12.4 ENTREPRENEURSHIP AND AGENCY

Innovation seems to offer everything that a historian could want from a theory of historical development. It highlights the unpredictability and uncertainty of change. It rejects progress in favour of a discontinuous understanding of change.

At first blush, the answer is simple: the entrepreneur. One analogue of the entrepreneur for intellectual historians, as we noted above, is Skinner's innov­ating ideologist. But here we have to pause. Are the actions of the innovating ideologist actually responsible for bringing about conceptual change? The evidence is equivocal. In the essay that first introduced the concept of the innovating ideologist, Skinner was careful to disavow any causal claim about the role of Protestant vocabulary in the rise of capitalism. After commenting that it had ‘become a commonplace amongst historians to repudiate any suggestion that the principles of Protestant Christianity played a causal role in the development of capitalist practices', Skinner went on to defend the conclusion that, even if large-scale industrial capitalism pre-dated the Reformation, ‘the Protestant work ethic was particularly well adjusted to legitimating the rise of capitalism, and in this way helped it to develop and flourish'.^[857] Providing a conducive ideological environment for a social prac­tice is of course very different from actually creating such an institution. Skinner would, I assume, agree: his insistence on the relative looseness of fit between changing social practices and normative vocabulary is one reason why he has been so eager to repudiate any ‘general theory' of the relations between conceptual and social change.

Here, then, is the puzzle. Even in Skinner's account of innovation, which of all of the approaches we have considered takes very seriously the import­ance of the role of leadership in bringing about conceptual change, the causal responsibility of agents in bringing about cultural change is unclear. When describing the actions of the innovating ideologist, Skinner works with a deliberately thin notion of agency, which makes innovating ideologists stake­holders, so to speak, in the process of cultural change, but which also makes their actions, in many cases, ‘precipitates of their contexts'.^[859] Both Hacking's and Kuhn's models of conceptual innovation, meanwhile, leave even less space for instrumental action in the process of conceptual change. Kuhn scandalised the creativity theorists by arguing that creativity was not a property of personality, but of ‘community structure'. The root of this difficulty with the agency problem - with who or what brings about innovation - can, I think, be traced to an ambiguity in the concept of entrepreneurship itself. In a word, the theory of enterprise looks like the description of a type of agent, but in practice there is no such agent or discrete social type.

I can begin to explain what I mean here by repeating Schumpeter's definition of the concept of entrepreneurship in The Theory of Economic Development.

Despite appearances, however, entrepreneurship does not describe a stable class of persons or a particular kind of activity. Rather, it is a function defined by a set of abilities. In Schumpeter's schema, this function was of course the carrying out of new combinations. It followed that one could call an entrepre­neur ‘all who actually fulfil [this] function'. This was a broad definition insofar as it could include not just ‘“independent” businessmen', but even ‘“depend­ent” employees of a company'. The requisite abilities of such an entrepreneur included ‘“initiative”, “authority” and “foresight”'. This sensitivity to oppor­tunities, along with the power to create new combinations, described a form of nous and boldness that could hardly constitute an office. Innovation ‘can no more be a vocation than the making and execution of strategic decisions, although it is this function and not his routine work that characterises the military leader'. Accordingly, the existence of entrepreneurship was often obscured by the fact that it could emerge within many roles, whether of the independent businessperson, the salaried manger, or the financier. ‘But what­ever the type, everyone is an entrepreneur only when he actually “carries out new combinations,” and loses that character as soon as he has built up his business, when he settles down to running it as other people run their businesses.' Finally, ‘[b]ecause being an entrepreneur is not a profession and as a rule not a lasting condition, entrepreneurs do not form a social class in a technical sense, as, for example, landowners, or capitalists, or workmen do'.^[861] An economy could allow for displays of entrepreneurial genius just as warfare might allow for the expression of military genius, but such performances were necessarily sui generis, and could in the end be identified only by their consequences.

