Nature of Naturalism: Adrian Haddock

McDowell understands human sensibility both as a natural item and as involving higher cognitive capacities. However, there are complete constitutive explanations for natural items in terms of law-governed first nature. Naturalistic explanations of human sensibility therefore threaten to blank out explanations of it in terms of the space of reasons. This threat can be deflected in three basic ways, familiar from McDowell:

(1) Sensibility is part of first nature, but the space of reasons isn’t sui generis.
(2) Nature is first nature, the space of reasons is sui generis and sensibility is not part of nature.
(3) The space of reasons is sui generis, sensibility is not part of first but of second nature.

As is well known, McDowell rejects option (1) and settles for option (3). The apparent advantage of that option is that there is no exhaustive first-nature explanation of items in second nature so that sensibility can (partly) be accounted for in terms of the space of reasons without abandoning naturalism altogether.

Adrian Haddock argued in his engaging paper that given McDowell’s intentions, the best option for him is not the one he chooses but in fact option (2). In particular, McDowell’s choice of option (3) undermines his recent Hegelian critique of Kant’s position on the contribution of our sensibility to our knowledge. For Kant, the conjunction of sensible intuitions and categories to which the objects of sensible intuitions conform is meant to provide us with empirical as well as a priori knowledge. This account is deficient, in McDowell’s view, because it does not allow us to know a priori what the appropriate character of sensibility (spatiotemporal or other) is, that is, the character of sensibility in virtue of which the objects intuited by us conform to our categories. Without this knowledge, specifically human sensibility and the sensibility of thinkers qua thinkers may come apart. To keep them together McDowell suggests along Hegelian lines that the specific character of sensibility must issue from the higher cognitive faculty.

Obviously, this doesn’t square with McDowell’s second-nature account of human sensibility. Second nature, to which human sensibility belongs, changes when first nature changes. The effect is the same as that of the impossibility, on Kant’s picture, to have a priori knowledge of the character of sensibility. While changes in second nature don’t affect the higher cognitive faculty of thinkers, human sensibility may change and become non-human, for instance, so that specifically human sensibility is not the sensibility of thinkers qua thinkers. Adrian concluded that to ensure that the sensibility of thinkers has the appropriate character, sensibility should be an aspect of the higher cognitive faculty, not of nature, as option (2) states. McDowell should abandon his compromised empirical concept of the thinking self and adopt the concept of a (more fully Hegelian or Roycean) absolute ego instead.

In the discussion Adrian added that this conclusion – option (2) makes more sense of McDowell’s intentions than McDowell’s preferred option (3) – could also motivate an endorsement of option (1). As long as empirical knowledge is the only relevant issue, there is no harm in giving a complete constitutive explanation of human sensibility in first-nature terms. This might prove more appealing overall than operating with an absolute concept of the self. Another point raised in the discussion concerned the variability of first nature. If first nature were to include everything in space and time, for instance, it would be essentially static, and evolving second nature would not simply follow first-nature changes. Consequently, changes in human sensibility, an item of second nature, would not necessarily disturb the co-ordination between sensibility and higher cognitive faculty in the way suggested by Adrian’s conclusion.