McDowell introduces his Mind and World this way: “My aim is to propose an account, in a diagnostic spirit, of some characteristic anxieties of modern philosophy – anxieties that center…on the relation between mind and world…I aim at explaining how it comes about that we seem to be confronted with philosophical obligations of a familiar sort, and I want the explanation to enable us to unmask that appearance as illusion” (xi).
What are these obligations? Well, consider a “minimal empiricism”: The belief that p is true if and only if p (xi-xii). More straightforwardly: “Thinking that aims at judgement, or at the fixation of belief, is answereable to the world” (xii). And this answerability is at least in the first instance empirical, “since…we confront the world by way of sensible intuition” (xii). This “minimal empiricism” is “the idea that experience must constitute a tribunal, mediating the way our thinking is answerable to how things are” (xii).
But, McDowell suggests, such an empiricism, full grown, would yield to skepticism as to the possibility of the mind’s contact with world (xiii). (Hume) Consider Wilfrid Sellars’s attack on “the Myth of the Given”: “knowing” isn’t simply a description of some state of affairs, but a normative state, placed within “the logical space of reasons, of justifying and being able to justify what one says” (quoting Sellars, xiv).
The opposite of the space of reasons might be “the space of nature…the realm of law” (xiv-xv). But surely “the idea of receiving [a sensory] impression is the idea of a transaction in nature” (xv). Now, “the logical space in which talk of impressions belongs is not one in which things are connected by relations such as one thing’s being warranted or corrected in light of another” (xv). “Being warranted by” is a logical relation, not a causal one. It can’t obtain between two physical states; “warrant” entails the presence of a mind.
If you’re committed to both the minimal empiricism and to the categorical distinction of the space of reasons from the space of nature, McDowell, you’ll generate an antinomy that will push you toward skepticism (xvi). That is, the minimal empiricism will suggest that “experience,” totally prior to all conceptual judgments, has to furnish some kind of backstop for beliefs; but the “space of reasons,” which for McDowell is most purely represented by Davidson’s coherentism, will suggest to you that only beliefs and never “pure facts” can justify other beliefs.
“One option [for resolving this tension] would be to renounce empiricism,” in the manner of Davidson,” but McDowell finds that this “does nothing to explain away the plausibility of the empiricist picture” of accountability to experience (xvii). Or, we could resolve it by “rejecting the dichotomy of logical spaces,” by making the space of reasons epiphenomenal to the space of nature (“bald naturalism”) (xviii).
McDowell proposes rejecting both of the above, maintaining the distinction of space of reasons (SOR) from space of nature (SON), while maintaining “both that the very idea of experience is the idea of something natural and that empirical thinking is answerable to experience” (xix). In short, “we can affirm that the idea of experience is the idea of something natural, without thereby removing the idea of experience from the logical space of reasons” (xix). So, the SON includes the SOR, but also includes a distinct logical space within which “the natural-scientific kind of intelligibility is brought to view” (xix). “The mistake here,” McDowell observes, “is to forget that nature includes second nature. Human beings acquire a second nature in part by being initiated into conceptual capacities, whose interrelations belong in the logical space of reasons” (xx). That is, our capacity for judgments in the space of reasons is continuous with our capacity for judgments just by receiving sensory impressions: “In receiving impressions, a subject can be open to the way things manifestly are” (xx).