Chapter Eight
Language acquisition
Outline
Introduction
Research methods in the study of language development
- Diaries and Parental Reports
- Observational Data
- Interviews
- Experimental Techniques
- Research Design
The development of speech perception
- Methods for studying speech perception in infants
How speech perception develops the child's lexicon
- Before First Words
- First Words
- Some Words Are More Difficult to Learn than Others
- Lexical Organization and Word Association
Learning to make & understand sentences
- Assessing Syntactic Knowledge
- Methods for Assessing Syntactic Understanding
- Moving from Words to Sentences
- Learning to Make Sentences in English
- The Role of Word Order Strategies in Sentence Formation and Comprehension
- Combining Sentences
- Some Later Acquisitions
- Metalinguistic Awareness
Learning to communicate: Early social uses of language
- Learning to Take Perspective: the Demise of Egocentrism
Theories of child language acquisition
- What Must Theories of Language Development Account for?
- General Features of Theories
- Major Dimensions in Language Development Theories
- Linguistic/innatist Theory
- Learning Theory
- Cognitive Theory
- Social Interactionist Theory
- Connectionist Theory
Perspectives: What do the data tell us about the theories?
Summary
Key Concepts From the Book
I. Developmental psycholinguistics is the study of how and why children acquire language. Developmental psycholinguists have discovered that children learn language in an orderly and predictable way, and have attempted to discover the biological and social processes that make language development possible or, perhaps, inevitable. (Page 348)
II. The systematic study of children's language is relatively new. Early modern studies, some carried out as early as the 18th century, focussed primarily on what children said, usually based on observations of the author's own children which were kept in diary form. Although diary studies can be a valuable adjunct to other research on children's language, they tend to focus on what is unusual or interesting rather than on what is daily and ordinary. (Page 349)
III. Spurred by dramatic changes in linguistic theory during the 1960s, developmental psycholinguists began to collect tape-recorded conversations and conduct experimental research based on children's ability to produce and comprehend specific structures in English. One of the most important early studies conducted by Roger Brown at Harvard University during the 1960s, examined the language development of three children known as Adam, Eve, and Sarah. These studies fostered our earliest understanding of how children acquire basic sentence structures in English. (Page 349)
IV. Most studies of child language involve the intensive scrutiny of small numbers of children. This is so because of a number of specific requirements of studying the language of children, including the need to study a child's linguistic attempts over time, and because current context, previous context, pronunciation, and gestural support are important to understanding a child's early utterances,. The results of these studies usually agree with each other. (Page 350)
V. During the 1980s, a large number of finely detailed transcripts of early child language were entered into a computer data base called the Child Language Data Exchange System (CHILDES). This data base enables researchers to examine and pool language transcripts already in existence. (Page 350-351)
VI. In general, we explore children's language systems indirectly. To study metalinguistic development, however, interview techniques are useful. One might ask a child about his or her favorite word, what is a long word, etc. Very young children are unable to distinguish the word from its referent. (Page 351)
VII. Various experimental techniques assess infant and child knowledge. Child language studies may be either cross-sectional or longitudinal. Observational studies may be naturalistic or they may be controlled observational studies, carried out in a laboratory playroom. Experimental studies involve interference on the part of the researcher. These involve an experimental group and a control group. (Pages 352-355)
VIII. Researchers can perform categorical perception experiments on young infants by taking advantage of the fact that babies suck on pacifiers at a higher rate when a sound is new to them and stop sucking vigorously as they enter a stage called habituation. This research design is called the high amplitude sucking paradigm (HASP). It has been used to show that infants are able to discriminate voice-onset-time and place of articulation. Another way to monitor infant interest is through cardiac deceleration. (Page 355)
IX. Infants of 6-18 months are usually tested using the conditioned head-turn procedure. The child is conditioned to turn her head to see a novel object whenever a novel sound is introduced. (Page 356)
X. Infants can hear before they are born and demonstrate a preference for their mothers' voices shortly after birth. By about 4 days of age, they can also discriminate between utterances spoken in their mothers' language as opposed to a foreign language. Within the first weeks, they perceive voice-onset-time and can discriminate differing places of articulation as well as the difference between male and female voices and stimuli of different intonation contours. They also show an early preference for language broken into clausal chunks, an ability that may be a bootstrapping strategy to make further advances in language analysis.(Page 356)
