It has turned wintery in Hyde Park.
Wednesday was the last meeting of the Philosophy of Mind Workshop this quarter. After discussing Chris Ferro's paper "The Normativity of the Mental", we turned to Making It Explicit, Chapter 2, parts IV and V.
PART I: The Set Up
We had a full workshop tonight. David F. got things moving with a quick summary of the most important topics in the sections we read:
1. Conceptual content: Brandom wants to understand conceptual content as inferential role, and he claims both Sellars and (more surprisingly) the early Frege as inferentialist forebearers. This way of understanding conceptual content is meant to contrast with a "representationalist" understanding of content, which would have us understand content as deriving from some relation between the constituents of content and things in the world (objects and properties, for example).
2. The inferences that are constitutive of content should be understand as material, rather than formal, inferences. There are inferential transitions between sentences we make in ordinary practice whose goodness does not need to be underwritten by a supressed conditional. For example, "If it's raining, the streets will be wet" or "Today's Tuesday, so tomorrow will be Wednesday".
3. Logic should be understood as playing an expressive role, making explicit the commitments that we implicitly undertake in our pre-reflective inferential practices. We shouldn't think of formal logic as somehow "standing behind" or already implicitly involved in our ordinary inferences. Logic is the device of semantic self-consciousness (and not, e.g. a necessary condition for thought as such).
David F. then raised two questions for the Brandom (not necessarily corresponding to the three central points just described):
Q1. Two of the examples Brandom gives of material inferences are (i) "Today is Wednesday" so "Tomorrow will be Thursday" and (ii) "Pittsburgh is to the west of Philadelphia" to "Philadelphia is to the east of Pittsburgh". Both of these examples look like "grammatical" inferences--intuitively, one could make them simply in virtue of knowing the meaning of the terms involved, and the inferences are necessarily (logically?) true. But Brandom's other example, "Lightning is seen now" so "Thunder will be heard soon" doesn't (seem to) have these features. I might understand the meaning of "lightning" (it's a certain kind of electrical discharge), but I might be a kind of creature insensitive to sound, so I wouldn't be disposed to conclude from seeing a lightning bolt that I will hear anything soon. Or it is possible that I see lightning but that I and everyone within earshot is struck dead--so no one hears any thunder. So, David F. asked, just what is a material inference?
Q2. Brandom says that conceptual content is inferential role. But an obvious concern for such an inferentialist account of conceptual content is how to explain the content of observational concepts, like "red". These concepts are sometimes invoked non-inferentially. Is their content then exhausted by their "downstream" inferential consequences? Brandom takes up this issue on p. 119, and says that with observational concepts like "red", "one is (among other things) committed to the propriety of the inference from its circumstances to its consequences of application". What exactly does it mean to infer from the circumstances of application of a concept like "red"? As Brandom says, the "circumstances of application need not themselves be linguistic" (119), and inferences (I would think) are transitions between linguistic (or sufficiently linguistic-like) things. So how can the circumstances figure in any inferences at all?
PART II: Discussion
I asked a question related to David's Q1. I pointed out that Brandom says that "It is the concepts 'Wednesday', 'Thursday', 'today' and 'tomorrow' that make the second inference correct, and the contents of the concepts 'lightning' and 'thunder', as well as the temporal concepts, that underwrite the third" (98). While it seems correct to say that it is the content of the concepts "Wednesday", "Thursday", "today" and "tomorrow" that make the first inference correct, it seems odd to say that the content of the concepts "lightning" and "thunder" make the inference about lightning and thunder correct. Following Rumfitt (1997), I asked whether it wouldn't be better to say that it is something about the world itself that makes the inference about lightning and thunder a good one. I think Jason B. was sympathetic to this question--arguing with David F. later, he glossed my (inarticulate) question as a worry (roughly) about frictionless spinning: "What is it about the facts that will fix content, when content is fixed by inference?"
An oh-so-very-philosophical moment occurred when Will S. asked how necessary the inference from "Today is Wednesday" to "Tomorrow will be Thursday" really is. He said that if the world were to explode, then tomorrow wouldn't be Wednesday. David and Jason simultaneously cried out, "Yes it will!" (But if time itself came to an end before midnight today [a scary thought], then would "tomorrow" even refer?)
There was some concern expressed by Zeke R. about Brandom's "K-vocabulary" (logical vocabulary). Jason B. shared the concern. He saw no principled way for Brandom to distinguish logical from non-logical vocabulary.
David F. and others tried to work out Brandom's account of observational concepts from the meager resources in the sections we read today, by thinking about how Sellars accounts for the content of observational concepts. But Zeke R. sensibly pointed out that even Brandom says we will have to wait until chapter 3 to get his account of the content of these concepts.
And wait we will. The Workshop will return during the first week of winter quarter.