[Back to Main | Show Dates | Photos | Language Index]

Rules and Representations

As with the rest - this file is eternally undergoing modification. Many of the bigger issues will later (if they do not already) include links to a more microscopic analysis. But again, we can only code so fast. For the time being, enjoy.

If we stick to the assumptions elaborated by Ferdinand de Saussure, we can take the study of language to be the study of the structure of language. We can further constrain the object of study by suggesting that language is a system which tranforms sound structures into conceptual structures(See: The elements of the sign. Here the sound structure is the mental representation of a sound and the conceptual structure is the mental representation of some "real-world" situation. Langauge, then, is the process by which one representation is transformed into another.) In this way, language can be taken to be a structure which, in turn, manipulates structures. In an attempt to express the nature of language "from the bottom up" we have begun to reach an impasse. Certainly, we cannot fully characterize a system of structures if we are not entirely aware of the character of the structures themselves. Conversely, if Saussure is correct in assuming that to "think of a sign as nothing more [than the combination of a certain sound and a certain concept] would be to isolate it from the system to which it belongs"; that "it would be to suppose that a start could be made with individual signs, and a system constructed by putting them together"; that "the system as a united whole is the starting point, from which it becomes possible, by a process of analysis, to identify its constituent elements" (p112: emendation from context - crw), then we need to have a knowledge of the system before we can explore the character of the structures. Does anyone smell an infinite regress? It stands to reason, then, that there must be a system of analysis which is capable of manipulating arbitrary relations, and, which is also capable of determining those relations during the early stages of childhood. A remarkable achievement, to say the least.

It is from this perspective that we can begin to see the importance of The Innateness Hypothesis as it is formulated by Noam Chomsky in the form of his Universal Grammar. This issue will be discussed later in more detail.

First, a brief contextualization of the evidence to be presented is necessary.

(1) Many of the operations of the mind involve systems of representations and also sets of rules for manipulating those representations. (Rules and Representations)
(2) The Language Faculty is a system of rules and representations and one of its operations (its grammar) is the process by which the acceptable sentences of a language can be generated.
(3) The rules and representations referred to are reflective of a specific apparatus with specific structural properties.
(3b) These are, in turn, reflective of a specific internal organization.

The assumptions laid down in (1)-(3) are indicative of the approach advocated by Chomsky which associates an internal organization with the processes involved in the language faculty. The necessary questions being begged here are: What is the character of the Language Faculty? and How does it relate to the process by which a child comes to acquire a (real) language? Chomsky addresses the first question by employing the notion of a generative grammar. By attempting to view the empirical facts of Natural Language in terms of the predictions made by such a grammar and by attempting to adjust the details of such a grammar according to these empirical facts, we can begin to see how a finite brain might be capable of processing/producing a (potentially) infinite set of different sentences. By adding a transformational aspect to our grammar we can express many of the regularities of a language. But the source of these mechanisms is still a mystery and even more mysterious is the process by which those rules specific to a person's native language are learned.

In order to handle the problem of language acquisition, Chomsky introduced the notion of Universal Grammar(UG). It is helpful to think of UG as a process, rather than as a thing. Roughly, we can take UG to be the process by which children learn their native language from environmental conditions. In this way, UG is not actually a grammar but, rather, a method by which to build a grammar from the evidence present in a child's environment. Since any infant from any culture is capable of learning any language if placed in the appropriate environment, this method should be common to every (normal) child and, therefore, should be biological (genetic) in character. In other words, Chomsky argues that Universal Grammar is innate. A language, by this theory, is acquired through a process (UG) which is biologically endowed to our species.

These assumptions provide a stark backdrop against which to view the empirical facts of Natural Language. They are, however, merely the assumptions of an approach to the study of linguistic regularity. That is to say, they act as a set of constraints for deciding between various explanations of a linguistic phenomenon, while, simulateneously, attmpting to guide research by differentiating between those questions which are interesting and those which are not.

Of course, this is only the tip of the iceberg compared to the treatment given in the literature. I would direct anyone interested in the generative tradition and universal grammar to read the primary sources. The text described here is a good start, but the best analysis seems to be in Aspects of the Theory of Syntax (1965) and The Minimalist Program (1992?)
-crw