In this paper, I am concerned to show that a wide and interesting range of phenomena commonly described as ‘weakness of will’ should be understood as manifesting a defect of what I shall call ‘executive cowardice’ rather than a strong kind of irrationality. More specifically, I claim that such cases should not be understood as an irrational bypassing of an all-things-considered judgment about the thing to do—a view succinctly described by Peacocke thus:

Laplacean determinism remains a popular theory among philosophers and scientists alike, in spite of the fact that the Copenhagen Interpretation of quantum mechanics, with which it is inconsistent, has been around for more than fifty years. There are a number of reasons for its continuing popularity. One, recently articulated by Honderich, is that there are too many possible interpretations of quantum mechanics, and the subject is too controversial even among physicists to be an adequate basis for overturning determinism. Nevertheless, quantum mechanics is an enormously successful theory, considering the quantity and variety of its predictions which have been verified under conditions never dreamt of by its originators; and the Copenhagen Interpretation is the only widely accepted interpretation of it. Although a hidden variable theory consistent with the results of quantum mechanics is not impossible, one of its major advocates admits that it is highly speculative, and far from adequately developed. Yet such a theory would be needed to reconcile Laplacean determinism with quantum mechanics; most of the controversies alluded to by Honderich are irrelevant.

Kant supposed that we possess two distinct cognitive capacities, which he referred to as ‘intuition’ (Anschauung) and ‘understanding’ or ‘intellect’ (Verstand). This ‘two-faculty account of cognition’ lies at the foundation of his theoretical philosophy, and almost everything he has to say in the Critique of Pure Reason presupposes it. But it is also problematic. At the outset of the Critique Kant simply assumes the validity of the distinction, without in any way attempting to justify it. And one looks in vain through the Kantian corpus for any explanation that might legitimate it. To make matters worse, Kant does not always draw the distinction in the same way. Most notoriously, he presents two quite different accounts of intuition, defining it in some places as ‘singular representation’ (A713=B741; Logic §6), in others as ‘immediate cognition’ (A19=B33).

The most embarrassing thing about ‘facts’ is the etymology of the word. The Latin facere means to make or construct. Bruno Latour, like so many other anti-realists who revel in the word’s history, thinks facts are made by us: they are a social construction. The view acquires some plausibility in Laboratory Life: The Social Construction of Scientific Facts (hereafter LL) which Latour co-authored with Steve Woolgar.1 This work, first published a decade ago, has become a classic in the sociology of science literature. It is in the form of field notes by an ‘anthropologist in the lab.’ This may seem an odd place for an anthropologist, but Latour finds his presence easy to justify. ‘Whereas we have a fairly detailed knowledge of the myths and circumcision rituals of exotic tribes, we remain relatively ignorant of the details of equivalent activity among tribes of scientists … ’ (LL, 17).