Evans, Gareth (1946–1980)
As an undergraduate from 1964 to 1967, Gareth Evans, a British philosopher of language and mind, studied for the PPE degree (philosophy, politics, and economics) at University College, Oxford, where his philosophy tutor was Peter Strawson. In 1968, less than a year after completing his degree, Evans was elected to a fellowship at University College. He took up the position in 1969, succeeding Strawson, who had become Waynflete Professor of Metaphysical Philosophy at Oxford. During the 1970s Evans and his University College colleague John McDowell played leading roles in developing a distinctive conception of truth-theoretic semantics, drawing on the work of Strawson, Michael Dummett, and especially Donald Davidson. Their coedited collection, Truth and Meaning: Essays in Semantics, appeared in 1976.
While philosophy of language enjoyed a central position in Oxford philosophy of that period, Evans did not share the view (regarded by Dummett as constitutive of analytic philosophy) that philosophy of language is foundational and so takes priority over philosophy of mind in the order of philosophical explanation. He attached particular importance to the mentalistic notion of understanding, and his work on the theory of reference was set within a theory of thought and especially thought about particular objects. Evans's published work ranged over philosophy of language, metaphysics, philosophy of mind, and philosophy of psychology. In 1979 he was elected to the Wilde Readership in Mental Philosophy at Oxford. He died in August 1980, at the age of thirty-four. His book The Varieties of Reference (1982), incomplete at the time of his death, was edited and brought to publication by McDowell. A collection of thirteen of his papers and two shorter notes appeared in 1985, and a further note was published in 2004.
Names and Reference
In his first published paper, "The Causal Theory of Names" (1973/1985) Evans contrasted two theories about the reference of names: the description theory and the causal theory. Evans agreed with Kripke (1972) in rejecting the description theory of reference, which he regarded as drawing support from a flawed account of what is involved in thought directed toward a particular object. In opposition to this description-theoretic account of object-directed thoughts, Evans maintained that a subject may think about a particular object in virtue of standing in a contextual relation to it and without being able to frame any description that the object uniquely satisfies. However, Evans did not accept the causal theory of reference suggested by Kripke's remarks. In the Kripkean picture, the reference of a name is established by an initial baptism and is then passed on from earlier to later users of the name. Evans challenged this picture by highlighting the fact that a name may change its reference over time and, more generally, he argued that a bare causal connection is not sufficient to underwrite reference. As against both the description theory and the Kripkean causal theory, Evans proposed that the bearer of a name is the object that is the dominant source of the body of information that speakers associate with the name.
Many of the themes of his early paper on names—including opposition to description-theoretic accounts of object-directed thoughts, rejection of causal theories as insufficiently demanding, and appeal to the notion of information—recur in The Varieties of Reference, set against the historical background of Gottlob Frege and Bertrand Russell. Evans read Frege as committed to the principle that if a name has no reference, then a sentence containing the name has no truth-value and does not express a thought; a speaker using the sentence does not literally say anything. This "no reference, no thought" principle is in line with Frege's view that the semantic function of a name is to introduce an object, but it appears to rule out the possibility of names with sense but no reference, a possibility that Frege clearly allowed once his distinction between sense and reference was in place. Evans sought to reduce the tension that this reading finds in Frege's position by appealing to Frege's assimilation of the use of empty names to fictional uses of language, which express pretended senses or "mock thoughts."
Evans held that many singular terms—especially demonstratives such as "that ball" or "that vase"—conform to the "no reference, no thought" principle, and he called such expressions "Russellian" singular terms. He also held, following Russell, that definite descriptions, even though they appear superficially to occupy name positions, are not really referring expressions but rather quantifier expressions; "the F " semantically resembles "some F " and "every F." The contrast between a Russellian singular term, whose significance depends on its having a referent, and a definite description, whose significance can be grasped independently of whether it has a denotation, was fundamental for much of Evans's work on reference (Sainsbury 1985).
Although The Varieties of Reference begins and ends with philosophy of language (returning to the topic of names and name-using practices in its final chapter), the central chapters address the issue of thoughts directed toward particular objects. According to the description theory of object-directed thoughts, thought about a perceived, remembered, heard-about, or recognized object, about an occupied place or about a present time, is a matter of the object, place, or time uniquely satisfying a descriptive condition that the thinker frames and deploys in thought. Alternative theories of de re (or object-directed) thought appeal to the causal relations implicated in perception, memory, and testimony and to contextual relations to places and times (Burge 1977). While Evans was opposed to the description theory, he was also concerned, here as in philosophy of language, that causal theories were liable to be insufficiently demanding. He was particularly critical of what he called "the photograph model of mental representation," according to which the causal ancestry of a mental state is sufficient to determine which object the state represents (as causal ancestry is sufficient to determine which object a photograph is an image of).
