Page 352 - Encyclopedia of Philosophy of Language

Formal Semantics
For instance, the common noun 'red' is associated with the given property red, and its extension is defined to be the class of all individuals to which this property applies. Property theories circumvent the problem of omniscience. To continue the above example, the sentence, 'Fermat's last theorem is true' refers to a proposition and so does the sentence, Two plus two is four.' But since propositions are primitive, these will be different objects, and hence the faulty inference is blocked. (There are, in fact, property the- ories which obtain propositions from other inten- sional objects. They block the inference, since the structure of the propositions involved will differ.) However, property theories face another problem, in that the intensional objects they use are often too fine- grained. For instance, it seems that the proposition expressed by 'John walks and talks' is exactly that of 'John walks and John talks,' but this is not always
guaranteed in such theories. Consequently, the quest is for objects which have a granularity intermediate between the fine non-compound abstract objects and the coarse sets of possible worlds.
Much in the above ' . . . offends the aesthetic sense of those who have a taste for desert landscapes' (Quine 1961:4). On the nominalistic view, of which Quine and Goodman are ardent defenders, explaining mean- ingful language in terms of abstract entities is explain- ing the obscure by the still more obscure. They strive to give an explanation in terms of particulars. The sole intensionality they would allow is that of the linguistic sign, which is identified with the het- erogeneous space-time region comprising all of its uses. This, too, is an extremely fine-grained notion of intensionality, for changing but one word in a sign results in a completely different particular.
3. Sentence Meanings
In translating a text, preservation of meaning or inten- sion is often stated to be the main aim. But what are the philosophical positions concerning the nature of meaning? These questions, evoked by the phenom- enon of sameness of meaning, are closely related to the problem of universals, evoked by the phenomenon of sameness of attributes, which has haunted phil- osophy. It is therefore no surprise that none of the theories of meaning proposed thus far is accepted by all workers in the field. Here, some of the main pos- itions are sketched using the familiar tripartition from the literature on universals; namely the realistic, the mentalistic, and the nominalistic position.
Frege (1892, 1918) held the realistic view that the intension of a sentence was the thought it expresses. These thoughts were taken to be free from the subjec- tive, poetic qualities which are attached to our every- day thoughts. Instead, Frege took them to be abstract entities existing independently of our human minds in a Platonistic realm different from either physical or
cognitive states and processes. In the Tractatus (1922), Wittgenstein, too, holds that the meaning of a simple sentence is an objective structure, but rather than Fre- gean thoughts these entities are identified as possible states of affairs in logical space. The world consists of the states of affairs which are realized, and hence a sentence would be true if its sense were to correspond to such actual states of affairs.
The semantics of Frege and the early Wittgenstein make use of notions which are far removed from actual linguistic practice. So these semantics leave open the important question of how to explain, in terms of these concepts, the ability to use language in order to make statements, queries, and the like.
According to a 'mentalistic' viewon meaning, inten- sions only exist within a mind. Meaningful language is now seen as a speaker's main tool to communicate the thoughts and pictures he senses. In order to do so, he somehowcodes these subjectiveexperiences in the form of perceptible signs, often written or spoken sentences. A hearer, in turn, decodes these signs result- ing in subjective experiences akin to those the speaker wished to express.
In this strongly simplified picture, there is an appar- ent danger of misunderstanding: how to find out whether the decoded message is the one intended? In order to bar the conclusion that misconstrual of intension is almost inevitable, one often appeals to universal principles governing the human mind, which would secure at least the possibility of successful trans- fer of information. What these principles are is a mat- ter for science to decide. Such 'innate knowledge' is also used to answer the question of how the signs come to have the meaning they in fact have. Some hold that there is stock of basic concepts, present at birth, perhaps in a mental language linked by con- vention to a spoken language, which gives the foun- dation for becoming a native speaker when exposed to a sufficiently rich linguistic environment. As will be seen shortly, the idea that meaningful language can be explained in this subjective mentalistic manner, is criticized by Wittgenstein (1953).
On the 'nominalistic' view, one strives to explain meaningful language in terms of particulars, rather than in terms of objective or subjective abstract enti- ties. As was said in the previous section, these par- ticulars induce a fine-grained notion of intensionality. In contrast, the abstract entities posited as meanings are often much coarser, in that they are said to remain constant across different languages. The feeling that such constancy occurs, prompted their use to begin with. This constancy is hard to explain on the nom- inalistic view, but Quine's argument of the inde- terminacy of translation purports to show that the assumed preservation of meaning is an illusion.
4. AnalyticversusHolisticTheoriesofMeaning
Until the 1990s sentences were taken to have meaning
330