Phonology

A Cognitive View

Jonathan Kaye

• Publisher: Routledge
• Year: 2013
• eBook ISBN: 9781136562358
• Topic: Sprachen & Linguistik
• Subtopic: Sprachwissenschaft

Designed to acquaint the reader with the field of phonology -- the study of the systems of linguistically significant sounds -- this book begins with a brief introduction to linguistics and a discussion of phonology's place within that field. It then goes on to cover a variety of topics including the nature of phonological units, phonological rules, which types of phenomena interest phonologists, and the evolution of phonological theory.

Suitable for many applications, this volume assumes no previous knowledge of linguistics. An excellent text for use in first or second year phonology courses, it will also be of value to those involved in cognitive science, neuroscience, artificial intelligence and computer science.

1

Introduction

WHAT IS LINGUISTICS?

Phonology may be defined as the study of the systems of linguistically significant sounds. At this point such a definition is rather meaningless. First, we need to know what “linguistically significant” means. When does a sound have this property? How do we know? Clearly “linguistically significant” must mean something like “of significance to linguists or to the object of their study.” It is reasonable then to begin our discussion of phonology with an explanation of the nature and the goals of linguistics. A clear idea of the field will provide us with a setting in which we can properly place phonology.

I should make it clear at the outset that the portrait of linguistics that I am about to sketch is by no means the only one. The field of linguistics represents a variety of different scholarly traditions, and each one carries its conception of what it is about (see the volume by Morgan, in prep., in this series). Linguistics departments may be found in the faculties of humanities, social sciences, behavioral sciences, and natural sciences. Such diversity is not purely administrative. People called linguists have dramatically different views as to what it is they are studying, what kind of results they are looking for, and how such results should be interpreted. I emphasize again that I will not present a balanced, objective, and comprehensive view of linguistic theory and practice nor of that branch of the field that deals with problems of a phonological sort. The conception of linguistics that I am about to present may be characterized rather broadly as that of generative grammar. This approach, proposed in its initial form by Noam Chomsky in the mid 1950s, has certainly become the dominant school of linguistic theory both in terms of the numbers and the geographic diversity of those who utilize it. As is to be expected, the field of generative grammar has constantly evolved in the course of the last three decades.¹ It would be inaccurate to characterize it today as a single approach. It represents rather a family of approaches sharing a certain vision of the field but differing in fundamental ways in both form and content.

What is it then that linguists study? A standard joke among linguists is that the first question one gets asked after having revealed one's profession is “Oh, so you're a linguist. How many languages do you speak?” The general view is that a linguist is somebody who studies languages and consequently winds up speaking a lot of them. It is true that some linguists speak an impressive number of languages, but this is typically due to personal taste and opportunity rather than to professional necessity. It is quite untrue, however, that linguists study languages. Indeed, it is unclear what it would mean to study languages (beyond learning to speak, read, or write them).

In fact, what linguists study is not languages but rather the particular systems that underlie them: linguistic systems or grammars. What then is a system that underlies language? In a word, linguists do not study languages but rather Language or, more precisely, that cognitive faculty that underlies our linguistic capabilities. Why should anyone believe that such a thing exists? To answer these questions we must consider what it means to “speak a language.” Following Chomsky, we can assume that this implies the ability to produce and to understand an unlimited number of sentences proper to the language in question and that the overwhelming majority of these sentences have never been heard or uttered before. In our day-to-day world, most of what we hear and speak is novel. A moment's reflection suffices to convince oneself that these novel utterances pose no particular difficulty. We produce and understand them with no apparent effort.