Schumpeter's contemporaries took different positions on the precise nature of the entrepreneurial function and the abilities it demanded, but they shared this general view of entrepreneurship as a productive factor within an eco­nomic system. Frank Knight, as we have seen, sought to justify free enterprise as a response to the uncertainty of productive activity. A producer had to ‘take the responsibility of forecasting the consumer's wants'; as an economy grew larger and more complex, ‘a large part of the technological direction and control of production are still further concentrated upon a very narrow class of producers, and we meet with a new economic functionary, the entrepreneur'. The function of the entrepreneur was to bear the uncertainty of economic life, to take responsibility for decisions on production and to reap the profits or losses that resulted. In practice, then, ‘the primary function or problem' of the entrepreneur was not executing a plan, but ‘deciding what to do and how to do it'. The more diversified and technologically complex an economy became, the greater the uncertainty involved in this function of decision. This generated in turn a greater specialisation of the decision function in entrepreneurs, and placed a greater premium upon the abilities that Schumpeter had identified: the ability to exercise foresight and seize the initiative. Entrepreneurs were ‘acting, competing on the basis of what they think of the future'. As with Schumpeter, Knight rejected the notion this ability defined a vocation: it was a position of authoritative decision-making into which a person could come from many existing occupations and social positions. The problem of entrepreneurship was thus one of determining the optimal supply of a factor, ‘entrepreneurial ability'.^[862]

Once entrepreneurship is made over in this way into a factor of production - albeit of a peculiar sort - the question arises of the means by which it is supplied.

The drawing forth of entrepreneurial ability thus depended upon the evolutionary pressure exerted by the fact of ineradicable uncertainty in eco­nomic life. Knight did not romanticise the realities of free enterprise: business competition in the modern world, said Knight, was marred by monopolisation and inefficiency. Nevertheless, the ever-present spectre of uncertainty, and the promise, however distant, of profit it offered, did encourage the development of a system of private owners ready to gamble their assets in search of business success. The only alternative to letting free enterprise evolve as a quasi-natural process was to substitute public for private ownership.

Trust ‘natural' processes of competition; resist the conscious planning of production - this was the message of innovation theory in economics. This observation helps us, I think, to explain the tendency of accounts of concep­tual innovation to invoke naturalistic tropes, and to underplay the importance of intentional collective action in bringing about conceptual change. Models of conceptual change that employ the analytics of enterprise will often treat innovation as a property of a system which evolves in a non-intentional manner - hence the descriptions of conceptual change as quasi-natural moments: ‘speciation', ‘epidemics', volcanic ‘eruptions', and so on. What looks on the face of it like a theory of conceptual change that places agency front and centre will often in practice back off from the claim that innovation is a collective, self-conscious process. Innovation is, in this sense, something we must understand, and prepare for, but which we can never catch up to or reliably control. This is perhaps its own kind of philosophy of history, albeit one that is profoundly open-ended. In fact, Reinhart Koselleck's definition of the modern idea of progress comes close to what we have in mind when we think of innovation. In the concept of progress, Koselleck writes,

is contained the idea that following industrialization and the growth of technology, the conditions of our prior experience will never suffice to predict coming surprises and innovations. Since the eighteenth century, progress produces a necessity for planning, but its goals must be constantly redefined as a result of the steady influx of new factors. The concept of progress encompasses precisely that experience of our own modernity: again and again, it has yielded unforeseeable innovations that are incomparable when measured against anything in the past.⁶

It is worth underscoring, in conclusion, the extent to which this fatalism about the twin inevitability and uncontrollability of innovation under modern indus­trial capitalism rests on matters of political judgment. For Knight, only an ideal form of democracy, radically different from the reality of competitive democratic politics, stood any chance of nurturing entrepreneurial ability in a socialised economy.^{[865] [866]} Kuhn, too, bucked the consensus among advocates of the democratic control of basic scientific research by insisting that only a decentralised system of inquiry, guided by esoteric professional norms, could produce innovation.^[867]

Yet for some of its theorists, innovation did not necessarily demand a kind of neoliberal regime, in which competitive conditions, but not developmental ends, were set by political institutions. Schumpeter, for his part, began to allow himself over the years to consider the possibility that innovation might either cease under socialism, or simply become automated.^[868] He came less and less to think of innovation as the act of a demiurge-entrepreneur, and more as the function of an ability that was scattered throughout a population.^[869] There was no special reason why a socialist regime could not develop this productive factor, just as it was not (as Schumpeter saw it) impossible for a socialist economy to solve the calculation problems involved in allowing consumer demand to determine the allocation of resources to production.

