ABSTRACTS |
|
We are proposing a semantic theory that is quasi-Quinian. We argree with Quine that there is no sharp distinction between synthetic and analytic statements, and we conclude that semantic relationships can be indeterminate. We agree with Quine that word meanings are systematic and not atomistic, but we conclude that the appropriate system is a semantic field specified by a subject matter. We agree that responses of subjects are the appropriate evidence base, but we distinguish between responses germane to determining intra-linguistic relationships, which we call sense, and those germane to applications of a word, which we call reference. A theory of meaning is, we claim, a theory of sense and reference.
An adequate semantic theory of reference and sense must explain how reference is determined by a variety of factors, which includes sense as a central determinant. Such a theory must account for widely recognized features of vagueness and indeterminacy. It must explicate the role of pragmatic determinants. It must accommodate the fact of semantic change. It must include an analysis of the influence of experts and the limits of their influence on reference and sense. It must contain as a salient feature an account of the connection between sense, the network relationships between words, and reference, the relationship between the word and the world. Finally, the theory should clarify how sense and reference combine to yield the interpretation of a word in idiolects and languages. We shall present a semantic theory which combines these various factors, a theory of the aggregation of vectors, to account for reference and sense.
We propose a sketch of such a theory, first showing why theories of both sense and reference are necessary, providing examples of indeterminacy and variation, but concluding with a vector model that presents a plausible model of utterance interpretation and meaning assignment. What is most impressive about our language ability is that we can use and exploit the dynamics of meaning by negotiation and decision. We can stipulate a precise technical definition whenever we consider this necessary, a process that is always dynamic; we can talk about our disagreements in sense and reference and either adopt one usage or understand the usage of the others in a conversation. But successful negotiation is only possible if there is stability elsewhere in the system, since we use other words to define the words in question. We may rebuild the ship of language at sea, but we need a stable working place within the ship for our constructive efforts.
Questions for Quine:
1. Why ignore speakers' reports of semantic relationships within a language? Reports are behavior, too, and as observable as the application of words.
2. Why limit the study of semantics to the way in which people apply words? Reference underdetermines sense and is often less stable than intralinguistic relationships.
3. If we reject the atomicity of word meaning, why do we have to accept holism? A semantic field delimited by a subject matter can be the appropriate unit of inquiry.
University of Arizona, Tucson, USA
ABSTRACTS*QUESTIONES FROM QUINE/KARLOVY VARY '95 |
41 |
In "Two Dogmas of Empiricism", Quine characterizes and rejects three approaches to making sense of analyticity. One approach attempts to reduceputative analytic statements to logical truths by synonym substitution; thus, supposing "unmarried men" is synonymous with "bachelor" "All bachelors are unmarried men" reduces to "All bachelors are bachelors." A second approach isto identity analytic statements with "semantic rules," or "meaning postulates." A third approach relies on the verificationist theory of meaning. According to that theory, "every meaningful statement is held to be translatable into a statement (true or false) about immediate experience" or, less radically, "each statement, taken in isolation from its fellows, can admit of confirmation or infirmation". Analytic statements are, then, those that are confirmed come what experiences may. If either version of the verificationist approach were correct, then there would be objective facts about the extensions of terms from intuitive semantics, for example, "'...' is synonymous with '___'" and "'...' is analytic," across all languages. In short, such metalinguistic terms would be transcendent. In this essay, I focus my discussion primarily on the third, verificationist, approach - though what I shall have to say will bear on the second approach as well. In particular, I explain the models of language contained in verificationism and the weaker confirmationism, and why most positivists abandoned the former for the latter. I explain the connection between confirmationism and intuitive semantics and why considerations of holism alone are incapable of severing that connection. More specifically, I argue that the acceptance of holism by itself does not undercut the confirmationist account of analyticity: a sentence is analytic just in case it is confirmed come what experiences may. Something more is required if that connection is to be severed. That "something more," implicit in , is what I call "the transcendence requirement," and I argue that neither verificationism nor confirmationism can meet that requirement. Consequently, attempts to ground analyticity on considerations of verificationism or confirmationism are forlorn.
Ruttgers University, New Brunswick, USA
Undeniably, Quine's arguments against de re modality and quantified modal logic set the stage for contemporary discussions of modality. Yet, discussions of Quine's master argument have tended to focus on a single point, because the argument's key critics have attacked this one part of the argument. However, the focus on this first stage of Quine's argument is quite unfortunate, because it has drawn attention away from the central insight of that argument. The more unfortunate, I shall argue, because key criticisms such as that due to R. Smullyan (and recently championed by S. Neale) are not sufficient to undermine Quine's argument. I will offer a reconstruction of the argument which is valid, independent of the theory of descriptions, and not restricted to the conception of necessity as analytic truth. Moreover, the reconstructed argument only uses assumptions of the undeniable and highly desirable sorts. I think that Quine's argument puts definite constraints on how we deploy the mechanism of quantified modal logic, and puts legitimate pressure on the notion of de
ABSTRACTS*QUESTIONES FROM QUINE/KARLOVY VARY '95 |
42 |
re modality. Some of the antecedently desirable assumptions of the argument must be surrendered in order to avoid a skeptical conclusion. Yet, this can be done in a well-motivated way, and Quine himself has done so. I will suggest which of these assumptions we should want to lose, and sketch a very Quinean account of genuinely de re modality.