XI. Infants are able to make fine discriminations among speech sounds as early as the first weeks of life, and are able to hear subtle differences between sounds that adults in the community cannot. By the end of the first year, when the child has begun to learn the sounds of the speech community around her, she loses the ability to discriminate among sounds not in her native language. Perceptual loss is a consequence of the infant's continued interaction with her native language.(Page 356-357)
XII. Caretakers tend to anticipate children's competencies and will impute intentions to their children before the intentions are actually there. Late in the first year, the child actually begins to demonstrate intentions and express them in prelinguistic ways that include gestures, consistent word-like sounds, proto-declaratives and proto-imperatives. (Page 358)
XIII. Children begin to produce recognizable words of their language at about 1 year of age. By 18-20 months they typically have a vocabulary of 50 words; by 2 years about 200-300 words. The first words spoken by children in communities around the world are similar in phonetic form as well as in kinds of meanings that underlie them. First words consist mostly of open syllables and tend to refer to things within the child's environment she can actively interact with. During the one-word, or holophrastic stage, children use a single word to function as a complete sentence. (Page 358)
XIV. Language development in normal children follows a predictable course. At the earliest stage, children produce one meaningful word at a time, inevitably a concrete content word rather than a function word or an abstract word. At this stage, children use language to signify a variety of intentions such as negation, recurrence, nonexistence and notice. Children learning sign language as a first language may begin to use words slightly earlier than children learning oral languages: the production of words may depend somewhat on oral motor coordination. (Page 359)
XV. Early words usually are restricted to discussion of on-going context. Children appear not to make referential use of language until after they have learned their earliest words. (Page 360)
XVI. At this stage, children understand about five times as many words as they can produce. For many children learning English, the lexicon consists primarily of nouns, which tend to be exemplars of basic level categories. They also typically engage in overextension, by, for instance, calling anything that can be sat in a "chair."Underextension is less common, but still frequently observed. Occasionally children mismatch an overheard word to a wrong meaning.(Pages 360-361)
XVII. Children may use various strategies to determine what words mean. These include:
A. Reference
B. Extendibility
C. Object scope
D. Categorical scope
E. Novel name-nameless category
F. Conventionality (Page 361)
XVIII. Researchers debate how words are represented in the child's mind. An early proposition was that children acquired semantic features for words. Prototype theory has also been proposed as an organizing principle for children's lexical learning. (Pages 361-362)
XIX. Not all words are equally easy for children to learn. Children show a preference for concrete vocabulary terms and have trouble with relative concepts until they are somewhat older, giving instead absolute meaning to these words. Children are often 8-10 years old before they appreciate that the language includes ambiguous words. They also do not fully understand kinship terms until they are 7 or older. Although children have smaller lexicons than adults, they take longer to recognize and retrieve words in experimental tasks. (Page 362-363)
XX. By the time children enter kindergarten, they know an estimated 14,000 words and will acquire about 300 new words each year. Some evidence indicates that children begin to organize these words into semantic networks, undergoing some sort of lexical reorganization during childhood. Whereas adults usually produce synonyms, antonyms and other members of the stimulus's lexical category (paradigmatic responses) on word association tests, young children often give syntagmatic responses. (Pages 364-365)
XXI. One method of assessing syntactic knowledge is asking children to imitate adult sentences: children are usually unable to imitate sentences above their level of syntactic development. Another way is to ask children to act out sentences using objects. For younger children, the preferential looking paradigm has been used to test children's understanding of grammar. (Page 364-365)
XXII. During their second year, when children have a productive vocabulary of about 50 words, they begin to put together two-word sentences. Everywhere in the world, children at this age express the same kinds of thoughts and intentions. In English, these sentences lack articles, prepositions, inflections and grammatical modifications. Such speech is said to be telegraphic. (Page 366)
XXIII. Although children follow regular patterns of language development, they also display considerable individual variation, for instance, some children may use adult-like prosody while still in the two-word stage. (Page 366-367)
XXIV. In English, where the root form of a verb can often stand alone, children learn the inflected forms more slowly than in languages like Spanish in which uninflected forms of the root are not permissible. The emergence of grammar is determined by a number of factors including the pervasiveness and regularity of a language's grammatical constructions. (Page 367)
XXV. Some theorists believe that universal grammar specifies parameters for some variable features of language. These parameters may also have default settings. (Page 368)
XXVI. Once children begin to acquire grammatical markers, most of them do so in basically the same order within a particular language. After children begin to learn regularized plurals and past tenses, they create some regularized forms of their own: this is called overregularization, and occurs in all languages that have been studied. (Page 368)
