Tuesday, October 11, 2005
The method of dispositional conceptual analysis
On Smith's view, which seems to me to be fairly standard, the project of providing an analysis of a concept C is the project of articulating a concept C* such that (a) for any x falling under C, x also falls under C*, and vice versa; (b) C* is unobvious; and C* 'will tell us something new and interesting about C.'
A dispositional conceptual analysis accomplishes this task by gathering together the various platitudes associated with the term which designates the concept in question. What Smith means to appeal to with the invocation of platitudes are the various simple declarative sentences which a competent user of the term would be expected to assent to. For example, one of the platitudes which Smith asserts to be associated with the term 'right' is the sentence, "When A says 'x is right' and B says 'x is not right' then A and B disagree with each other" (33). A platitude said to be associated with the phenomenology of color experience is that, "Everything we see looks coloured."
The notion that this collection of platitudes amounts to an analysis of a concept is based on a further contention about the role these platitudes play in our inferential practice. In discussion of a set of platitudes offered as a partial dispositional analysis of color terms, Smith writes that, "we come to master colour vocabulary by coming to treat remarks like these as platitudinous" (30). Treating the terms as platitudinous, in turn, seems to amount to accepting the various inferences and judgments implicit in the remarks as being valid a priori. So a proffered dispositional analysis of a term amounts to an analysis in that it is able to "capture the inferential and judgmental dispositions vis-à-vis the [term] of those who have mastery of the term" (30).
Smith provides the following summary gloss to his initial discussion of dispositional conceptual analysis:
…the dispositional analysis is perhaps best seen as an attempt to encapsulate or to summarize, or to systematize, as well as can be done, the various remarks we come to treat as platitudinous in coming to master [a] term…In this admittedly vague sense, it can therefore lay some claim to giving us knowledge of all the platitudes. And, accordingly, it can therefore lay some claim to constitute an analysis. (31-2)
This may be enough to show that a dispositional analysis meets condition a above, but it remains unclear how such an analysis could meet the other conditions. Smith addreses this issue a few pages later:
Why are analyses unobvious and informative? Because even though someone who has mastery of some concept C must have certain inferential and judgemental dispositions, it may not be transparent to her what these inferential and judgemental dispositions are, and so, a fortiori, it need not be transparent to her what the best summary or systematization of the platitudes that describe these dispositions is. Whereas mastery of a concept requires knowledge-how, knowledge of an analysis of a mastered concept requires us to have knowledge-that about our knowledge-how. It might therefore take time and thought to see whether or not C* constitutes an analysis of C because it takes time and thought to figure out what the relevant inferential and judgemental dispositions are and what the best systematization of the platitudes describing these dispositions is. (38)
1 This characterization is drawn from the discussion on p. 36-37.
2 It is, of course, possible that the sentences which we take to be platitudes might turn out to be false, or that the set of platitudes taken together might turn out to be inconsistent. Such possibilities don't undermine the project of giving a dispositional conceptual analysis, but rather, Smith argues, merely show that analysis should sometimes encourage us to reform our inferential habits. He goes on to write that, "At the limit, to give up on the platitudes associated with…terms…is to give up on using the…terms altogether. For they themselves have a prima facie a priori status, and gain a priori status simpliciter by surviving as part of the maximal consistent set of platitudes constitutive of mastery of the term" (31).