Brandom’s paper presents a distinct social reading of knowledge. In addition, he gives a more concrete characterisation of what one might mean by the phrase ‘the space of reasons’. It stands – he suggests – for an abstraction from the actual practice of giving and asking for reasons. That is a human practice with rules governing what the right ‘moves’ are. So if knowledge is a standing in the space of reasons, then knowledge is not so much a mental state as a social status. It is ascribed according to rules. Thus it is normative. (By contrast, happiness is a state of mind and is not a normative. Happiness is not correct or incorrect.)
Brandom’s positive picture of knowledge.
Brandom’s idea is to give an account of knowledge based on how we ascribe knowledge to others. He does this using two explicitly normative (ie correctness involving) notions: commitments and entitlements. (The reason for picking these is that they do work in his broader account of language works. But they will help shed light on knowledge if we think they are clearer, more basic, simpler than ‘knowledge’ ‘justification’ etc.) Thus belief, being justified and being true work like this:
Construed as a standing or status, belief will correspond to some sort of commitment, while justification (being justified) will correspond to some sort of entitlement to that commitment. So taking someone to have a justified belief will be understood as attributing two sorts of standings: a commitment and an entitlement. What about the truth condition? To take someone to have the status of a knower one must take it that the justified belief in question is also true. What is it to do that? Taking the belief in question to be true is not a matter of attributing a commitment, but of undertaking one-endorsing the claim oneself. For taking-true is just believing, that is, committing oneself, adopting a standing or status. [903]
So ascribing knowledge involves adopting three different attitudes: attributing a commitment, attributing an entitlement, and undertaking a commitment.
Note, now, that thinking of knowledge by thinking of it from this perspective involves us, say, ascribing knowledge to him, say, by noting that he has taken on a commitment (he is committed to it being raining), by us ascribing an entitlement (we think he can see the rain from where he’s standing) and by us thinking his commitment is true. That is, we take on the same commitment.
So he does not need to have a view about his own entitlement. We need to ascribe it to him.
McDowell’s response
Although Brandom says that he is attempting to provide an interpretation of McDowell and to add to it an explicit social dimension his characterisation of McDowell comes as a bit of a surprise.
Here is the bit of Brandom in which he summarises McDowell’s strategy:
McDowell's argument is structured by a botanization that classifies approaches to knowledge as coming in four flavors: sceptical, dogmatic, hybrid, and extreme externalist. We can group these further according to whether they conceive justification and truth as internally or externally related-or as I will say, according to whether they aggregate or segregate these conditions. The sceptic and the dogmatist take it as a criterion of adequacy on a notion of justification that any claim or belief that is sufficiently justified is true. They are right that if a claim or belief has the status of knowledge, it is guaranteed to be true. But they also take it that justification of a certain sort is what distinguishes knowledge from other belief. If that is right, then justification must be truth-guaranteeing. While agreeing on this basic principle, the sceptic and the dogmatist disagree about whether a notion of justification meeting this condition is to be had. The sceptic arrives at the false conclusion that knowledge is not possible by combining the false claim that justification must be incompatible with falsehood with the true claim that justification that rules out the possibility of falsehood is not to be had. The dogmatist arrives at the true conclusion that knowledge is possible by combining the false claim that justification must be incompatible with falsehood with the further false claim that justification that rules out the possibility of falsehood can be had. McDowell rightly does not rehearse at length the difficulties of these views; their unsatisfactoriness is widely acknowledged. [Brandom 1995: 899]
But this is really odd. Brandom says that the sceptic and the dogmatist agree that justification must be truth-guaranteeing and that McDowell rejects both. But a key point of McDowell’s rejection of putting justification on one side of a divide (what a subject can ensure by her own powers unassisted by a favour from the world) and truth on the other (what is provided by the world) is that then justification is never sufficient for truth. In other words: McDowell thinks that when one has knowledge one does have a justifiucation which is sufficient for truth, which is truth-guaranteeing.
Here’s a passage from McDowell’s reply to this paper:
Compare Brandom’s §II (899-901). where he purports to restate my argument. Brandom lists four positions that he says my argument rules out. About one of them, “dogmatism,” he writes, supposedly in agreement with me (899): “The dogmatist arrives at the true conclusion that knowledge is possible by combining the false claim that justification must be incompatible with falsehood with the further false claim that justification that rules out the possibility of falsehood can be had.” But both these supposedly false claims are true by my lights. What I urge in my paper is precisely that justification adequate to reveal a state as one of knowing must be incompatible with falsehood and can be had. “Dogmatism,” one of the four positions Brandom says I reject, is precisely what I defend. [McDowell 2002: 98]
So much for whether Brandom is a good reader of McDowell. But if Brandom has this wrong, what of his own positive position? In the next passage McDowell points out that Brandom falls prey to the very thing McDowell criticises: putting truth and justification on either side of a divide.
What I object to is interiorizing entitlements, in the sense of refusing to let the connivance of the world enter into constituting them. Applied to the entitlements that perceptual, for instance visual, experience affords, the interiorizing move restricts them to appearances, conceived as a highest common factor between seeing that such-and-such is the case and having it merely look to one as if such-and-such is the case. I argue that it is not satisfactory to leave entitlements thus interiorized but add that what the putative knower takes to be so is in fact so, conceiving this as an extra condition over and above an interiorized entitlement. Now Brandom’s socially perspectival hybrid conception of knowledge attributions has just that shape. It makes no difference that he can take over my phrase, “standing in the space of reasons,” and define it so that it includes the satisfaction of the extra condition. The extra condition is still seen as extra to the knower’s entitlement, and that is what, according to the argument of mine that Brandom purports to endorse, precludes making sense of the status in question as one of knowledge. [McDowell 2002: 102]
Here’s the McDowellian worry. On Brandom’s account, two subjects can be equally entitled to a commitment (= equally justified in a belief) but one can be true and the other false. If so, McDowell argues, even the subject with a true belief lacks knowledge. Brandom’s social account thus undermines the possibility of knowledge, if McDowell is right about this.
A further lingering worry.
Does Brandom’s account really connect up with the world in a sufficiently objective way? After all, Brandom captures the truth aspect or condition of knowledge by asking how we ascribe truth (or knowledge) to another. We do that by sharing the commitment. But whilst that unpacks the social rules governing the space of asking for and giving reasons, might it not simply stay within the realm of what we - socially - take to be true, not what is true? It is not so much that he has attempted to reduce what is truth to what we take to be truth. Rather, he has not attempted to unpack the notion of truth at all. Still, might we not want some reassurance that truth will not just collapse into what we take to be true?