XXVII. Brown and other researchers developed the Mean Length of Utterance (MLU), to track the development of language in children. The MLU is a measure of the average number of morphemes in a child's utterances. (Page 368-369)
XXVIII. Acquisition of the negative in English progresses through three major stages: 1) negative markers, such as "no" and "not" simply precede the utterance; 2) negatives follow the main verb and "can't" and "don't" appear in sentences; 3) negative markers are placed on the verb auxiliary, creating forms such as "isn't." (Page 369)
XXIX. In learning to ask questions, children usually learn well formed Yes/No questions before Wh- questions. What, where and who questions usually precede why, how and when questions. (Pages 369-370)
XXX. Young children learn the S-V-O word order of English very quickly and have difficulty understanding passive and dative sentences that violate S-V-O word order. Many studies suggest that children younger than 5 will use either an S-V-O word order strategy or event probability to comprehend passives and datives. They may adopt a minimal distance principle, assuming that the subject for any verb is the noun that immediately precedes it. (Page 372)
XXXI. Children's first compound sentences use the conjunction and, and tend to link objects. Linking of subjects comes later, as does use of other conjunctions such as but and because. In many cases an order of mention strategy appears to apply. Embedded clauses are rarely used by children. (Page 372)
XXXII. Ambiguity, especially that based on syntax rather than word meaning, is difficult for children. The ability to paraphrase develops during the later elementary school years. Understanding idioms and figurative language are also later acquisitions, as is metalinguistic awareness. (Page 373)
XXXIII. The early social intentions of language use include drawing attention to the self, and showing, offering, and requesting objects. Researchers have suggested that some children are more social in their use of language than others, drawing a distinction between children who are referential users of language and those who are expressive users of language. (Pages 373-374)
XXXIV. Children's acquisition of polite forms occurs in conjunction with explicit, active, teaching on the part of adults. (Page 374)
XXXV. Theorists of language acquisition are divided on the extent to which language is innate and to what extent it is learned behavior. Folk theories of language development include the notion that children simply imitate what they hear. However, imitation is not a sufficient explanation for language development.(Page 376)
XXXVI. Theories of language development vary along these major dimensions:
A. Nature or nurture? To what extent is language hardwired into the human brain.
B. Continuity or discontinuity? Does language proceed as an orderly flow, or in stages?
C. Universal competence or individual variation? To what extent is linguistic competence invariant?
D. Structure or function?
E. Autonomy or dependency? Is language a separate faculty of the human mind, or is it dependent on other cognitive developments?
F. Rules or Associations? Is the child internalizing a set of abstract cognitive principles or learning language without recourse to rules? (Pages 376-377)
XXXVII. Innatist theorists believe that the principles of language are inborn and not learned, because children universally acquire a successful grammar between 1 and 6 years of age, even without access to consistent grammatical models. Deaf children who are not exposed to any language in infancy develop their own manual sign systems that incorporate many of the formal features of language.(Page 378)
XXXVIII. Other evidence of children creating language comes from a study of Hawaiian pidgin. Hawaiian children who heard only pidgin gradually transformed their language into creole, producing grammatical forms more complex than those their parents used. (Page 378)
XXXIX. Proponents of the innatist view list three major concerns that underlie the assumption that language learning is dependent on specific biologically determined abilities:
A. The poverty of stimulus problem
B. The fact that children learn to use and understand sentences that never occur in their language learning environment
C. The negative evidence argument (Pages 378-379)
XL. Linguistic theorists rely on a special abstract mental mechanism called a Language Acquisition Device (LAD), that enables children to attend to the language that adults around them speak, make hypotheses about how it works, and derive an appropriate grammar. (Page 379)
XLI. According to Chomsky's theory, the principles underlying all possible human languages are considered to be innate and all languages adhere to a set of rules known as Universal Grammar. According to this view, many aspects of language development are preprogrammed in the individual and the child does not require explicit teaching or experience in order to acquire language. Children are also assumed to come innately endowed with linguistic switches, or parameters that they set once they hear the language of the community around them. Language is considered an autonomous faculty, separate from intelligence. (Pages 379-380)
XLII. Behaviorists or learning theorists claim language is acquired according to the general laws of learning. According to this view, language behavior is reinforced and shaped by adults. Learning theory includes at least three kinds of learning:
A.Classical conditioning causes children to associate words with things in the real world through the processes of stimulus and response.