Perhaps the most illuminating commentary on the politics of innovation came from a figure a generation younger than Schumpeter and Knight, the British economist Maurice Dobb. In Capitalist Enterprise and Social Progress (1925), which we can loosely call a pre-Marxist study by an author soon to embrace socialist economics more fully, Dobb made clear that the presence of uncertainty was to an important degree itself a political choice. A good deal of the uncertainty that Knight identified, such as uncertainty of income and uncertainty about the intentions of others, was ‘relative only to individualist conditions'. If it was recognised that the entrepreneur function (which Dobb defined as the capacity for adjustment and innovation within an economic system) could be realised in at least three systems of enterprise, including a socialist one, then ‘the importance of uncertainty to the problem and the burden it imposes is not so great as those who generalize uncritically from an individualist society lead one to suppose'.^[870] Radical uncertainty was the product of decentralised, competitive production - it was one of the wages of capitalism.^[871]

When, in the mid-1930s, Dobb made his contribution to the socialist calculation controversy, his argument turned on the issue of how free enterprise and socialist economies dealt with the challenge of rapid innov­ation and the increased rate of capital accumulation such periods of rapid technical change produced. In a system of private enterprise, Dobb noted, it would only be profitable to invest in new technical processes once interest rates had fallen in the face of a growth in capital accumulation. But that would mean that investments in these new technical processes would come at some later date, and that in the meantime capital would be invested in plant that was already on the way to obsolescence. This entailed both waste of investment in an already outmoded technology and the loss of benefits derived from the longer-term usefulness of the most innovative new technologies. In contrast, the state in a socialist economy could lower the rate of social discount on those proven new technical processes that would only be appealing for investors in a private enterprise economy perhaps ten or twenty years hence. Instead of giving over the process of innovation to the evolutionary develop­ment of the capitalist economy, the collective action of a political community through the state could actively secure the benefits of innovation as a public good. Dobb gave a vivid illustration of this difference in a discussion of the ‘pursuit-curve':

A dog is situated at right angles to the path along which his master is bicycling. The dog is running towards his master, and, influenced by a simple conditioned reflex, runs always in the direction of his master at the given moment; with the result that his path in pursuit of his master is a curve. But if the dog could have acted on forethought and calculation, he would have taken a straight line to the point along the path which his master would presently reach. A planned economy, it would seem, should take a similar line towards a technical level of the future; and the ultimate economising of capital to produce a given result (or, conversely, the more rapid rate of technical advance financed by a given rate of investment) will be the difference in length between the straight line and the curve.^[872]

At the time, Dobb's intervention pleased neither his socialist interlocutors nor the critics of planning. But it was directed at issues of economic innovation and political control that waxed in importance as the calculation controversy cooled. In the post-Second World War years, questions of economic innov­ation and growth moved to the centre of the political stage. Debates about the optimal rate of social discount swirled around precisely the dilemmas of investment in long-term public infrastructure projects that Dobb had con­sidered. The policy of lowering the social discount rate for projects with pay­offs for multi-generational publics (but not necessarily for consumers in the present) received a favourable hearing from a new cohort of economists with a good understanding of problems in political philosophy.^[873] Meanwhile, discus­sions of the optimal allocation of resources to research and development increasingly emphasised the shortcomings of market provision and the need to treat scientific research - that is, innovation - as a public good.^[874] Dobb's conception of innovation as a political value, and his view of uncertainty as to some degree a political choice about the organisation of an economy, were thus not wholly misguided.

I record these debates not because there is any easy translation of arguments for innovation planning into discussions about conceptual change. I do so simply because the tendency to think of innovation as a kind of force of nature, defined by radical endogenous change, uncertainty, incommensurability, and daimonic creativity, itself bears the marks of ideo­logical struggle. In its efforts to model conceptual change on the theory of innovation, the study of conceptual change is itself embroiled in a political argument about the kinds of control we can exercise over the processes of cultural and ideological development. Most often, writers on conceptual change have underplayed agency and collective action and embraced natur­alistic metaphors in making sense of cultural change. Such a move concedes, in practice, the argument to those who believe that the process of discovery - the creation of novelty through research - cannot be subject to democratic control and collective decisions on the ends of inquiry. At the very least, the vigour with which this claim was contested by economists of technical change during the twentieth century suggests that we need not share this fatalism about the wellsprings of conceptual change. Here, then, is another way in which we can speak of a politics of context - or perhaps, as the foregoing suggests, a political economy of context.