“Kikker, waar ben je?” Narratives in Dutch speakers with Williams syndrome Barbara Schmiedtová Peter Indefrey Monique Flecken, Neeltje Verstegen Peter.

“Kikker, waar ben je?” Narratives in Dutch speakers with

Williams syndrome

Barbara SchmiedtováPeter Indefrey

Monique Flecken, Neeltje VerstegenPeter Hagoort

Dieke Lagers-vanHaselen & colleagues from Rotterdam

Erasmus Universiteit Rotterdam, Dept. of Neuroscience, July 14, 2005

2

Outline

I. Recapitulation of project’s goalsII. Populations

- examples: chat format

III. Coding procedures- examples of coding categories

IV. Results- comparison to English results

V. Discussion

3

General research question

How general intellectual impairment with a specific genetic basis may affect the process oflanguage acquisitionandnarrative development?

4

Focus on Narratives

• typical development– morphosyntactic structures acquired by age 5

(e.g. Gretsch 2001, Bowerman 1992, Slobin 1985)

BUT– application of morphosyntactic forms in

discourse, that is considering information & temporal structure, much later (Halm 2005, age 14)

5

Dissociation between language and cognition

• Williams syndrome– distinct behavioral phenotype (Bellugi et at. 2000,

1999)

– mild to moderate retardation: IQ 55 (rarely reaching above 80)

– productive language skills preserved (in contrast to other genetic syndromes with mental retardation, e.g. Down Syndrome)

– extremely social and outgoing personality

6

Populations

• PatientsWilliams syndrome (N= 51)

Pseudo-Williams (N=8)

• Control groupsTypically developing children (N=49)

- IQ matched sample: 5-6 year olds

Adults (N=72)- Age matched sample: 15-18 year olds

7

Transcription & Coding

Patients: 2460 propositions

Adults: 4032 propositions

Children: 2254 propositions

Total: 8746 propositions

8

Coding procedures

• coding scheme (Reilly et al. 2004, 1998)

– 4 domains:

1. the overall story length

2. grammatical competence

3. narrative skills (episodic & thematic)

4. evaluation devices

9

1. Story Length

• In general, narratives vary in length => number of propositionsa proposition = predicate and its arguments

Coding criteria:1. Same subject?2. Two separate sentences?3. Do two clauses realize one/two events?

10

Number of propositions - examples

• One proposition“De jongen kruipt in de boom om te zien …..” (prop1)

• Two propositions “De volgende ochtend zien ze dus (prop1)

dat de kikker weg is.” (prop 2)

11

2. Grammatical Competence

A) Morphological Errors, e.g. (27 categories in total)

• Errors in pronouns (*hem heeft de kikker gevonden)• Omissions of auxiliaries and/or verbs (*kikker

weggelopen)• Verb tense (*toen loopt…)• Omission of determiners and gender errors (*het

jongetje, die…)• Noun plurals (*de bijs)• Number marking (*de jongen en de hond loopt)• Prepositional errors (*roepen achter)• Omission of ‘er’ / ‘te’• Word order errors• Omission of verbs/ non-existing words / too many words

12

2. Gram. Competence - cont’

B) Occurrences of complex syntax

• Coordinate sentences (maar, en, of)• Adverbial sentences (toen, want)• Verb complements (zien dat + X, proberen te +

V, beginnen te + V)• Relative clauses• Passives

13

3. Narrative skills

• Thematic (0-4)– Maintenance of story theme

• Frog missing & boy looking for him (0-2)

• Theme reiterated throughout the story0 - no additional mentions;2 - multiple additional mentions

14

3. Narrative skills – cont’

• Episodic (0-8)– Basic components of the story

1. Settings

2. Instantiation

3. – 7. The five main search episodes

8. The story’s resolution

15

4. Evaluation devices

1. Cognitive inferences (“De jongen denkt …”)

2. Social engagement(“Oooh, wat gebeurt er nou weer?”)

3. Affective states and/or behaviors(“De jongen mist zijn kikker, wat zielig!”)

4. Intensifiers + repetitions(“De jongen roept heel erg hard”)

5. Indication of certainty/uncertainty (“Misschien zit de kikker wel in die boom”)

16

Analyses

• Comparisons– Williams syndrome (WMS) vs. typically developing

children (TD) vs. adults (AD)– In some cases Pseudo-WMS subjects included– Dutch (WMS vs. TD) & English (WMS vs. TD)

• Compensation for the length of the story– Number of countable categories divided by

number of propositions

17

English Data Set

• Reilly et al., Brain and Language (88),2004, 229 - 247– Main focus on development in children– 3 different age groups – Comparisons

• SLI vs. TD• WMS vs. TD• SLI vs. TD• FL (unilateral focal brain lesions) vs. TD• SLI vs. FL

