While Chomsky was attending to natural language syntax, philosphers continued to pursue what was a essentially a different way of getting at the structure of language: the authority of Tarski manifest in the semantics of such diverse thinkers as Donald Davidson and Richard Montague became even more firmly entrenched. While Chomsky pursued matters of case, anaphora, and transformation ( " Move alpha " ), the philosophically inclined linguists and the linguistically inclined philosphers remained by and large capitivated by the model theoretic applications of Tarki's work initiated, attempting to explain difficulties tied to more traditional problems, such as intentionality and indirect speech. In short, for Chomsky there was and is a syntactic animus to virtually every linguistic endeavor. For the philsophers, there is an equally irrepressible semantic motivation to what they have been doing during their more productive moments. Aside from developments in the theory of reference (such as Kaplan's " Demonstratives " ) much of what has been most productive in this tradition belongs to those who have sought to apply formal techniques to metaphysical problems having to do with events (here Higgenbotham, Dowty and Parsons leap to mind) and issues surrounding the treatment of " aspects " and tense (for a good overview here try Time and the Verb A Guide to Tense and Aspect by Robert I. Binneck.)
My second objective is to argue convincingly that having naturalized language we may proceed to the conclusion that in language as in the physical sciences there are " facts of the matter." In linguistics the determination of what the facts of the matter are in the philosophically relevant sense of " fact, " is based on the intentions of the speaker. Such a possibility is likely to suggest a metaphysics of intentionality reminiscent perhaps of Brentano ( Psychology from an Empirical Standpoint ). But there is a way of remaining noncommital on this: narrow the domain of intentionality. The best way of doing this is to restrict the domain of intentional phenomena to those involving certain strategies. To have an intention is to have a strategy for implementing a desire. Basic actions (in the sense of Danto) are therefore not intentional. The desire here is either communication or the expression of thought; and here we will at some time in the future briefly digress into a discussion of Dummett's and Chomksy's views on the primary function of language.
My third objective is to support the way I accomodate my second objective by adducing evidence from syntactic considerations of binding phenomena.
A centerpiece of Chomsky's views on language is the innateness hypothesis. The innateness hypothesis will be argued to be without significance, philosophically speaking, except insofar as it reveals a telling clue as to the link between Chomsky and logical positivism's preoccupation at one point with " dispositional terms. "
In order to ferret out some of the important details it seems to me one has to return to Rudolph Carnap. I believe that on closer examination what appears in Chomsky to be an easy blow against behaviorism in his now classic review of Skinner's work was in fact laden with some of the weaknesses of Carnap's conception of the relationship between mind and our dispositions to behave. In effect, what I am suggesting is that Chomsky did not go far enough in his repudiation of behaviorism.
One of Carnap's most important ideas, one related to his work on the semantics of disposition terms, was that
The occurrent properties of psychology are logically analogous to the familiar dispositional concepts of physics ...more accurate insight into the micro-structure of the human body should enable us to replace dispositional concepts with actual properties.(originally published in 1936, also inAyer and here quoted from Essays on Other Minds p.96.)
The key notion here is that of " dispositional concept. " Now today the semantics of dispositional terms has received treatment as what is supported by a counterfactual, emerging from ealier debates on causation in Nagel, Pap, and Nelson Goodman. But here I am going to pretend that the actual semantics of dispositional terms is not an issue. Instead I am going to examine the role of dispositions in their own right, despite the allure of the standard semantic reductions. But before doing so, I would like the reader to appreciate the irony in the fact that dispositions for Carnap in the above quote were explicitely reducible to " actual properties, " whereas today one of the most popular accounts ( David Lewis Counterfactuals) appeals to other possible worlds. A profound distinction lies hidden here: what van Heijenoort called " Logic as Calculus and Logic as Language " . Hintikka has made use of what is underlying here to make some very general points concerning the ineffability thesis, etc. The reader is also encouraged to look at " Wittgenstein and Language as the Universal Medium " by Merrill B. Hintikka and Jaakko Hintikka.
The matter of the relationship between " intelligent " behavior and dispositional concepts persists. Most philosophers will no doubt already know that Quine believes that thinking is largely if not entirely a matter of " verbal dispositions " (vide. " Mind and Verbal Dispositions " .) But the degree to which " dispositional concepts " figure in Chomsky's theory of language is rarely regarded as significant, or so it would appear. I hope to show that this appearance to the contrary notwithstanding Chomsky's notion of cognitive structure is very closely related to what Carnap and others spoke of in terms of dispositional concepts.
