This philosophical treatise on the foundations of semantics is a systematic effort to clarify, deepen, and defend the classical doctrine that words are conventional signs of mental states, principally thoughts and ideas, and that meaning consists in their expression. This expression theory of meaning is developed by carrying out the Gricean program, explaining what it is for words to have meaning in terms of speaker meaning, and what it is for a speaker to mean something in terms of intention. But (...) Grice's own formulations are rejected and alternatives developed. The foundations of the expression theory are explored at length, and the author develops the theory of thought as a fundamental cognitive phenomenon distinct from belief and desire, argues for the thesis that thoughts have parts, and identifies ideas or concepts with parts of thoughts. This book will appeal to students and professionals interested in the philosophy of language. (shrink)
There is abundant evidence of contextual variation in the use of “S knows p.” Contextualist theories explain this variation in terms of semantic hypotheses that refer to standards of justification determined by “practical” features of either the subject’s context (Hawthorne & Stanley) or the ascriber’s context (Lewis, Cohen, & DeRose). There is extensive linguistic counterevidence to both forms. I maintain that the contextual variation of knowledge claims is better explained by common pragmatic factors. I show here that one is variable (...) strictness. “S knows p” is commonly used loosely to implicate “S is close enough to knowing p for contextually indicated purposes.” A pragmatic account may use a range of semantics, even contextualist. I use an invariant semantics on which knowledge requires complete justification. This combination meets the Moorean constraint as well as any linguistic theory should, and meets the intuition constraint much better than contextualism. There is no need for ad hoc error theories. The variation in conditions of assertability and practical rationality is better explained by variably strict constraints. It will follow that “S knows p” is used loosely to implicate that the condition for asserting “p” and using it in practical reasoning are satisfied. (shrink)
H. P. Grice virtually discovered the phenomenon of implicature (to denote the implications of an utterance that are not strictly implied by its content). Gricean theory claims that conversational implicatures can be explained and predicted using general psycho-social principles. This theory has established itself as one of the orthodoxes in the philosophy of language. Wayne Davis argues controversially that Gricean theory does not work. He shows that any principle-based theory understates both the intentionality of what a speaker implicates and the (...) conventionality of what a sentence implicates. In developing his argument the author explains that the psycho-social principles actually define the social function of implicature conventions, which contribute to the satisfaction of those principles. This challenging book will be of importance to philosophers of language and linguists, especially those working in pragmatics and sociolinguistics. (shrink)
Wayne Davis presents a highly original approach to the foundations of semantics, showing how the so-called "expression" theory of meaning can handle names and other problematic cases of nondescriptive meaning. The fact that thoughts have parts ("ideas" or "concepts") is fundamental: Davis argues that like other unstructured words, names mean what they do because they are conventionally used to express atomic or basic ideas. In the process he shows that many pillars of contemporary philosophical semantics, from twin earth arguments to (...) the necessity of identity, are unfounded. (shrink)
David Lewis, Stewart Cohen, and Keith DeRose have proposed that sentences of the form S knows P are indexical, and therefore differ in truth value from one context to another.1 On their indexical contextualism, the truth value of S knows P is determined by whether S meets the epistemic standards of the speakers context. I will not be concerned with relational forms of contextualism, according to which the truth value of S knows P is determined by the standards of the (...) subject Ss context, regardless of the standards applying to the speaker making the knowledge claim. Relational contextualism is a form of normative relativism. Indexical contextualism is a semantic theory. When the subject is the speaker, as when S is the first person pronoun I, the two forms of contextualism coincide. But otherwise, they diverge. I critically examine the principal arguments for indexicalism, detail linguistic evidence against it, and suggest a pragmatic alternative. (shrink)