B. Through operant conditioning, language behavior is selectively reinforced and behavior that is not reinforced becomes extinct.
C. Finally, social learning causes children to observe and imitate others. (Page 380-382)
XLIII. Cognitive theorists believe that language acquisition is tied to cognitive development. According to this theory, children first learn a concept and then are able to "map" a word onto that concept. It is difficult to know to what extent language development is dependent on or part of other development and to what extent it is autonomous. (Page 383)
XLIV. Certain language development milestones occur at around the same time as cognitive milestones. There are some general cognitive prerequisites for language, however, it is uncertain how much language is based on cognitive stages. (Page 384-385)
XLV. Speech to babies is different in many ways from the ordinary speech adults use to communicate with each other. Child directed speech (CDS), in our society is characterized by slow rate, exaggerated intonation, high fundamental frequency, repetitions, simple syntax and simple, concrete vocabulary. Although the exact features of the baby talk (BT) register vary from language to language, BT appears in all languages. (Page 385)
XLVI. Social interactionists believe that biological factors are necessary but insufficient to ensure language development. According to the social interactionist viewpoint, there is no biological critical period for language acquisition. Language is considered a communicative behavior that develops through interaction with other human beings: the language acquisition socialization system, (LASS) is offered as an alternative to Chomsky's LAD. (Page 386)
XLVII. Interactionists stress that children acquire language in part through the mediation of others. Thus, interaction rather than exposure is seen as necessary. Interactionists stress the functional basis of language, paying particular attention to the interpersonal reasons that children have for speaking, the way older speakers tailor their linguistic interactions with infants, and the effect of different kinds of input on children's development. (Page 386)
XLVIII. Social interactionists dispute the idea that children receive no negative evidence. Parents respond in a number of ways to unsuccessful language attempts, including the tendency to recast them and provide other cues to grammar. They also focus on phonology, the lexicon, and pragmatics. (Page 387)
XLIX. Connectionist or Parallel distributed processing (PDP) models explore how information is built into a system through neural connections. According to these models, memory for experiences is distributed across many processing units that become connected to each other in a neural network. Children learn language by establishing neural associations between words and objects. Sufficient exposure to grammar leads to the establishment of neural networks. According to this view, children do not learn rules for language: the ability to do such things as come up with the plurals of nonsense words is the result of an automatic neurological process. (Page 388)
L. Critics of PDP models point out that the most parsimonious explanation for the acquisition of grammar is still that the child internalizes a set of rules or principles. It has become increasingly plausible that multiple mechanisms may be involved in the full range of language skill acquisitions. (Pages 388-389)
LI. Assessment of the data on child language acquisition supports the role of both nature and nurture: while it is true that the ability to learn language is innate, some properties of language are clearly taught. (Page 389)
LII. Linguistic capacity relies upon neuroanatomical structures. Left hemisphere specialization appears to be innate: cerebral asymmetry has been observed in fetuses. Thus, humans are born with the neuropsychological endowments to learn language. (Pages 389-390)
LIII. Language development is also connected to cognitive, affective, and social development. The innate social and affective disposition of babies causes them to display many types of attachment behavior. Language use can be seen as serving the social and affective needs of both the child and the caretaker. Babies prefer seeing human faces to seeing objects, and seem to become attached to their mother's voices in utero. (Page 390-391)
|
 |
|
|