University of Florida, Gainesville, USA
The purpose of this paper is to explore the idea that there is genuine indeterminacy of identity that is due to how the world is, not just to a deficiency in the conceptual apparatus that we use to describe it. Leibniz's account of identity in terms of coincidence of actual properties and relations is adopted; then indeterminacy in possession of properties or standing in relations may (though it need not) result in indeterminacy in identities. This, we argue, is consistent with Quine's views on identity, including the "no entity without identity" principle. Further, quantification and counting are not deprived of meaning.
Indeterminate identity can be bewildering, so we propose a method of picturing situations containing indeterminacy, including indeterminacy of identity. In graphic form, the pictures resemble Venn diagrams, except that objects are pictured by small areas instead of points. An object image that lies across the boundary of a circle representing a property extension illustrates a situation in which it is indeterminate whether the object has the property. Object images that partially overlap illustrate that it is indeterminate whether the objects pictured are identical. A simple set theoretic characterization is given of the imagery; a picture of a sitation in the given terms automatically establishes the consistency of statements describing the situation.
Some writers suggest that the appearance of indeterminacy of identity is appearance only, and that any such indeterminacy can be resolved by refining the concepts that we use to characterize the world. We offer a brief critique of one version of this doctrine. We then relate this to Quine's suggestions in "What Price Bivalence?"
University of California, Irvin, USA
Quine has long been interested in set theory and in the problem of vagueness or indeterminacy. In this essay we consider the connection between these two issues. We set in "Worldly Indeterminacy of Identity" forth a theory of indeterminate identity that attempts to locate that indeterminacy in the objects themselves rather than in some failure of linguistic meaning. At the same time, the theory allows for lack of truth value due to indeterminacy of predication. When we consider the extension of the theory to higher-order objects such as sets or properties, certain natural questions arise. Foremost among these questions is whether indeterminacy of a predicate qua predicate (that is, indeterminacy arising from its predication of some object, as in "x is pink") is a necessary and/or sufficient condition for its indeterminacy qua object (in the sense of our theory, as in "Pink is a
ABSTRACTS*QUESTIONES FROM QUINE/KARLOVY VARY '95 |
43 |
color"), an issue closely related to Quine's insistence on the difference between occurrences of set-designating expressions on the left as opposed to the right side of the membership sign. We will set forth and defend a particular answer to this question in the context of a comprehensive account of indeterminate identity of sets.
University of California, Irvine, USA
The propositional attitude idioms have long been a source of trouble in semantics in virtue of their creating opaque contexts following the attitude verbs. As Quine has noted in 'Word & Object', and elsewhere, this feature of the attitude idioms provides one of the principal reasons for introducing propositions, entities to which the attitudes relate us, and to which that-clauses can be construed to refer. If admitting propositions were the only way to provide a semantics for attitude verbs, we would be faced with the prospect of admitting them, with their obscure conditions of individuation, into our ontology, or relegating the attitudes to at best a second-class status in our world-view. One of the most promising suggestions that has been made about how to bring intensional idioms into the fold of an extensional semantics, without appeal to propositions or meanings as entities, is contained in Donald Davidson's paratactic analysis of indirect discourse. Davidson's proposal, while generally acknowledged to be of striking originality and of great interest, has attracted a considerable amount of critical discussion. While many have been attracted to the proposal, it is safe to say that there is no consensus that it is correct, or that the device that Davidson introduces can be extended to handle other apparently intensional contexts, without appeal to propositions as entities-something which would undermine much of the appeal of the proposal. We undertake to show that the major criticisms that have been advanced against Davidson's proposal can be met while preserving its central insight, and that it can be extended in a straightforward way to attitude sentences and other apparently opaque contexts, without appeal to meanings, properties, relations, or propositions as entities which must be quantified over in our theory of truth.
University of Florida, Gainesville, USA
In an earlier paper ("Quine's Truth", Inquiry 37, 4, December 1994), I have argued that certain passages in Quine's work indicate that he should accept what I call an empiricist conception of truth. According to such a conception, a sentence is true, roughly speaking, if it is entailed by an empirically adequate theory. In his response (Inquiry, same issue), Quine is unwilling to accept my proposal, and he indicates certain problems, to which it gives rise. In this paper, I discuss Quine's response, and I consider some further problems connected with the empiricist conception of truth.
