Quine and “Word and Object”
Ozge Ekin
December 9, 2003
In this paper I intend to give an overall assessment of Quine’s famous book: “Word and Object” (1960). My aim is first to reveal essential ideas in early chapters of the book, particularly Quine’s well-known “Indeterminacy of Translation” thesis and his arguments on learning language .I believe Quine’s suggestions on regimentation of language in Ch.5 are quite valuable and deserve an extensive review, however I will not give a detailed explanation since this topic itself deserves another paper. After setting out the rules of regimentation, intensions (propositions, properties, and relations) withstand the rules (Cannonical Language) and stand as a remainder of the language. Quine, however, manages to eliminate intensions in terms of cannonical language.
1.Physical Things
According to Quine we know physical things only through their actions on our bodies, and this is the qualification of empirical meaning of our talks about physical things. We do not question the reference (physical things). Hence we are free to reflect our own referential apparatus. Quine emphasizes this over and over, and argues that “…the points of condensation in the primordial conceptual scheme are things glimpsed, not glimpses.”1
Moreover however Quine is suspicious with mental items in any case, this does not matter since their explanatory value is empty for context of language.
2. Learning of Language
Quine believes that there are two ways of learning words: in isolation as a one word sentence and contextually or by abstraction. The words that learned in isolation are substantives, adjectives and verbs and learned contextually are prepositions, conjunctions and other words that are fragment of sentences learned as wholes. Some examples in terms of the way we learn them can be given as follows:
‘unicorn’ à by description/ contextual
‘sake’ à contextual
‘tile’ (refer to objects) à by description/ contextual/ in isolation
‘molecule’ (refer to objects) à contextual and by description
‘class’ and ‘attribute’ à contextual (however Quine at that point specifies his doubts about whether they refer to the object or not)
3.Indeterminacy of Translation
Quine seeks to express a related indeterminacy of thesis in behaviouristically respectable terms, hence in terms of use. He believes that we attribute meaning “holophrastically”, e.g. when we attribute the meaning of ‘rabbit’ as a term we need to understand ‘fur’, ‘white’, ‘jumps’, etc. This is our referential apparatus for ‘rabbit’, and everybody might have different referential apparatus, moreover every language i.e. every culture has different referential apparatus for the same referred object. Hence there is an indeterminacy of translation when we have “terms” to deal with. What Quine’s suggestion is to take ‘Rabbit’ as a one-word sentence. Hence referring to ‘Rabbit’ we present a scene for an assent or dissent when ensuing query ‘Rabbit?’ is at stake for a native we are not familiar which her language is. Therefore stimulus synonymy2 works in this case. When we deal with this issue in the same language the problem is reduced to terms that stimulus synonymous where learning of these terms differs. Hence ‘brother’ and ‘male sibling’ are stimulus synonymous and we learn ‘brother’ by verbal connections with sentences about childbirth, and ‘sibling’ by verbal connection with ‘brother’ and ‘sister’. However when it comes to the synonyms that we learn one’s meaning by associating the sample objects and the other by association of words with words we cannot talk about the stimulus synonymy for the very reason that their referential apparatus differ.
It is important to note here that indeterminacy of translation is not a problem; it is a thesis. Quine’s point of view is that “There is no fact of matter” e.g. translation is not determined empirically or by the totality of facts nor by the totality of truths about nature.
4. Referential Opacity
Quine defines purely referential position as when a sentence is true of the object and when a singular term is used in a sentence purely to specify its object, then sentence truth-value will not change when any other singular term is replaced referring to the same object. An example for purely referential position is the use of singular terms under predication. It is important to note that a term might be referential but not purely referential. It is obvious from the following example that referential position cannot be confused by purely referential position (as in the example given it the lecture):
(1) Giorgiovanni (G) was so called because of his size.
Giorgiovanni refers to a certain man. Now suppose Giorgiovanni = Jack
(2) Jack was so called because of his size.
In (1) we have non-referential position for (G).
We create opaque context when we use single quotes. Consider the following example:
(3) Giorgiovanni (G1) was called ‘Giorgiovanni’ (G2) because of his size.
(G1), in this example is purely referential and (G2) has opaque context.
Moreover, Quine uses opaque context for belief and argues that when belief is a binary relation between man (subject of the sentence) and propositional attitude then belief is an opaque context but when belief is multiple relation (as in Russell’s theory of multiple relation) then it is transparent.
5. Intensions and Regimentation
Quine gives the definition of regimentation as follows: “ At the height of the regimentation sentences were constructible only by adjoining general terms (including ‘=’ and ‘'’) predicatively to variables and applying quantification, truth functions, and other operations on sentences (thus ‘O’, ‘Ox‘ [a variable bounding binary predicate (refer to § 38)])” (p.203). In regimenting the language, Russell had eliminated definite descriptions, and Quine eliminated classes. Intensions defined as “object of belief” in his terms (propositions, properties/ attributes, relations) were hard to eliminate at first hand.
Quine takes propositions as objects of propositional attitudes and meaning of eternal sentences. Hence if we can talk about synonymy of eternal sentences then we can conclude that two propositions [p] and [q] are identical (‘p’ and ‘q’ would be eternal sentences meant by these propositions). He suggests that we use structural synonymy3 for this purpose and dispense with “propositions”. For “relation-in-intension”, Quine considers it as a binary relation where it stands as intension of a binary predicate. Still it is not easy to dispense with relations, and with attributes also. [I will not state clearly here how Quine eliminated these terms since I have not get a clear idea whether he really did it. It seems more like he never solved this problem.]
1.W.V. Quine, Word and Object, and New York: John Wiley and Sons, 1960: 1
2. Two sentences are stimulus synonymous if they have the same stimulus meaning where stimulus meaning is ordered pair of two classes of stimuli: ({the ones that prompt to assent} à positive stimulus meaning, {the ones that prompt to dissent} à negative stimulus meaning)
3. Let S and S’ be eternal sentences. S’ is structurally synonymous with S if S’ can be reduced to S by transformations of propositional logic and predicate logic + by interchanging stimulus synonymous terms (definition given in the class)