A note on Foucault: limits to the conditions of possibility?29

The apparatuses of scientific (and arguably of social science) production produce something akin to what Michel Foucault described as the conditions of possibility. If we go with the economic metaphor then they set necessary limits – more or less permeable, but nevertheless limits – to those conditions.

So how does the present argument differ from that of Foucault? One answer has to do with empirical scope. Foucault and his interpreters insist that there is endless possibility for variation and creative innovation within the existing conditions of possibility.30 Nevertheless, it is also well known that Foucault argued that the current conditions of possi- bility were established at the end of the eighteenth century in a set of strategies laid down within what he called the modern episteme. The argument is that at the beginning of the twenty-first century we are still being produced by that episteme and its conditions of possibility.

This may or may not be right. However, the picture of natural (and social) science production proposed by Latour and Woolgar and other STS scholars is drawn on a smaller scale. Perhaps there are larger limits set by modern disciplinary strategies that lie within and are being enacted by the different inscription devices and practices of modern natural and social science. But Latour and Woolgar’s suggestion is more modest. It is that the limits to scientific knowledge and reality are set by particular and specific sets of inscription devices. The relations between these become an empirical matter.

Given the flexibility of the modern episteme, the position is not necessarily inconsistent with that of Foucault. Further it shares with him the commitment to the idea that it is not simply knowledge of realities, but also realities themselves, that are generated in the practices of production. My question, and one to which I will return in Chapter 3, has to do with singularity. Latour and Woolgar tend to assume that inscription devices (and so their hinterlands) mesh together fairly well. This seems to me uncertain.

Дата добавления: 2015-11-14