I argue that an observation sentence S is true on occasion O if every competent speaker of the language to which S belongs would assent to S if he or she were witnessing O. Moreover, an observation categorical of the form 'Whenever X, Y' is true if, and only if, for every occasion O, if X
ABSTRACTS*QUESTIONES FROM QUINE/KARLOVY VARY '95 |
44 |
is true on O, then Y is true on O. Finally, a theoretical sentence S is true if S is entailed by some tight theory which entails every true observation categorical and no false observation categorical. Underdetermination presents no problem here; different total theories are never incompatible, since meaning is immanent to theory. The empiricist conception of truth is compatible with truth as correspondence and with the existence of unknowable truths. Also, it is compatible with the existence of mistakes concerning the truth-values of observation sentences.
Stockholm University, Sweden
In modal metaphysics, there are two opposing positions, known as actualism and possibilism. The dispute between them is usually formulated in terms of existence: actualism holds that the only concrete entities that exist are one spatiotemporally connected to us here now, whereas possibilism holds that other concrete entities exist as well. So formulated in terms of existence, the dispute is bogus. Quine's famous slogan "To be is to be a value of a variable" helps us understand why. Quine's rhetorical questions concerning the possible fat man and the possible bald man in the doorway, on the other hand, help us see what the right formulation of the dispute should be. Once the dispute is correctly formulated, it is easy to see that Quine's other slogan "No entity without identity" succinctly encapsulates the case against possibilism. Unlike Quine, however, I deny that the dispute should be decided in favor of actualism. I deny that merely possible entities are "well-nigh incorrigible". I also deny that perception gives us evidence for existence in a sense which warrants skepticism about merely possible entitites.
California State University, Northridge, USA
There is a deep disagreement between Quine and Davidson about invoking sensory input or data in the theory of knowledge, mind and language. Quine, taking a firm naturalistic stance, treats these data as triggering of our sensory receptors leading to our assent of observation sentences, and, in the long run, to comprehensive scientific theories of the external world. In other words, it is private stimulus meaning, resulting from the stimulation of our 'surfaces', which in the first instance gives content to our thoughts. The public linguistic meaning, however indeterminate, must rest ultimately on stimulus meaning. Davidson believes that such an internalist theory, according to which what really matters to meaning or constitutes the contents of our thought occurs within the skin of an individual, is untenable: it makes truth relative to individuals and leads to skepticism. He proposes instead an externalist theory insisting that the role played by the individual is only that of possessing the thoughts, not that of determining their contents. The thoughts we form and entertain are 'located' in the world we inhabit. In other words, meaning is determined all the way down by public objects from our environment, with which we causally interact and of which we speak in our shared language.
ABSTRACTS*QUESTIONES FROM QUINE/KARLOVY VARY '95 |
45 |
The paper compares these two opposing views and tries to show why we should stick to some Quinean internalist intuitions in the full account of meaning and mental content. At the same time it acknowledges that Davidson pointed out an important difficulty which puts into serious doubt the philosophical significance of the category of naturalized sensory input. Finally it considers whether substituting this category with the notion of non-conceptual perceptual content bears the promise of resolving Davidson's difficulty.
Catholic University of Lublin, Poland
Saying that sentences are translations of each other because they have the same meaning presupposes what should be proved. So meaning is not appropriate for philosophical explanation. Quine introduces the stimulus meaning in order to escape pitfalls of meaning. Sentences 'Rabbit' and 'Gavagai' belong together because they have the same stimulus meaning. Stimulus meaning may be called analytical in that it directly experientally supports a finite number of sentences. Stimulus meaning for a given speaker is either affirmative or negative, it is a class of all stimulations which would prompt speaker's reactions of assent or dissent to a sentence. Here are some problems with the stimulus meaning:
(1) Proximality and distality. Is stimulus proximal, thus happening at the receiving surface of the organism, or behind the surface? Or is it distal, taking place in the external world of objects? Because of the behaviorist disregard for the internal, stimulus can not be inner to the organism. The distality of stimulus, such as proposed by Davidson, is also rejected. Quine opts for an intermediate position, locating stimuli at the organism's surface.
(2) Privacy and community. Stimulus meaning shows its effect in the intersubjective space of the translation situations, and so it seems to be shared by many organisms. If stimulus meaning is determined as similar for various organisms, this would bring us to the old problems with meaning. Stimulus meaning is thus determined as belonging to a single organism.
(3) The structure of the stimulus. Stimulus is physical and thus close to reception. On the other hand, stimulus is adequate for encoding of the sentential information. In order to serve as a basis for response to the situations described by sentences, it has to have an adequate structure. The structure of the stimulus is not only pure reception. Structure comes to a stimulus because of its involvement in a wider setting. It is argued that the structure of the stimulus may be explained by categorization.
Categorization is not necessarily internal. It can be accounted for as fitting to the organism's receptor surfaces. This is indeed an usual way to describe categorization as a behaviorally accountable mechanism.
Categorization is a story told about the sensation by the science. It is happening at the sensory and at more complex levels. Meaning comes at this second level, adequate to the sentential description. But higher categorization would not be possible without reception level protocategorization.
Consequences of this are given for the issues of proximality/distality and privacy/community.
University of Ljubljana, Slovenija, matjaz.potrc@uni-lj.si