It is not clear just how Sellars thinks of the "framework of givenness" against which he is arguing. Sometimes he seems to identify it with any commitment to immediate knowledge or justification (immediate in the sense of not owing its epistemic status to other knowledge or justified belief). But I will take his target to be a more distinctive one -- the idea that there is immediate knowledge or justified belief that acquires that title from support by a nonconceptual, nonpropositional direct experience (in which something or other is "given" to the awareness of the subject).

    Much of Sellars's philippic against the given is directed at its sense-datum form. Since I hold no brief for that version, I will defend the given in the form it takes in the Theory of Appearing. According to that theory, at the heart of perception are facts of the form X appears P to S, e.g., the book looks red to Jones. When that is true Jones enjoys a direct, nonconceptual awareness of the book. Such a fact is irreducibly relational. Thus the givenness involved in the Theory of Appearing is a givenness of particulars. I agree with Sellars that there is no nonconceptual givenness of facts, and that we must be careful to keep distinct these two kinds of alleged givens.

    Where Sellers makes contact with the Theory of Appearing is in his discussion of looks. He proposes to interpret 'X looks green to Jones' in terms of its relation to 'Jones sees that X is green'. "Whereas the latter both ascribes a propositional claim to Jones's experience and endorses it, the former ascribes the claim but does not endorse it". (Empiricism and the Philosophy of Mind [EPM], 145.) This construal leads him to deny that looks is any kind of a relation. Thus he comes into conflict with the Theory of Appearing both in that denial, and in the thesis that any perceptual experience includes a "propositional claim." More basically, he opposes the commitment of the Theory of Appearing to a kind of cognition of perceived objects (the one that is involved in being appeared to by objects) that is not conceptually structured. And the above account of look-locutions is at the heart of his argument. I contend that his account ignores the diversity of ordinary look-concepts and the light this throws on what it is for X to look P to S. In particular, there is the difference between a phenomenal concept, in terms of the intrinsic qualitative distinctiveness of the look, and a comparative concept, in terms of what sort of object will typically (standardly, normally) look that way. The former is more fundamental than the latter, for it makes explicit what is indicated only indirectly in the latter -- just how X looks to S. Sellars's account fails to deal with the intrinsic character of looks, and as a result does not adequately bring out what is being said by 'X looks P to S' (Sellars's theoretical concept approach to "private episodes" in the latter part of EPM does attempt to take account of the intrinsic character of perceptual experience. But the fact remains that he ignores the way in which ordinary look concepts do so.) Hence his criticism leaves standing the kind of givenness of particulars embodied in the Theory of Appearing.

    Since Sellars's attack on the given is obviously driven by an epistemological concern to discredit immediate knowledge, it is also pertinent to indicate how nonconceptual appearings can serve as a basis for knowledge and justified belief. The intuitive plausibility of the claim that when a belief that X is P stems from X's looking like P is thereby prima facie justified, is shown not to be shaken by claims that a causal role of experience is incompatible with a justificatory role. Sellars, more explicitly elsewhere, attempts to rule out justification by direct experience by maintaining that justification is only by reasons (propositionally structured justified beliefs or knowledge). I suggest that his claim owes whatever plausibility it possesses to a conflation of being justified and the activity of justifying.