18

1. Length of Story:number of propositions

0

10

20

30

40

50

60

Nr. of Propositions

Me

an

Nr.

of

oc

cu

rre

nc

es

AdultsChildrenWMSPseudo

*

p < .05

English Data Set: no difference between TD vs. WMS

ENG NL

19

2. Grammatical skills: morphological errors

0

10

20

30

40

50

60

70

Morphological ErrorsMe

an

Nr.

of

Mo

rph

olo

gic

al E

rro

rs

AdultsChildrenWMSPseudo

p < .05

*

English Data Set: strong difference between TD vs. WMS

ENG NL

20

2. Grammatical skills:complex syntax

0

10

20

30

40

50

60

70

80

Complex Syntax

Me

an

of

Nr.

of

oc

cu

rre

nc

es

AdultChildrenWMSPseudo

p < .05

*

ENG NL

English Data Set: strong difference between TD vs. WMS

21

3. Narrative structure:Theme Maintenance

0

1

2

3

4

Theme Maintenance (0-4)

Me

an

Sc

ore Adults

ChildrenWMSPseudo

p < .05

*

English Data Set: strong difference between TD vs. WMS

ENG NL

22

3. Narrative structure:Story components

0

1

2

3

4

5

6

7

8

Story components (0-8)

Me

an

Sc

ore Adults

ChildrenWMSPseudo

p < .05

*

*

English Data Set: strong difference between TD vs. WMS

ENG NL

23

4. Evaluation devices

05

101520253035404550

Evaluation Devices

Me

an

of

Nr.

of

oc

cu

rre

nc

es

AdultsChildrenWMSPseudo

p < .05

*

*

English Data Set: strong difference between TD vs. WMS

ENG NL

24

Summary:Dutch vs. English WMS

No differences– Nr. of propositions– Grammatical

competence• Morphological errors• Complex syntax

– Narrative skills• Story components

Differences– Narrative skills

• Story maintenance

WMS groups together with TD children

Age related?

25

Dutch WMS: age groups

32%

32%

36%

young:< 12middle: 12 - 19 older: ≥ 20

26

Dutch WMS: Age & Theme maintenance

0

1

2

3

4

Theme maintenance

Me

an

Sc

ore

YoungMiddleOlder

p < .05

* young vs. middle, young vs. older; n.s. middle vs. older

*

27

Dutch WMS: Age & Story components

0

1

2

3

4

5

6

7

8

Theme maintenance

Me

an

Sc

ore

YoungMiddleOlder

* young vs. middle, young vs. older; n.s. middle vs. older

*

p < .05

28

Dutch WMS:Age & Number of propositions

0

10

20

30

40

50

60

Number of Propositions

Me

an

Nr.

of

oc

cu

rre

nc

es

YoungMiddleOlder

* young vs. middle, young vs. older; n.s. middle vs. older

*

p < .05

29

Dutch WMS:Age & Evaluation Devices

0

10

20

30

40

50

60

70

Evaluation Devices

Me

an

Nr.

of

oc

cu

rre

nc

es

YoungMiddleOlder

* young vs. middle, young vs. older; n.s. middle vs. older

*

p < .05

30

Dutch WMS:No changes with Age

• WMS ”weak“ points– Grammatical competence

• In all age groups, a very high number of morphological errors

• Complex syntax: stable number of syntactic structures (=> no increase in complexity when the story length goes up with age!)

31

Summary: age in WMS

Developmental changes in WMS: Critical period between the age of 12 and 19

• Changes inMaintenance of Theme - better

Story length – stories longer (BUT not more complex)

Evaluation Devices – fewer

32

Dutch WMS:Age & Blokpatronen (BP)

0

5

10

15

20

25

BP Score

Mea

n S

core

YoungMiddleOlder

* young vs. older; n.s. young vs. middle & middle vs. older

*

p < .05

33

Dutch WMS:Age & Boston Naming Test (BNT)

0

10

20

30

40

BNT Score

Mea

n S

core

YoungMiddleOlder

*

p < .05

* young vs. middle, young vs. older; n.s. middle vs. older

34

Dutch WMS:Age & Peabody Picture Vocabulary Test (PPVT)

0

20

40

60

80

100

120

140

PPVT

Mea

n S

core

YoungMiddleOlder

*

p < .05

* young vs. middle, young vs. older; n.s. middle vs. older

35

Summary: development in Dutch WMS & other tests

Development in all three investigated tests• Blokpatronen• Boston Naming Test• Peabody Picture Vocabulary Test

36

Future plans

• More correlations between the Frog- story-measurements & other tests

• Investigation diversity in each category (token/type ratio)

37

THANK YOU!