Vista previa parcial del texto
¡Descarga reading semantics y más Apuntes en PDF de Idioma Inglés solo en Docsity!
UNIT 1 What Is Linguistics? Leading Questions In beginning the study of any field, one good way of orienting yourself is to find out what problems the field works on. What leading questions does it seek to answer? In the approach to linguistics we will follow, the leading questions are very easy to formulate. In day-to-day conversation, we routinely speak of people “knowing English” or “knowing Japanese and Korean.” We talk about a language as a body of knowledge that people do or do not possess. The leading questions of linguistics arrange themselves around this commonplace way of talking: they address knowledge of language. Whenever someone can be said to know something, a number of basic questions present themselves. | Knowledge of X What exactly does NA that How is that ? ? this person know? owledge? knowledge used? 10 Unit 1: What Is Linguistics? Knowledge of chess (5) fe say people “know chess” or “don't know chess.” The basic questions about knowledge of chess are these: »s What exactly does How did the How is the knowledge someone know when erson learn of chess actually he or she knows the P used in the process of the game? . game of chess? playing a real game? o o y» A list of rules for By having some- ñ h Do players mentally moving pieces? one explain the construct a table of rules? Strategies for rules? By watch- Do they use those rules moving them? ing other people to produce some kind play? of mental image of the board that they manipu- late in their heads? 2 Unit 1: What Is Linguistics? other than Homer (“who”) is expected to be the surpriser, and the surprisee (“him”) may be either Homer or some third party. Finally, sentence (3) is identical to sentence (2) minus the word him, but now Homer again must be the surpriser, rather than the surprisee. These facts are remarkably intricate and subtle, yet immediately obvious to anyone who has mastered English. But what principles are we following in making these judgments? In fact, we don't have a clue—not initially, at least. True, we can make complex judgments about sentences like these. But we cannot directly grasp the basis of our judgments. People don't consciously know why, when they say 1 wonder who Homer expected to surprise him, the name Homer and the pronoun him will be taken to refer to different people. The knowledge that we possess of our language is almost entirely uncon- scious or tacit knowledge. In this respect, language appears to be similar to other important parts of our mental life. Sigmund Freud is famous for having proposed that much of the mind”s functioning and contents lies entirely hidden to con- sciousness. Freud held that unconscious phenomena and processes are no less psychologically real than conscious ones, and that appeal to them is just as necessary for an understanding of human cognition. l handle unconscious ideas, uncon- scious trains of thought, and uncon- scious impulses as though they were no less valid and unimpeachable psy- chological data than conscious ones. [Ana] of this | am certain—that anyone who sets out to investigate the same region of phenomena and employs the same method will find himself com- pelled to take the same position ... —Fragment of an Analysis of a Case of Hysteria (“Dora”), p. 232 Sigmund Freud 1856-1939 For the most part, the principles and operations behind knowledge of language lie outside the range of consciousness and cannot be recovered by simply sitting down, staring off into space, and thinking hard. A Talking Analogy 13 A “Black Box” Problem Tf we can't directly intuit what's in our minds, then our only option is to approach the investigation of internal things (like knowledge and mental states) as we would approach the investigation of external things (like birds and planets). That is, we must formulate explicit theories about what we know, and we must find ways to test, refine, and extend those theories in order to reach a satisfactory explanation of the facts. Since we can't look directly at what's inside the mind, our job will be to figure out what's inside on the basis of what we can observe from the outside. Problems of this kind are sometimes called black box problems. In a black box problem, we have an unknown mechanism that receives observable input data and produces observable output behaviors. The task is to figure out what's inside the box on the basis of inputs and outputs alone. Observable input Y «y Observable output In the case of human language, the observable input is the speech data that people are exposed to as children, the language that they hear around them. The output is their various linguistic behaviors as children and as adults: the sentences and other expressions that they produce, their judgments about their speech and the speech of others, and so on. By carefully examining this kind of information, the linguist must deduce the language mechanism that lies within the human mind. A Talking Analogy To make the black box nature of the problem more concrete, consider a simple analogy (due to MIT linguist James Harris). For many years, toymakers have produced talking dolls of various kinds. Some have a string on their back or neck that you pull. Others have a button on their wrist or stomach. Still others talk when you talk to them (although these must be turned on initially with a switch). A Talking Analogy 15 open and look inside. For example, since the doll produces a very limited range of utterances and all repetitons of a particular utterance are identical, it is very likely that the utterances are stored within the doll as whole chunks, not con- structed online. That is, it is likely that the doll contains a storage unit loaded with all of its utterances; pulling the string causes a whole, individual stored utterance to be played back from its beginning. It is likely that the doll contains a storage unit loaded with all of its utterances; pulling the string It is unlikely that the doll mechanism constructs utter- ances online from smaller causes a whole, individual stored utterance to be played back from its beginning. parts, with the parts and their rules of combination being what's stored. Deducing what's