Вы находитесь на странице: 1из 36

SLI as a Model System of Language Acquisition

Mabel L. Rice SRCLD-IASCL 2002 Madison, Wisconsin July 2002

Questions

1. What accounts for growth (nongrowth) in childrens language? 2. Is growth (nongrowth) synchronized tightly across all elements of language?

Possibilities:

Sorting out useful distinctions about linguistic growth and the nature of linguistic competence in children

Pitfalls Variance across children in levels of performance Variance across elements of language

Background: What are the Mechanisms That Underlie Childrens Acquisition of Morphosyntax?

Assumption of uniform robustness: All normal children acquire language effortlessly, following the same timing mechanisms and the same general sequence. An emphasis on invariant properties of language acquisition. In general, language acquisition is a stubbornly robust process; from what we can tell there is virtually no way to prevent it from happening short of raising a child in a barrel. Pinker, 1984, p. 29.

Updated assumption: Otherwise normal children can have language impairments (SLI); there is unexpected and unexplained variance across children Some relatively invariant properties of morphosyntax show unexpected individual variability

Variation in onset timing: Late activation of language acquisition mechanisms?

A late start for an intact language system versus a late start for an underspecified grammar

The value of the 3-group design: Affected, age-matched, languagematched Affected < Age matches = Language Impairment Affected < language-matched = Language impairment beyond general language delay

Variation in acquisition timing mechanisms for TNS, ages 3-8 years


children start later, and show slower acquisition timing although similar growth curves
SLI

Com posite T ense

Mean Percent Correct

100

80

60 5N 40 20 3 3;6 4 4;6 5 5;6 6 6;6 7 7;6 8 3N SLI

Age in Years

Young children show variation that disappears by age 5 years, at adult grammar

Composite TNS Scores Over Time

100 80

Score

60 40 20 0 3 3;6 4 4;6 5 5;6 6 6;6 7 7;6 8 3N SLI

Age in Years

SLI children show variation in a range far below age expectations

At the Same Time of Variation in TNSMarking, Other Morphology is Nonvariant


R g la -s p ra (sp n n o s d ta eu r lu ls o ta e u a )
10 0

Mean Percent Correct

8 0

6 0

4 0

5 N 3 N

2 0 3 3 ;6 4 4 ;6 5 5 ;6 6 6 ;6 7 7 ;6 8

SI L

A in Y ge ears

Lexical indices Show Consistent Variation Across the Growth Curve, and Do Not Differentiate SLI from Language-Equivalent Group

# Different Words # Verb Types # Verb Tokens % General All Purpose Verbs PPVT Raw Scores

Number of Different Word Roots


120 100

Score

80 60 40 20 SLI 3N 5N

Group

Verb Types

80 70 60

Score

50 40 30 20 10 Group SLI 3N

Time 1

Time 2

Verb Tokens
300 250 200

Score

150 100 50 0 Group SLI 3N

Time 1

Time 2

Percentage GAP Verb Use

.8 .7

Percentage

.6 .5 .4 .3 .2 .1 0.0 SLI 3N Group

Time 1

Time 2

PPVT Raw Scores


120 100

Score

80 60 40 20 0 3.0 4.0 5.0 6.0 7.0 8.0 5N 3N SLI

Age

Cross-clinical Comparisons as a Way of Unraveling the Relationship of TNS, MLU, Cognitive and Lexical Acquisition: A Comparison of SLI and WMS Children Matched for MLU, Disparate for Cognitive and Lexical Acquisition

MLU equivalent across SLI and WMS CA different between SLI and WMS WMS IQ < SLI WMS < SLI on number of words comprehended SLI < WMS on three measures of TNS: regular past ed 3rd person singular s BE copula and auxiliary

Mean Length of Utterance


8 7 6

MLU

5 4 3 2 1 SLI 3N 5N WMS

Group

Chronological Age by Group


150 125

Age in Months

100 75 50 25 0 SLI 3N 5N WMS

Group

IQ Levels by Group
140 120 100

Scores

80 60 40 20 SLI 3N 5N WMS

Group

PPVT-R Raw Scores


100

80

Scores

60

40

20 0 SLI 3N 5N W S M

G roup

Regular Past -ed, Spontaneous


1.0

Proportion Correct

.8 .6

.4 .2 0.0 SLI 3N 5N WMS

Group

Regular 3rd Person Singular -s, Spontaneous


1.0

Proportion Correct

.8

.6

.4

.2 0.0 SLI 3N 5N WM S

G roup

BE, Spontaneous
1.0

Proportion Correct

.8

.6 .4

.2 0.0 SLI 3N 5N WMS

Group

SLI = WMS on non-TNS morphemes


Plural s In/On

Plural -s, Spontaneous


1.0

Proportion Correct

.8

.6

.4

.2 0.0 SLI 3N 5N WMS

Group

In and On, Spontaneous


1.0

Proportion Correct

.8

.6

.4

.2 0.0 SLI 3N 5N WMS

Group

Working Conclusions

Outcomes are compatible with linguistic models of the adult grammar that posit a relatively discrete morphosyntax and TNS-marking as obligatory features of clause construction Evidence is compatible with a 2-phase maturational model, one that controls initial appearance of language (i.e., start-up) and another that controls certain grammatical properties versus general lexical growth and overall clause construction

Overall, support for an Extended Optional Infinitive period that co-exists with a generally slowed linguistic system in children with SLI Selective slowing of certain grammatical properties is evident in the obligatory properties of clausal structure

Implications for Genetic Studies

What elements of language acquisition are vulnerable to phenotypic variation onset? general delay? delay-withindelay? Is a grammatical marker age-dependent, or also evident in older children and adults?

Вам также может понравиться