direct economic significance) are satisfied.
The above described factors explain the rigidities of R&D in response to changing economic conditions and demands. Dosi until now did not give an explanation for a sudden technological break-through. He ends section one by noting that: "extraordinary technological attempts emerge either in relation to new opportunities opened up by scientific developments or to the increasing difficulty in going forward on a given technological direction. (but to me it seems that we did not get a lot further seen from the starting point from this article with this rather vague formulation).
In section 4 Dosi goes on to elaborate on the consequences of technological trajectory selection for economic theory, especially economic sector structure. He distinguishes two phases.
1. The ‘emergent industry’ phase in which a lock-in in a technological trajectory has not yet materialized. There are two selectors in this phase. Institutions channel the paradigms roughly. In this selector mode the ‘Schumpeterian’ entrepreneurs establish themselves in the market which results in a selected matrix of paradigms. By trail and error they fulfill a ‘trailblazer’ function (they try to set a standard). The winning strategy here is to erect dynamic entry barriers, i.e. to go up the learning curve as fast as possible. In the second selector mode of market-competition an arena is created by introducing a market-mechanism. In this mode 2 one paradigm will be selected. It will be decided which entrepreneurs will be champion. Their way of producing and working with the paradigm will be the standard for the rest of the industry. In this phase competition is beneficial for innovation, because it will select the most vital companies and production paradigms.
2. The second phase is the oligopolistic phase. There are only a few big competitors left in the market. Here innovation is better served by an oligopolistic structure, because innovative activity becomes more dependent on the surplus the oligopolistic firm can realize in the maturing market, by creating static entry-barriers.
Section 5: Conclusions and policy implications.
Dosi starts of with saying that the analogy between a scientific paradigm
and a technological one should not be taken as an identity. Still, the
analogy can throw some light on the problem of continuity versus
discontinuity in the studying of technological change. First: Radical
innovation happens more in the first Schumpeterian phase of innovative
activity, incremental innovative activity happens in the second maturing
or oligopolistic phase. Second: The paradigm selection and natural trajectory
notions can also explain something about the procedure through which technological
change occurs. The search for new products and new technological opportunities
are never a random process, but the result of the cognitive window which
the paradigm generates and the searching routines which become established
if a lock-in in a new trajectory materializes. 3. The observed cumulativeness
of technical advance is a way of reducing the uncertainty which is inherent
to innovative activity. Ex ante (beforehand) and even ex post
nobody knows what the pay-off of a new paradigm will be, so one sticks
to the routines and paradigms one already practices.
After this Dosi elaborates on the notions mentioned in section 4. To me the picture doesn’t contribute to the explanation power of the notion in section 4 because he remixes the relative clear notions he described before. I will try to summarize a couple of things:
* market-induced innovative activity, a notion developed by Smookler, may be consistent with the selector 2 mode in phase one of innovative activity. Dosi calls this confusingly ‘normal’ innovative activity. It is opposed to discontinuous technological advance which should be corresponding, I gather, to selector mode 1 in phase one.
* In the long run development of technology (so phase one, selection mode 1) there is a relative autonomy vis-a-vis short run adjustment and inducement mechanisms of the economic systems (changes in prices, quantities, profitabilities etc.). This is a period where: "A reconstruction of the history of technology and history of science would almost certainly show the long run influences of the evolution of the social and economic structure upon the emergence of new scientific and technological opportunities.". So no market-induced innovative activity here, i.e. a clear distinction with neo-classical theory.
* Dosi tries to reconcile Nelson and Winter’s ‘natural trajectory’ theory with his own by placing it in phase two of innovation, because their models focus primarily on the endogenous dynamics of technical progress in an oligopolistic environment. I gather that in this phase the incentive for innovation is not so much market induced, but differential innovative success, so ‘technology-push’. In the real world there is a lot of price competition in an oligopolistic environment though, especially if we enter the last part of the maturing phase (for example cd-players now).
* Dosi tends to focus with his theory more on selector mode one- phase two in which institutions and the broad socio-economic environment act as a selector and focusing device in a chaotic world of potential technological paradigms.
Last remarks:
- maybe there is a possibility here to investigate the problem of the
differential development of countries in innovation. There will be no convergence
in growth if oligopolistic structures in countries with technological lead
prevail.
discussion points:
* There continues to be some difficulty in reconciling the endogenous nature of technological growth and price induced innovation, i.e. the gap between demand-pull and technology-push theories is not sufficiently bridged in my opinion.
* Dosi’s theory tells us something about the selection of technological paradigms, but doesn’t explain a lot about why at a certain time a new paradigm is born, i.e. when a click between different subsystems is effectuated. What is the relation between the Kondratieff theory of long conjectural waves. Does a low conjectural part of the wave induce a new innovative click?, i.e. is this the starting point for the emerging of a new technological paradigm?