Evans's own theorizing about object-directed thoughts was guided by Russell's principle, which says that to think about a particular object, a thinker must know which object is at issue. Evans interpreted the principle as requiring discriminating knowledge, that is, the capacity to discriminate the object of thought from all other things, and this, at least initially, sounds so demanding as to make object-directed thought an extraordinary achievement. But Evans's examples of how to meet the principle make it seem more tractable: presently perceiving the object, being able to recognize it, knowing discriminating facts about it.
When a thinker meets the requirement of Russell's principle by having discriminating knowledge of a particular object, the thinker is said to have an adequate Idea of the object. In this technical use of the term, an Idea deployed in thought about an object is analogous to a concept deployed in thought about a property. Evans was particularly concerned with cases (centrally, cases of demonstrative identification) in which a thinker's Idea of an object depends on an information link between the thinker and the object, so that the Idea of the object, and thoughts in which the Idea is deployed, are information-invoking. The picture here is not that the information link contributes to the thought about the object, because the thinker frames a descriptive condition along the lines of "the object, whichever it is, that is the source of this information." It is the information link itself, and not a thought about the information link, that plays a role in making object-directed thought possible.
If, as a result of malfunction or hallucination, there is really no information link to an object, then the thinker has no adequate Idea of this information-invoking kind. A thinker who is unaware of the problem may essay a thought and yet fail to think about any particular object at all. Information-invoking thoughts (centrally, demonstrative thoughts) are object-dependent; where there is no object, there is no thought. Evans was especially interested in cases where understanding a singular term requires an information-invoking thought, and hence object-dependent thought, on the part of the hearer. For in such cases, it is possible to argue that the singular term is Russellian, that its significance depends on its having a referent.
Despite the central role played by Russellian singular terms in The Varieties of Reference, Evans did not equate the categories of Russellian singular terms and referring expressions. In "Reference and Contingency" (1979/1985), he considered descriptive names (names whose referents are fixed by description). His example was the name "Julius," introduced with the stipulation "Let us use 'Julius' to refer to whoever invented the zip fastener [or zipper]." "Julius" behaves epistemically and modally like the definite description "the actual inventor of the zip." Evans offered "If anyone uniquely invented the zip, Julius invented the zip" as an example of a sentence whose truth can be known a priori, even though it is contingent. Evans argued that the thought expressed by the nonmodal sentence "Julius is F " is the same as the thought expressed by "The inventor of the zip is F "—a thought that can be grasped whether or not "Julius" refers to anyone. But he rejected the suggestion that descriptive names belong semantically with definite descriptions and maintained that, although the descriptive name "Julius" is not a Russellian singular term, it is still a referring expression. His argument for placing descriptive names in the category of referring expressions, alongside Russellian singular terms and separate from definite descriptions, involved two main points. First, as the introducing stipulation makes clear, the semantic contribution of "Julius" is stated using the relation of reference, no less than is the semantic contribution of a Russellian singular term ("John" refers to John). Second, even in a semantic theory for a modal language, the semantic contribution of a descriptive name, like that of a Russellian singular term, can be stated using a reference relation that is not relativized to possible worlds, but this is not generally so for definite descriptions.
Information and Nonconceptual Content
The notion of information, as Evans used it, is not the notion of what a subject believes. Indeed, Evans suggested that we should take the notion of being in an information state as a primitive notion, not to be explained in terms of belief, judgment, and reasons. Because perceptual-information states can be present in a creature that does not think or apply concepts, Evans maintained that the representational content of perceptual states is a kind of nonconceptual content. To be in states with such content (perhaps the nonconceptual content that a sound is coming from direction d ) a creature does not need to apply, or even to possess, the concepts that we use to specify the content of the states (concepts such as those of sound and direction).
Evans held a distinctive view of the relationship between perceptual-information states and perceptual experiences according to which conscious perceptual experience requires that perceptual-information states should function as inputs to a system for thinking and reasoning. Thus, only a creature with concepts can enjoy perceptual experiences. Nevertheless, a perceiving, thinking, concept-applying creature need not possess all the concepts that would be required fully to specify the content of a perceptual experience and, in having the experience, need not employ even those concepts that are possessed. Evans allowed that the representational content of perceptual experience need not be conceptual content, and in subsequent work, the notion of nonconceptual content has played a major role in accounts of the representational content of perceptual experience (Crane 1992, Gunther 2003, Peacocke 2001).
