Page 80 - Encyclopedia of Philosophy of Language

Language and Mind
of presentation in the two judgments. If not, the per- Fodor is not an eliminativist about intentionality; son is inconsistent. Also the mode of presentation the representational folk theory can be salvaged for employed will determine the thought's explanatory science provided the representations are suitably
role. For example, the indexical 'self mode of pres- entation and demonstrative modes of presentation of objects link in special ways to agency.
5.2.3 Real-world Referents
Thoughts about objects, kinds, and properties can be classified in a way that is sensitive to the entities that they represent. If the thought's identity is made to depend upon its worldly object, then its content is likewise world-dependent or 'externalist.' Evans (1982) defines a 'Russellian singular thought' to be one whose object is literally a constituent of it. Without the object, there would be no genuine thought. Similar accounts have been given of thoughts about natural kinds. Putnam (1975) showed that the meanings of natural-kind words are not fully encoded in the heads of speakers or hearers; the actual referent of 'water' helps to fix its meaning. The same could go for natural- kind concepts. Perhaps the actual referent of the con- cept water helps to individuate that concept. A differ- ent referent would make a different concept even if the two concepts seemed subjectively the same.
The distinctions already made between Fregean modes of presentation can be grafted onto the referent-sensitive typology to yield a set of fine- grained modes of presentation individuated jointly by cognitive role and by real-world referent.
6. Is Intentionality in the Head?
How one classifies a person's mental state depends on one's interests, and there seem to be plenty of ways to choose from. But can these crosscutting taxonomies really coexist peacefully? Is there not a 'right' one, or one that is 'best for scientific purposes'? A full treatment of the many arguments cannot be given here. A central issue concerns the location of contents: are they wholly in the head? It is useful to look first at singular thoughts, then at general thoughts.
It would be hard to deny that Evans-style singular thoughts, if they exist, are object-involving, and folk psychology certainly makes heavy use of indexical thoughts whose referents are contextually fixed. If such thoughts are individuated by their truth conditions, they are not wholly in the thinker's head. Yet two-factor theorists claim that such thoughts can be dissected into an in-the-head component, and a external component (see Loar 1981; McGinn 1982; Block 1986). The inner component is said to have 'narrow content.' Narrow contents cannot be specified by 'that' clauses, but they can be got at indirectly by subtracting the contribution made by the subject's context to the determination of the object of thought. Fodor, in Psychosemantics (1987), holds that the nar- row content of an indexical type is a function from contexts and mental episodes onto truth conditions.
trimmed. The trouble with singular thoughts, in his view, is that they crosscut the scientific classification. A content-based taxonomy should facilitate causal generalizations about types of thought, and this means abstractingfromthecontextsinwhichthoughttokens occur. The causal interactions between thoughts inside the head must hold in virtue of their narrow contents. This is the traditional view. Note, however, that if narrow contents were causally epiphenomenal, computational psychology would not need them any more than it would need wide, world-involving contents. All its causal generalizations would be over neural states typed physically or syntactically (Stich 1983).
Turning now to general thoughts which are not directed upon particular objects, if content- externalism says that contents are world-involving, can it possibly apply to these? McGinn (1989) dis- tinguishes two versions: weak externalism holds that a given content requires the mere existence of an object or property in the world at large, while strong exter- nalism insists that real relations (e.g., a causal relation) must be instantiated between the subject and worldly items. The strong version seems appropriate for, say, demonstrative thoughts; the subject must be in the context of the object at the time of thinking the demonstrative thought about it. But perhaps some thoughts require that there should have been causal interactions between the subject and things in the past, that is, that S's life-historical environment had to be a certain way. The intuition here is that S's concept water (say), exercised at time t, had to develop out of S's earlier experiences. If the formative experiences prior to t had not been interactions with genuine HH) samples, the concept exercised at t would not have been the concept water. And perhaps the same goes for concepts for simple qualities: past experience of genuine instances of quality Q is necessary in order for the current concept to count as concept Q. So externalism has some plausibility for general thoughts. The subject has indeed strayed a long way away from Brentano.
7. Normative Aspects of Intentionality
According to Peacocke (1986b), the nature of concepts and contents can be illuminated by investigating what it is for a person to possess concepts. His approach immediately brings normative considerations to the fore, since possessing a concept involves knowing how to employ it correctly.
Frege took concepts to be abstract entities, grasp- able by many minds but not dependent for their existence upon minds. Wittgenstein said that mastery of a concept was an ability to follow a rule. Both emphasized that, because a concept can be exercised
58