...I am using the term 'language' to refer to an individual phenomenon, a system represented in the mind/brain of a particular individual.( Language and the Problem of Knowledge: The ManaguaLectures p.26.) Unviversal grammar for Chomsky has received a reasonably uniform characterization throughout his work. It pertains to the " ...set of properties, conditions, or whatever that constitute the " initial state " of the language learner " ( Rules and Representations p. 69). There are occasions, however, when Chomsky speaks of Universal Grammar more generally as an attempt to " ...formulate the principles that enter into the operation of the language faculty. " ( Language and the Problem of Knowledge p. 61). Still these notions are not really that far apart. In order to see why this is so, consider what Chomsky has to say concerning the relation of behavior and language.
...if we are interested in the problem of 'causation of behavior' as a problem of science, we should at least analyze the relation of experience to behavior into two parts: first LT, which relates experience to cognitive state, and second a mechanism, Mcs, which relates stimulus conditions to behavior, given the cognitive state CS.
( Reflections on Language p. 16.)
Perhaps the most general feature of Chomsky's approach to language is his belief in linguistics as a part of psychology ( Rules and Representations . p.4). Please refer to thevabove quotation from Carnap.There is the suggestion, taken together with Chomsky's belief here, that language might be reducible to certain states or properties of individuals, just as Carnap argued that solubility could be described in terms of physical states of individual substances. A brief review of Carnap's influence on cognitive psychology prior to Chomsky will prove worthwhile.
Carnap insisted on an operationalist approach to the introduction of theoretical terms such as solubility. The logical characterization of introducing such terms within an experimental context was illustrated using the dispositional concept of solubility as a paradigm. The standard formulation is exemplified by 1.
1. If x is placed in water, then (if x is soluble then x will dissolve)
Notice that we have three things here: first, a set of initial experimental conditions (of being placed in water), second a theoretical notion (being soluble) and third, a behavior that ensues, viz. that of dissolving.
In the case of temperature as opposed to solubility we would have another set of experimental conditions associated with other behavior. What is important here is that solubility " intervenes " between observable conditions and observable results. In addition, since dispositional terms are associated with a number of experimental conditions such definitions as offered by Carnap were not explicit, but relied on sets of such reduction sentences as (1). The details, at this point are not crucial. What is important is the influence or Carnap's thinking on the matter of reduction and confirmation in psychology and indirectly at least, or so I maintain, on both Chomsky and the behaviorists. This influence has been widely recognized:
Legitimate instances of such " pure " intervening variables are Skinner's 'reserve', Tolman's 'demand', Hull's 'habit strength' and Lewin's 'valence'. These constructs are the behavioral analogue of Carnap's "disposition concepts" such as 'solubility', 'resistance', 'inflammability', etc.(Kenneth MacCorquodale and Paul Meehl p.33.)
When Chomsky speaks of " cognitive structure " what he means, I shall argue, is a second order disposition. There is evidence to suggest to the contrary; that is, there is evidence that while linking dispositional concepts and cognitive capacities Chomsky prefers to keep a safe distance. For example he remarks:
there has been a tendency to employ the notion 'disposition' or 'capacity' where the more abstract 'cognitive structure' is , I believe, more appropriate.( Reflections on Language p.23)
Nevertheless, talk of " innate capacities " sounds very much like disposition talk:
Secondly, I want to consider mind (in the narrower or broader sense) as an innate capacity to form cognitive structures, not first order capacities to act.
( Reflections on Language p.23)
While I still maintain the essential content of the main thesis of this article, I would today modify some special points. Perhaps the most important is the following.In the article I regarded a psychological term, say " excited, " as designating a state characterized by the disposition to react to certain stimuli with overt behavior of certain kinds...But at least for those of scientific psychology ... it seems to me more in line with the actual procedure of scientists, to introduce them not as dispositions concepts, but rather as theoretical terms ...( Other Minds p.109) .
This move is very similar to Chomsky's move from disposition to " cognitive structure. " But the problem is not obviated by this manuever: what has in fact taken place is that Carnap is suggesting only a new way of treating all theoretical terms and then subsuming dispositions under this rubric. His new method introduced " correspondence rules," a subject worthy of subsequent discussion, but here well outside our purview. Cognitive structure remains dispositional on the face of it, related as second order to first order dispositions: those first order dispositions being akin with Quine's speech dispositions.
Chomsky's development of the concept of a " rule " will provide a way of distinguishing psychological dispositions from cognitive structure. My point has been that Chomsky did not go far enough in rejecting the behaviorist dispositional approach. He provides the wherewithal for making the distinction explicit but to the best of my knowledge he never makes this explicit.
What I wish to do is focus on the concepts of rules and principles, developing a semantics based on game theoretical principles which will dispense with an arcane metaphysics inherited from classical positivism: a metaphysics that took us to new heights but which must be left behind as we come to view the concepts with which we think as even more intimately connected with natural language than many in the posititivist tradition believe.