Davidson's general approach to metaphysics follows a long-standing tradition in trying to derive the basic features of reality from the structure of language. The particular angle he introduces comes from his suggestion that the structure of language and its revelations about the large features of reality are refracted in the effort to formulate a comprehensive, formal theory of truth for a canonically regimented version of natural language (see Davidson, 1984, essay 13; for some criticisms of this approach to metaphysics, see Rovane, 1986).
The most original specific application of this method to metaphysics is to be found in Davidson's claim that such a theory for that fragment of language which contains sentences such as “The boiler exploded”, yields an ontology of events. That is to say, it yields the idea that events such as explosions exist as particulars, in the way that boilers do. There is something very natural and appealing about the suggestion because, among other things, it allows for our re-description of the same event in different ways. Thus one can say of the same event that it was an explosion and that it was a domestic disaster (see Davidson, 1980, essay 9; for a different, more universalist view of events, see Chisholm, 1970. For a different view of events as particulars, see Kim, 1976. Davidson discusses the criterion for individuation of events in 1980, essay 8, and in Davidson, 1985.)
Davidson extends the point to sentences such as “Jones buttered the bread”, arguing that these sentences about human agency too are to be treated as quantifying over events which get intentional descriptions. He points out that this affords a satisfying theoretical treatment of sentences with adverbial modifiers. In the standard predicate calculus, adverbially modified predicates are usually represented as distinct predicates, but that representation fails to capture the seeming validity of the argument which goes from, say, “Jones buttered the toast with a knife” to “Jones buttered the toast.” There is simply no sanction for Ga from Fa. But the intuitive validity of the argument is preserved if we take events to exist as particulars, and treat the canonical representation or “logical form” of such sentences as quantifying over them. Thus we may go from Ǝx (x was a buttering of the toast by Jones, and x was with a knife) to Ǝx (x was a buttering of the toast by Jones) (see Davidson, 1980, essay 6. There are, of course other suggestions in the literature for handling adverbial modification, such as, for instance, that one should introduce not quantification but modifiers of predicates.)
He further exploits the point to give an analysis of sentences citing causes, such as. “His pressing the button caused the explosion.” The ontology of events allows him to make a distinction between two different aspects of what these sentences convey: causal relations which hold between events and which are purely extensional, and causal explanations which are intensional in the sense that they, unlike causal relations, depend upon how the events are described (Davidson, 1980, essay 7).
Davidson, then, develops this distinction to provide a solution to the traditional MIND/BODY PROBLEM (see the extended essay). Mental events are identical with physical events, but when we gather these mental events into types, there is a principled objection to their being identical with the types we gather physical events into. This is because unlike the particulars (the tokens of mental events), the types are essentially dependent on the concepts we employ in describing the events; and there are no lawlike correlations between these mental concepts and the concepts we employ in physical descriptions of events. So, at the level of concepts there is no reduction of the mental to the physical, but, as far as ONTOLOGY goes this does not imply a dualism since any mental event is identical with some physical event or other. He calls this hybrid position, “anomalous monism”, and the view of identity it proposes, “token-identity” (Davidson, 1980, essay 11).
In addition to token-identity, Davidson's metaphysics of mind posits a dependency relation between the mental and the physical, which he calls SUPERVENIENCE. This is the idea that a psychological predicate will not distinguish anything that is not also distinguished by some physical predicate(s). The underlying motivation for such a dependency was to make a claim in the philosophy of mind parallel to a position in the study of values which both denied a reduction of values to facts of nature at the same time as it did not make values mysteriously autonomous. In positing supervenience, Davidson was able to claim that a denial of a definitional as well as a nomological reduction of mental properties to physical properties was compatible with a dependency relation between them which disallowed one from saying that two things were indistinguishable physically but were different in some mental respect. This allowed the scientific study of physical nature maximum comprehensiveness in its dominion without any concession to mind/body reductionism. More recently he has also invoked supervenience to quell a worry which has loomed at least since Descartes – that the mental is epiphenomenal. In contemporary discussions this worry has sometimes been expressed as the worry that mentality makes no difference to causal relations. (See Kim, 1984 and Sosa, 1984 who raise this worry for Davidson's anomalous monism, in particular. For a response on behalf of Davidson, see LePore and Loewer, 1987.) So expressed, supervenience is not needed to deal with it. The worry is handled in Davidson's metaphysical framework by simply appealing to the extensional nature of causal relations. Since events which enter causal relations are often described in mental terms, mental events (uncontroversially) enter causal relations. But for him it makes no sense to go on to ask whether mental events make a difference to causal relations in the sense that they cause other events in virtue of being mental. Causal relations being purely extensional do not hold or occur in virtue of anything, mental or physical. However the worry about epiphenomenalism is sometimes expressed as not being about causal relations in particular, but more generally as, say, the idea that we may alter (in the limit, even strip all) mental properties without at all affecting the physical properties of things. It is to this worry that Davidson responds by pointing out that if it were true it would contradict the weak dependency relation of supervenience as characterized above. However this subject throws up many questions issuing from modal intuitions about identity, intuitions whose relevance Davidson has always been suspicious of; these questions are at present the subject of much controversy in metaphysics and the philosophy of mind.
There are other aspects of Davidson's philosophy – such as his views on REALISM, OBJECTIVITY and the nature of truth – which may be treated as being part of metaphysics, but because of their integral relation with epistemological themes, are best discussed within epistemology.
See also EVENT THEORY; REDUCTION, REDUCTIONISM; THEORIES OF TRUTH.
American, b: March 1917, Springfield, Massachusetts. Cat: Philosopher of mind; philosopher of language. Ints: Casuation; meaning. Educ: ...
1917-2003 Donald Davidson was an American philosopher who developed some distinctive ideas about language, truth, and interpretation as well as...
b. 1917, Springfield, Massachusetts, USA Philosopher Davidson works in the Anglo/American or analytic tradition and is known for his essays in...