This is all well and good, but what does it tell us about the linguistic systems that underlie this ability? We have seen that speaking a language implies the ability to produce and understand a potentially infinite class of utterances or sentences. Furthermore, for any particular individual, this ability is limited to a tiny subset of the world's languages. The ability to speak, say, English does not have as an automatic consequence the ability to speak Zulu, French, or Navajo. This means that speakers of English possess a particular knowledge that permits them to speak and understand this language. Those who do not speak English do not possess this particular knowledge. They will possess a knowledge of whatever language or languages they happen to speak. We call this knowledge linguistic competence. It must be borne in mind that I am employing the term knowledge in a very special way, that is, I do not mean conscious knowledge. Speakers of English are no more able to express the nature of this knowledge than they are to explain the stereoscopic nature of their vision or the manner in which they recognize people they know. It is the understanding of the exact nature of this linguistic competence that is the primary objective of linguistics.

This characterization of linguistic competence can aid us in determining certain properties of the linguistic systems that are its source. For example, it is quite clear that our linguistic system cannot be simply a list of sentences. Such an hypothesis encounters serious difficulties: In the first place, there are an unlimited number of sentences associated with any human language. The storage capacity of the human brain, while doubtless quite large, is nevertheless finite. Second, it is difficult to imagine a scenario for acquiring a linguistic system that is a list of sentences. I have already pointed out that the vast majority of linguistic utterances are novel; they have never been heard or produced before by the individual in question. This excludes simple imitation as a learning theory. One cannot imitate something one has never heard before.

There are a number of ways (some trivial) in which we can illustrate the unlimited nature of linguistic competence. Consider the sentence:

(1) I ate 42 oranges.

Clearly, one can replace the integer in (1) by any other and the result will still be a grammatical English sentence. In coordinate constructions (those involving and, or, etc.), there is no obvious limit on how many terms may be conjoined. For example:

(2) I ate an apple and a pear, and a peach, and an aardvark,.…

Expressions such as noun phrases (a noun possibly accompanied by various articles, modifiers, quantifiers, and so on) may themselves contain noun phrases. This recursivity once more leads to a potentially infinite number of utterances, as in (3).

(3) I ate the apple that the boy whom Harry saw yesterday threw at the fence that the girl painted with the brush that fell from the truck that John's brother drove.…

Sentence (3), while stylistically somewhat strained, is perfectly comprehensible. Notice that there is no point at which it must be terminated. It may be extended indefinitely in the ways already mentioned.

The three examples also illustrate the fact that we cannot consider a language to be a list of words. A list of words, being a list, runs into the same sort of difficulties that we have just discussed vis-à-vis sentences. There are other problems. Anyone who has attempted to understand a newspaper written in a foreign language using only a dictionary has a small idea of what these problems are. Let us consider some very simple examples.

(4) John saw Jane.

Sentence (4) consists not only of three words but of certain grammatical relations that connect them. Any speaker of English knows that in (4) John is the agent, that is, the one who sees, and Jane is the object or patient, the one who was seen. Changing the order of these words changes or destroys the meaning. Thus, (5a) does not mean the same thing as (4), and (5b) and (5c) are not possible English sentences.

(5) a. Jane saw John.

b. *Saw Jane John.²

c. *John Jane saw.

Thus, in addition to knowing English vocabulary, a speaker must know the principles of sentence construction: the syntax. To a certain extent, these principles may vary from language to language. If we simply translated (4) word for word into Japanese or Desano (a South American Indian language) the resulting form would be quite ungrammatical. In fact, (5c) would be the corresponding form in both these languages. In a similar vein, in English a verb such as look requires a preposition, whereas a verb such as discuss does not.

(6) a. Mary looked at the pictures.

b. *mary looked the pictures.

c. Mary discussed the equation.

d. *Mary discussed of the equation.

In French, the facts are exactly the reverse. The verb regarder (look at) must not be followed by a preposition, whereas the verb discuter (discuss) req...

Table of contents

• Cover
• Half Title
• Tutorial Essays in Cognitive Science
• Full Title
• Copyright
• Dedication
• Contents
• Preface
• 1 Introduction
• 2 Phonological Phenomena
• 3 The Search for Explanations
• 4 Beyond Tones: Extending Nonlinear Phonology
• 5 Current Phonological Research and Its Implications
• References
• Author Index
• Subject Index