This philosophical treatise on the foundations of semantics is a systematic effort to clarify, deepen and defend the classical doctrine that words are conventional signs of mental states, principally thoughts and ideas, and that meaning consists in their expression. This expression theory of meaning is developed by carrying out the Gricean programme, explaining what it is for words to have meaning in terms of speaker meaning, and what it is for a speaker to mean something in terms of intention. But (...) Grice's own formulations are rejected and alternatives developed. The foundations of the expression theory are explored at length, and the author develops the theory of thought as a fundamental cognitive phenomenon distinct from belief and desire, argues for the thesis that thoughts have parts, and identifies ideas or concepts with parts of thoughts. This book will appeal to students and professionals interested in the philosophy of language. (shrink)
My goal is to define intending. I defend the view that believing and desiring something are necessary for intending it. They are not sufficient, however, for some things we both expect and want (e.g., the sun to rise tomorrow) are unintendable. Restricting the objects of intention to our own future actions is unwarranted and unhelpful. Rather, the belief involved in intending must be based on the desire in a certain way. En route, I argue that expected but unwanted consequences are (...) not intended, examine the two senses of "desire," distinguish intending from being willing, and relate intending to a variety of other propositional at? titudes. (shrink)
The causal theory of reasons holds that acting for a reason entails that the agents action was caused by his or her beliefs and desires. While Donald Davidson (1963) and others effectively silenced the first objections to the theory, a new round has emerged. The most important recent attack is presented by Jonathan Dancy in Practical Reality (2000) and subsequent work. This paper will defend the causal theory against Dancy and others, including Schueler (1995), Stoutland (1999, 2001), and Ginet (2002).Dancy (...) observes that our reasons are neither psychological states nor causes, and that our reasons can be both motivating and normative. I argue that these observations are fully compatible with the causal theory. According to the reductive version I develop for both cognitive and optative reasons, what it is for an action to be done for a reason is for certain beliefs and desires to cause the action in a particular way. Our reasons for action are the objects of some of those beliefs and desires. The causal process has two stages. This theory explains not only Dancys observations, but also many other facts about reasons that alternative theories leave unexplained. I argue against Schueler and others that the non-appetitive desires entailed by acting for reasons are no less distinct and independent causal factors than the beliefs entailed. I go on to rebut arguments that the relation between psychological states and actions cannot be causal because it is non-empirical, rational, normative, or non-deterministic, and that explanations in terms of psychological causes are incompatible with explanations in terms of reasons. (shrink)
The use of ‘S knows p’ varies from context to context. The contextualist theories of Cohen, Lewis, and DeRose explain this variation in terms of semantic hypotheses: ‘S knows p’ is indexical in meaning, referring to features of the ascriber’s context like salience, interests, and stakes. The linguistic evidence against contextualism is extensive. I maintain that the contextual variation of knowledge claims results from pragmatic factors. One is variable strictness :395–438, 2007). In addition to its strict use, ‘S knows p’ (...) may be used loosely to implicate that S is close enough to knowing p for contextually indicated purposes. Here I explore another variable: belief about what is known. This factor is pragmatic rather than semantic in that it affects the use of ‘S knows p’ without affecting its truth conditions. While variation in strictness accounts for the variation in the bank, parking, and some lottery cases, variation in belief accounts for the variation in other lottery cases and the epistemology cases. Along the way, I sketch an insensitive invariantist semantics that is strict but non-skeptical, and show how it works with these pragmatic factors. (shrink)
Contextualist theories of knowledge offer a semantic hypothesis to explain the observed contextual variation in what people say they know, and the difficulty people have resolving skeptical paradoxes. Subject or speaker relative versions make the truth conditions of “S knows that p” depend on the standards of either the knower’s context (Hawthorne and Stanley) or those of the speaker’s context (Cohen and DeRose). Speaker contextualism avoids objections to subject contextualism, but is implausible in light of evidence that “know” does not (...) behave like an indexical. I deepen and extend these criticisms in light of recent defenses by contextualists (including Ludlow). Another difficulty is that whether certain standards are salient or intended does not entail that they are the proper standards. A normative form of contextualism on which the truth of a knowledge claim depends on the proper standards for the context is more promising, but still unsatisfactory whether the view is speaker or subject relative. I defend alternative explanations for the observed linguistic and psychological data: a pragmatic account for some cases and a cognitive account for others. 1. (shrink)
Abstract MacFarlane distinguishes “context sensitivity” from “indexicality,” and argues that “nonindexical contextualism” has significant advantages over the standard indexical form. MacFarlane’s substantive thesis is that the extension of an expression may depend on an epistemic standard variable even though its content does not. Focusing on ‘knows,’ I will argue against the possibility of extension dependence without content dependence when factors such as meaning, time, and world are held constant, and show that MacFarlane’s nonindexical contextualism provides no advantages over indexical contextualism. (...) The discussion will shed light on the definition of indexicals as well as the meaning of ‘knows,’ and highlight important constraints on the way meaning can be represented in semantics. Content Type Journal Article Pages 1-14 DOI 10.1007/s11098-011-9831-1 Authors Wayne A. Davis, Georgetown University, Washington, DC, USA Journal Philosophical Studies Online ISSN 1573-0883 Print ISSN 0031-8116. (shrink)
I critically examine Cappelen and Lepore’s definition of and tests for indexicality, and refine them to improve their adequacy. Indexicals cannot be defined as expressions with different referents in different contexts unless linguistic meaning and circumstances of evaluation are held constant. I show that despite Cappelen and Lepore’s claim that there are only a handful of indexical expressions, their “basic set” includes a number of large and open classes, and generates an infinity of indexical phrases. And while the tests can (...) be used effectively to combat contextualism concerning ‘knows’ and ‘actual,’ many expressions not in their basic set test positive for indexicality, including quantifier nouns, weather reports, and comparative adjectives. I rebut their claim that context-shifting arguments inevitably lead to radical contextualism, and that if there were any indexicals beyond their basic set, communication would be impossible. (shrink)
Grice’s Razor is a methodological principle that many philosophers and linguists have used to help justify pragmatic explanations of linguistic phenomena over semantic explanations. A number of authors in the debate over contextualism argue that an invariant semantics together with Grice’s (1975) conversational principles can account for the contextual variability of knowledge claims. I show here that the defense of Grice’s Razor found in these “Gricean invariantists,” and its use against epistemic contextualism, display all the problems pointed out earlier in (...) Davis (1998). The everyday variation in acceptable knowledge claims is better explained in terms of implicature than indexicality, but general conversational principles shed little light on whether ‘know’ is used hyperbolically, meiotically, or loosely in a context, although this issue is crucial in deciding what if anything ‘S knows p’ implicates. I present reasons favoring an account of the representative bank case in terms of loose use, making clear how they differ from Grice’s Razor. (shrink)
In part 4 of Meaning, Expression, and Thought, Davis rejects what he calls Fregean ideational theories, according to which the meaning of an expression is an idea; and then presents his own account, which states that, e.g., the meaning of 'Primzahl' in German is the property of meaning prime number. Before casting doubt on the latter ontology of meanings, I come to Frege's defence by pointing out that he was not an advocate of the position Davis named after him because (...) Fregean senses are not lexical meanings and Fregean thoughts are not types of mental events. (shrink)
Berg seeks to defend the theory that the meaning of a proper name in a belief report is its reference against Frege’s puzzle by hypothesizing that when substituting coreferential names in belief reports results in reports that seem to have different truth values, the appearance is due to the fact that the reports have different metalinguistic implicatures. I review evidence that implicatures cannot be calculated in the way Grice or Berg imagine, and give reasons to believe that belief reports do (...) not have the implicatures Berg attributes to them. I also argue that even if belief reports did have such implicatures, they would not explain why the belief reports in Frege’s puzzle seem to have different truth values. I point out that Berg has no reason to believe that Lois Lane believes Clark Kent is a reporter and Lois Lane believes Superman is a reporter are both true rather than both false, and that Leibniz’s Law cannot be used to defend substitutivity in belief reports because belief reports are not relational in the requisite way. Finally, I observe that some of the linguistic data Berg uses to argue for substitutivity in belief reports concerns the transparent interpretation of belief reports, whereas Frege’s puzzle concerns the opaque interpretation. (shrink)
It has often been said that 'desire' is ambiguous. I do not believe the case for this has been made thoroughly enough, however. The claim typically occurs in the course of defending controversial philosophical theses, such as that intention entails desire, where it tends to look ad hoc. There is need, therefore, for a thorough and single-minded exploration of the ambiguity. I believe the results will be more profound than might be suspected.
