The Discovery of Spoken Language.pdf

< BACK

The Discovery of Spoken Language

Peter W. Jusczyk

Acknowledgments

1 Surveying the Terrain

2 A Brief Historical Perspective on Language Acquisition

Research

3 Early Research on Speech Perception

July 2000

ISBN 0-262-60036-6

328 pp., 6 illus.

$30.00/£19.95 (PAPER)

4 How Speech Perception Develops during the First Year

5 The Role of Memory and Attentional Processes in the

Development of Speech Perception

6 How Attention to Sound Properties May Facilitate Learning

Other Elements of Linguistic Organization

7 Relating Perception to Production

8 Wrapping Things Up

Appendix Methodology Used in Studies of Infant Speech

Perception

Notes

ADD TO CART

Other Editions

Cloth (1997)

References

Name Index

Series

Bradford Books

Subject Index

Language, Speech, and

Communication

Chapter 1

Surveying the Terrain

Language involves a duality of patterning, as Hockett (1954) has noted.

On the one hand, there are patterns that pertain to the way that sounds

are organized; on the other, there are patterns that relate to how meanings

are organized. Although language use involves dealing with patterns at

both levels, it often seems as though the two are studied in relative isola-

tion of one another. This seems especially true with respect to the study of

language acquisition. Peruse a typical textbook in the field, and you are

apt to find that only a relatively small portion of the book is concerned

with the development of speech perception and speech production. There

are several possible reasons for this. It could be the case that there is

simply a lot less known about the way speech perception and production

capacities develop. Another possibility is that research on perception and

production makes relatively little contact with the rest of the research on

language acquisition.

Historically, there is support for both of these contentions. Although

investigations documenting the growth of speech production have a rela-

tively long history in diary studies (Ament 1899; Gregoire 1933, 1937;

Leopold 1939, 1947; Scupin and Scupin 1907; Stern and Stern 1928),

extensive research on infant speech perception had its inception during

the last quarter century. Thus, with respect to how the receptive side of

speech processing develops, there really was not a great deal of informa-

tion available until fairly recently. However, the same cannot be said for

developmental studies of speech production. Even the early diarists took

pains to record some of the changes in pronunciation that occurred in

the child's early words, although to be sure, the accuracy and reliability of

phonetic transcriptions have improved enormously with the advent of the

tape recorder and more sophisticated technology for analyzing speech

production. Hence, to fully explain the separation between speech and the

Chapter 1

rest of language research, we also have to consider the second alter-

native—that the two domains have made little contact with each other.

The fact that speech researchers typically receive very specialized training

does have something to do with this. Not every linguist is trained to do, or

is well-experienced in doing, phonetic transcriptions. Furthermore, speech

perception researchers are trained in the basics of acoustics and signal

processing. The kinds of analyses that they perform, the dimensions they

examine, and even the terminologies they employ are pretty foreign to the

rest of linguistics. Perhaps, then, it is not surprising that speech research

has not been better integrated with psycholinguistic research in general

and with language acquisition research in particular.

There was also another factor that helped encourage the separation of

speech research from other aspects of language acquisition. This had to

do with the way research in linguistics was conducted during the 1960s

and 1970s. Generative grammarians such as Chomsky (1965; Chomsky

and Halle 1968) and his colleagues held that each level of linguistic orga-

nization could be studied more or less independently from other levels. In

fact, according to this view, one could best capture the correct general-

izations about the organization of some component of language, such as

syntax, by studying it in isolation from the other components. Only when

the structures of each component were understood would it be possible to

examine the integration and interaction of the various components.

Finally, there was still another factor that likely played a role in limiting

the integration of infant speech-perception studies into language acquisi-

tion studies overall. Many of the early investigations demonstrated that

even very young infants (i.e., 2- to 3-month-olds) apparently have well-

developed speech perception capacities. These findings may have helped

give rise to the belief that all the basics for speech perception are already

in place well before the bulk of language learning even begins. Hence,

those investigating how the child's semantic and syntactic development

occurs could just assume that the child had already succeeded in seg-

menting words from fluent speech. In part, then, language researchers did

not need to worry about developmental changes in speech perception

capacities because there did not appear to be any real changes.

This picture has begun to change during the past decade, and it has

done so for a variety of reasons. First, the general Zeitgeist with respect to

whether components of language should be studied separately or not has

changed. More and more research is being devoted to the way different

linguistic levels interact during the course of language production and

Surveying the Terrain

comprehension (Bates and MacWhinney 1989; Marslen-Wilson and Welsh

1978; McClelland and Elman 1986; Seidenberg and McClelland 1989;

Tanenhaus et al. 1993). Second, we have a much clearer picture of the

basic speech perception and production capacities of infants. This picture,

in turn, has provided investigators with a backdrop for viewing develop-

mental changes in these capacities. Moreover, it has become evident that

many changes to speech perception and production capacities are occur-

ring within the first year of life and that these capacities are very much

influenced by the kind of input to which infants have been exposed. Third,

speech researchers themselves have begun to make efforts toward relating

their findings to other findings in psycholinguistic and language acquisi-

tion research. In fact, this book is one such attempt to situate the findings

of infant speech-perception research more squarely within the field of

language acquisition.

Some Characteristics of Speech Perception

For the average adult, there is little mystery in speech perception. It is just

a matter of hearing the words in the order in which they are spoken. Aside

from the times in which interfering noises are present, the whole process

of speech perception seems rather effortless. One simply hears the sounds

and grasps the meanings that they stand for. In fact, the transition from

sound to meaning is so seamless that we commenly hear right through the

sounds directly to their meanings. The whole process is so fluid that, other

than learning which sound patterns go with which meanings, it is hard to

believe that learning plays much of a role in speech perception. Yet, this

process of going from sound patterns to meanings, which is so easily

accomplished by humans, still has not been successfully implemented on

machines (Marcus 1984; Reddy 1976; Waibel 1986). Among the impedi-

ments to successful machine recognition of speech, boundaries between

successive words are not clearly marked in the speech stream (Cole and

Jakimik 1980; Klatt 1979, 1989), and the acoustic shapes of words are

frequently affected by the nature of the words in the surrounding context

(Liberman and Studdert-Kennedy 1978; Mills 1980). Hence, speech re-

searchers are well aware of the fact that the perception of fluent speech is

a lot more complex than it first appears.

There are moments when we do appreciate some of the complexities in

speech perception that we overcame when learning to speak and under-

stand a language. For instance, when we are with people speaking an

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: