Reading Immaterialism by Graham Harman, I’m struck by the overlap between his account of ‘duomining’ and Margaret Archer’s critique of conflation. As he writes on pg 27-28,
“If we reduce an object downward to its pieces, we cannot explain emergence; if we reduce it upwards to its effects, we cannot explain change.”
While Archer’s argument is made in terms of the structure/agency problem, it can easily be recast in terms of structure alone. If we reduce social structure to the individuals who comprise it (alongside other material elements, which Archer is less sensitive to), we cannot explain how certain arrangements of people and things assume characteristics which the same ‘pieces’ lack in other arrangements (upwards conflation). If we focus solely on the effects of social structure, identifying how it constrains and enables individuals, we cannot explain how that structure might itself undergo change because it is the only causal power we admit (downwards conflation).
However this is only an overlap, as Archer and Harman’s arguments about modes of reduction are made for different reasons and they later diverge. Archer is concerned with the analytical temptations which inhere in the structure/agency problem that social science invariably confronts, even when it attempts to suppress it through various means. In contrast, Harman is concerned with ‘undermining’ and ‘overmining’ as two fundamental forms of knowledge which cannot be avoided: “what a thing is made of” (undermining) and “what a thing does” (overmining) (pg 28). Archer is concerned with a denial of relationality, as well as its temporal unfolding, with downwards and upwards conflation charged with suppressing the interplay over time between the different kinds of entities which make up the social word. Harman is concerned with the denial of objects as such, reducing their reality to the parts and their effects, losing a grip on the entity which is composed of these parts and capable of these effects without being reducible to either.
Both approaches explore a tension between the analytical and the ontological. Harman’s notion of overmining, which I found much less straightforward to grasp than his notion of undermining, identifies its roots in the tendency to treat objects as mysterious and unknowable in themselves. An ontological claim licenses an analytical one, as the analyst focuses upon the effects of objects as something epistemically tractable in contrast to the objects themselves. Even if they continue to recognise the reality of the object, it is a notional recognition which doesn’t enter into their analysis. This is something Harman addresses explicitly on pg 28:
After all, any claim that a thing is convertible into knowledge cannot account for the obvious and permanent difference between a thing and knowledge of it: if we had perfect mathematised knowledge of a dog, this knowledge would still not be a dog. It will be said that this is a “straw man” argument, since philosophers are obviously aware that knowledge is different from its object. Yet it is not a question of whether philosophers are personally “aware” of this, but of whether their philosophies sufficiently account for it.”
To which we might add: ‘and whether they incline social scientists drawing on their ideas to factor this awareness into their explanations’. This interface between the ontological and the analytical one is one that has long fascinated me: how does theory constrain and enable the explanations which enter into social inquiry? What other forms of ‘conceptual slippage’ can we identify as ontological claims contribute to social analysis?