Christopher Peacocke has presented an original version of the perennial philosophical thesis that we can gain substantive metaphysical and epistemological insight from an analysis of our concepts. Peacocke's innovation is to look at how concepts are individuated by their possession conditions, which he believes can be specified in terms of conditions in which certain propositions containing those concepts are accepted. The ability to provide such insight is one of Peacocke's major arguments for his theory of concepts. I will critically examine (...) this "fruitfulness" argument by looking at one philosophical problem Peacocke uses his theory to solve and treats in depth. Peacocke (1999, 2001) defines what he calls the "Integration Challenge." The challenge is to integrate our metaphysics with our epistemology by showing that they are mutually acceptable. Peacocke's key conclusion is that the Integration Challenge can be met for "epistemically individuated concepts." A good theory of content, he believes, will close the apparent gap between an account of truth for any given subject matter and an overall account of knowledge. I shall argue that there are no epistemically individuated concepts, and shall critically analyze Peacocke's arguments for their existence. I will suggest more generally that the possession conditions of concepts and their principles of individuation shed little light on the epistemology or metaphysics of things other than concepts. My broader goal is to shed light on what concepts are by showing that they are more fundamental than the sorts of cognitive and epistemic factors a leading theory uses to define them. (shrink)
The idea that english has more than one declarative "mood" has been dismissed as superstitious by empirically-minded grammarians of english for centuries--with such spectacular unsuccess, however, that the indicative/subjunctive dichotomy stands today as a cornerstone for philosophical and logical speculation about "conditionals." let me be next into the breach. i shall urge that there is no grammatical basis for any such distinction. and as for the particular adjudications of mood logicians and philosophers actually propose, there is neither rhyme nor reason (...) to them. my bent, then, is basically destructive. but i shall also be outlining a better taxonomy. (shrink)
Lowe claims that "if oswald did not kill kennedy, someone else did" is a material conditional. he also claims that the difference in truth-value between this indicative conditional and the subjunctive "if oswald had not killed kennedy, someone else would have" does not support the conclusion of lewis and others that corresponding indicative and subjunctive conditionals are not always equivalent. i dispute both claims.
I shall conclude with a methodological moral. I have tried to show that there are several fundamentally different kinds of fear. One is a pure propositional attitude, one is partially a bodily state, and one is a relation between a person and a nonpropositional object. Other emotions come in similar varieties, such as hope and happiness, but with significant differences. The state of happiness, for example, does not entail any particular bodily state or feeling. So one lesson is this: it (...) is hard to generalize about “the emotions.” Further detailed, analytical studies of particular emotions are needed before a “general theory of the emotions” can be fruitfully attempted. (shrink)
According to Peacocke, concepts are individuated by their possession conditions, which are specified in terms of conditions in which certain propositions containing those concepts are believed. In support, Peacocke tries to explain what it is for a thought to have a structure and what it is for a belief to have a propositional content. I show that the possession condition theory cannot answer such fundamental questions. Peacocke’s theory founders because concepts are metaphysically fundamental. They individuate the propositions and thoughts containing (...) them, which in turn individuate the propositional attitudes that are relations to those propositions or thoughts. (shrink)
We have argued that Lehrer's definitions of coherence and justification have serious technical defects. As a result, the definition of justification is both too weak and too strong. We have suggested solutions for some of the problems, but others seem irremediable. We would also argue more generally that if coherence is anything like what Lehrer's theory says it is, then coherence is neither necessary nor sufficient for justification. While our current objections are directed at the ‘letter’ of Lehrer's theory, other (...) criticisms can be aimed at its very ‘spirit’. We would argue that coherence is unnecessary for justification because of the existence of ‘basic beliefs’, those about self-presenting states (‘I have a tingling sensation in my leg’) or self-evident truths (‘All men are men’). Such beliefs may be justified even though there are no other propositions in the subject's acceptance system that makes them more probable than competitors. Coherence is, moreover, insufficient for justification, because it ignores the inferential structure of the subject's acceptance system, and requires no justification of any kind for the subject's acceptance system itself. But we must develop these more fundamental objections on another occasion. (shrink)
Robert Cummins has argued that Jerry Fodor’s well-known systematicity argument begs the question. I show that the systematicity argument for thought structure does not beg the question, nor run in either explanatory nor inferential circles, nor illegitimately project sentence structure onto thoughts. Because the evidence does not presuppose that thought has structure, connectionist explanations of the same interconnections between thoughts are at least possibilities. Butthey are likely to be ad hoc.
The different occurrences of a word in a sentence cannot be identified with the one word type, nor with its many tokens. What then are occurrences of a word? How can one type occur more than once in another type? Is the conception of ?structural universals? that leads to these questions incoherent, as Lewis maintained? I argue against the answer Wetzel suggested, which identifies sentences with functions from numbers to expressions, and propose instead that occurrences of one type in another (...) are subtypes with different relative positions in the whole. Types have subtypes as well as tokens. Occurrences in non-symmetrical types like sequences are individuated by their relative positions in the type. Occurrences in a symmetrical whole have different positions but stand in the same relations to the other components. They are numerically different subtypes that do not differ from each other qualitatively. (shrink)