inside humans is vastly more complex than deducing what's inside the doll, but already we can see some things by contrast. For example, since we humans produce an enormous range of utterances, without exact repe- titions, it's very unlikely that we have utterances stored within us as whole chunks. Rather, we probably do construct our utterances from smaller parts as we speak, with the parts and their rules of combination being what's stored. With humans, then, something different and more complex is involved. As we will see in later units, the rich complexity of linguistic data—the speech we hear around us, the output we observe—allows us to conjecture a very rich mechanism inside the human mind. Deducing What's inside the Box from the Input The data we draw on in solving a black box problem come not only from “output behavior”: in our present case, the utterances produced by talking dolls, or the utterances and linguistic judgments produced by talking humans. They also come 16 Unit 1: What Is Linguistics? from the input the mechanism receives. Often we can deduce what kind of mechanism is inside the black box by seeing what kind of information initially went into it, For example, going back to our analogy, suppose you observe that, for the doll, “learning” the ten or so utterances that it produces involves a human being producing each of these utterances. Perhaps you visit the factory where the dolls are made and you observe a person speaking into a microphone that is connected to the doll by a wire. You observe that the doll's speech exactly repeats that of the person speaking into the microphone, that the utterances the doll ultimately produces are copies of the human's speech. Such evidence would clearly support your hypothesis that the doll contains some kind of storage and playback device— a disk, a tape player, or something similar. So, the circumstances in which the doll acquires its language can give us information about the mechanism inside it, even when we can't observe this mechanism directly. Let's play! Let's play! Comparisons with Human Language Applying this strategy to human language yields surprising results—indeed, some of the most fascinating results in all of the cognitive sciences. Clearly, humans do not learn language like our talking doll, or like a parrot. Although children do repeat expressions that they hear around them in day-to-day speech, often very closely matching the intonation, pitch, and timing of words, their speech goes far beyond what they hear. Children, and indeed humans generally, are extremely creative in their language use, routinely producing utterances they have never encountered before. Furthermore, the data that form the input to human language acquisition are not clean and precise. Our doll's utterances were “learned” from very precise, careful speech uttered into a microphone, perhaps in the sheltered environment of a sound booth. But these are not the circumstances in which human speech 18 Unit 1: What Is Linguistics? e Ttis tacit; we come to know many things that we don't know that we know. + Itis complex; it underwrites very subtle and intricate judgments. + Itis untutored; the vast bulk of it was never taught to us directly. + Ttis gained in the face of very impoverished input. One plausible explanation for this picture—perhaps the only plausible explana- tion—has been proposed by the linguist Noam Chomsky. Chomsky suggests that children come to the task of language acquisition with a rich conceptual apparatus already in place that makes it possible for them to draw correct and far-reaching conclusions on the basis of very little evidence. Human language learning involves a very powerful cognitive system that allows learners to infer their grammar from the meager data they are presented with in day-to-day speech. Chomsky terms this cognitive system Universal Grammar, or UG for short. (We may think of Universal Gram- mar as the system of principles that characterizes the class of possible grammars by specify- ing how particular grammars are organized (what are the compo- nents and their relations), how the different rules of these com- ponents are constructed, how they interact, and so on. ... Universal Grammar is not a grammar, but rather ... a kind of schematism for grammar. —Language and Responsibility, Nor. 180, 183 Noam Chomsky Institute Professor Massachusetts Institute of Technology Photo by Donna Coveney/MIT. Reprinted with permission. UG in humans is very roughly analogous to the mechanism inside our talking doll, Although the doll's device is not a deductive conceptual mechanism, it is one that allows dolls equipped with it to “learn” or at least be made to “speak” any language. By simply recording utterances in one or another language on the Universal Grammar 19 disk, drum, tape, or whatever device the mechanism uses for storing its messages, dolls can be made to utter sentences of German, Hindi, Maori, and so on. Furthermore, just as the dolls mechanism is part of its basic physical structure, is specific to that kind of doll, and is found in all dolls of that kind, so too the basic mechanism that makes it possible for humans to learn language is appar- ently part of our physical structure (our genetic endowment), is peculiar to the human species alone, and is found in all members of our species (putting aside cases of pathology). Evidence for Universal Grammar Evidence for basic linguistic endowment in humans comes from at least three SOUrces: + The acquisition process is surprisingly uniform for all children, even though the languages being learned may seem wildly different. + Although the languages acquired by children are superficially diverse, deeper investigation reveals significant, shared design features. + With equal facility and with no special training, all children, of whatever ethnic or genetic background, learn whatever language or languages they have significant contact with. No one has a racial or genetic predisposi- tion to learn one language more readily than another. These facts would be all but impossible to understand if normally developing human children did not come to the task of native-language acquisition equipped with a single standard acquisition device, provided by their biological makeup. 