Several of Evans's papers—beginning with "Identity and Predication" (1975/1985) and including "Semantic Structure and Logical Form" (1976/1985) and "Does Tense Logic Rest upon a Mistake?" (1985)—contributed to the foundations of semantics and particularly to constraints on semantic theories that show how the meanings of whole sentences depend on the meanings of their parts. In "Semantic Theory and Tacit Knowledge" (1981/1985), he connected the requirement that a semantic theory should reveal semantic structure in sentences with the idea that speakers of a language have tacit knowledge of such a theory. Evans developed a substantive account of tacit knowledge (see also Davies 1987, Peacocke 1989) and distinguished the nonconceptualized content of tacit-knowledge states from the conceptualized content of belief states.
Evans's account of the semantic properties of descriptive names, put forward in "Reference and Contingency" (1979/1985), led to developments in two-dimensional modal logic (Davies and Humberstone 1980; see also Evans 2004), and he made further use of the notion of a singular term with its reference fixed by description in seminal work on pronouns. In "Pronouns, Quantifiers, and Relative Clauses" (1977/1985) and in "Pronouns" (1980/1985), Evans developed an influential account of the semantic function of pronouns that depend for their interpretation on an earlier quantifier phrase yet without being interpretable as bound variables (Neale 1990; King 2005). Finally, "Things without the Mind" (1980/1985) and "Molyneux's Question" (1985), along with the central chapters of The Varieties of Reference, have had a profound influence on subsequent work in philosophy of psychology, particularly work concerningthe perception and representation of space and, more generally, the conditions for an objective conception of a spatial world (Eilan, McCarthy, and Brewer 1993).
See also Davidson, Donald; Dummett, Michael Anthony Eardley; Frege, Gottlob; Kripke, Saul; McDowell, John; Philosophy of Mind; Proper Names and Descriptions; Reference; Russell, Bertrand Arthur William; Semantics; Strawson, Peter Frederick.
works by evans
With John McDowell, eds. Truth and Meaning: Essays in Semantics. Oxford, U.K.: Oxford University Press, 1976.
The Varieties of Reference. Oxford, U.K.: Oxford University Press, 1982.
Collected Papers. Oxford, U.K.: Oxford University Press, 1985.
"Comment on 'Two Notions of Necessity.' " Philosophical Studies 118 (2004): 11–16.
Burge, Tyler. "Belief de re." Journal of Philosophy 74 (1977): 338–362.
Crane, Tim, ed. The Contents of Experience: Essays on Perception. Cambridge, U.K.: Cambridge University Press, 1992.
Davies, Martin. "Tacit Knowledge and Semantic Theory: Can a Five Per Cent Difference Matter?" Mind 96 (1987): 441–462.
Davies, Martin, and I. L. Humberstone. "Two Notions of Necessity." Philosophical Studies 38 (1980): 1–30.
Eilan, Naomi, Rosaleen McCarthy, and Bill Brewer, eds. Spatial Representation: Problems in Philosophy and Psychology. Oxford, U.K.: Blackwell Publishers, 1993. 2nd ed., Oxford, U.K.: Oxford University Press, 1999.
Gunther, York H., ed. Essays on Nonconceptual Content. Cambridge, MA: MIT Press, 2003.
King, Jeffrey C. "Anaphora." In The Stanford Encyclopedia of Philosophy, fall 2005 ed., edited by Edward N. Zalta. Available from http://plato.stanford.edu/entries/anaphora/.
Kripke, Saul. "Naming and Necessity." In Semantics of Natural Languages, edited by Donald Davidson and Gilbert Harman, 253–355. Dordrecht, Netherlands: D. Reidel, 1972.
Neale, Stephen. Descriptions. Cambridge, MA: MIT Press, 1990.
Peacocke, Christopher. "Does Perception Have a Nonconceptual Content?" Journal of Philosophy 98 (2001): 239–264.
Peacocke, Christopher. "When Is a Grammar Psychologically Real?" In Reflections on Chomsky, edited by Alexander George, 111–130. Oxford, U.K.: Basil Blackwell, 1989.
Sainsbury, R. M. Critical notice: The Varieties of Reference, by Gareth Evans. Mind 94 (1985): 120–142.
Martin Davies (2005)