But many organisms also have a more sophisticated innate capacity. They can learn to generalize according to other types of similarities, provided they undergo appropriate experiences. Consider, for example, the ostensive learning of the word 'toy'. This learning process again includes a generalization. But this generalization is clearly not based on the similarity derived from our innate standards of similarity. Children do not innately generalize from one toy to other toys, e.g., from a doll to a toy-train. Rather, the generalization is based on a functional (or acquired) similarity, the similarity that holds between different toys. But this similarity is only valid for the persons who have undergone certain experiences, namely, the experiences in which they observed that some objects share the "toy-function". (For some of the empirical evidence that shows that certain organisms have the innate capacity to learn to generalize according to functional similarities, see, e.g., the experiments cited in Quine, 1974, Stemmer, 1973, 1983; see also Goldiamund, 1966.)
Empiricists now argue that such functional similarities also account for our syntactic competence. Consider first relational terms such as 'x is-bigger-than y' which children may have learned by an exposure to appropriate ostensive pairing events. But learning the meanings of such relational terms includes not only a generalization from one 'is-bigger-than' situation to "similar" situations, but also the classification of the values of the variables in appropriate semantic categories. Moreover, and this is crucial, by performing a so-called distributional analysis of the relevant linguistic context, the children can extract specific functional features from this context, such as the property of often receiving a plural ending, to be preceded by an article, or to occupy certain positions in 'is-bigger-than' sentences (see, e.g., Maratsos, 1982; Stemmer, 1973, pp. 69-73). These features then enable the children to also classify the values of the variables in syntactic categories, probably Noun-Phrase in the present case.
But children also learn the meanings of more specialized relational terms such as of 'x who y', as in 'The man who receives the book receives the prize'. (Cf. Chomsky's discussion of 'The man who is tall is in the room', 1975, p. 31.) Clearly, the meanings of such relational terms are extremely subtle and the learning processes are probably very complex. But since children do not innately know the meanings of these English terms, they must undergo the semantic experiences that enable them to learn these meanings. Now, knowing the meaning of 'x who y' includes having the capacity to perform a structural analysis of the relevant text on the basis of the semantic structure that is determined by the clause 'x who y'. Further, once the children classify the variables of the clause 'x who y' in the syntactic categories that are based on the (highly complex) functional features revealed by a distributional analysis of the context, they also acquire the capacity to perform a syntactic analysis of the text, an analysis that gives them the syntactic structure of the text. And this capacity then enables the children to perform syntactic operations on the text, for example, to transform the above sentence into the passive sentence 'The prize is received by the man who receives the book'. (For more details, see, e.g., Stemmer, 1987).
In this way then, empiricists account for the children's syntactic competence. It derives from their sophisticated learning capacities, especially from the capacity to learn to generalize according to functional similarities. With the help of these capacities, the children then process the experiences they undergo, and among the effects of these activities are the acquisition of a structure-dependent syntax.
Empiricists favor a bottom-up approach for language-acquisition research, and they stress the study of both: the concrete experiences that children undergo when learning their language and the learning mechanisms that enable children to process the data they collect in these experiences. Our conclusions suggest that, by acknowledging the psychological reality of the above-described learning mechanisms, empiricists can account not only for the semantic but also for the syntactic competence of children. To be sure, the conclusions don't prove this. But in any case, adopting an empiricist approach will indeed contribute to Barbiero's project, because it includes a detailed, and empirically supported, analysis of "the different constraints and mechanisms that can be attributed to the human biological endowment for language".
Nathan Stemmer
Beth-David Institute,
Email: nstemmer@netmedia.net.il
This essay continues the discussion that began in How should the Innatenes Hypothesis be reframed? and On Framing the Innateness Hypothesis?
Back up to Chomsky for Philosophers.
Chomsky, N. (1959). Review of Skinner's Verbal Behavior. Language, 35, 26-58.
Chomsky, N. (1965). Aspects of the theory of syntax. Cambridge, MA: MIT Press.
Chomsky, N. (1975). Reflections on language. New York: Pantheon Books
Goldiamond, I. (1966). Perception, language, and conceptualization rules. In B. Kleinmuntz (Ed.), Problem solving (pp. 183-224). New York: Wiley.
Maratsos, M. (1982). The child's construction of grammatical categories. In L. R. Gleitman & E. Wanner (Eds.), Language acquisition: The state of the art (pp. 240-266). New York: Cambridge University Press.
Quine, W.V. (1960). Word and object. Cambridge, MA: MIT Press.
Quine, W.V. (1974). The roots of reference. La Salle, IL: Open Court.
Stemmer, N. (1973). An empiricist theory of language acquisition. The Hague: Mouton.
Stemmer, N. (1983). The roots of knowledge. Oxford: Blackwell (New York: St. Martin's Press).
Stemmer, N. (1987). The learning of syntax: An empiricist approach. First Language, 7, 97-120.