E L E N I B A L D I M T S I 1, E L E N I P E R I S T E R I 1, & I A N T H I T S I M P L I 1 , 2

1A R I S T O T L E U N I V E R S I T Y O F T H E S S A L O N I K I ,2U N I V E R S I T Y O F R E A D I N G , U K

An investigation of oral narratives in

children with High Functioning

Autism: aspects of microstructure and

macrostructure

13th International Congress for the Study of Child

Language, 14-18 July 2014

Aims

to compare the narrative performance of children

with High Functioning Autism (HFA) to that of

typically-developing (TD) children in terms of

referential cohesion and global narrative coherence

to examine whether lexical and syntactic abilities

correlate with cohesion and coherence in narrative

discourse

HFA

Individuals with HFA belong to the autism spectrum

with no learning disability (Frith, 2004)

Basic language functioning (phonology, syntax) is

age-appropriate while higher level linguistic abilities

(e.g. use of language in context, pragmatic

inferencing, intentionality) are inappropriate (Joliffe

& Baron-Cohen, 2000; Tager- Flusberg, Paul & Lord,

2005)

Verbal and non-verbal IQ ≥85

Narratives as linguistic and cognitive markers

of development

Linguistic development: reference, coordination,

subordination, temporal anchoring

Cognitive (and pragmatic) development:

constructing the episode(s) and building a mental

model of the narrative(Johnson-Laird, 1983; Bower & Morrow, 1990; Arnold, Benetto &Diehl, 2008)

Central coherence - Theory of Mind

A. Referential cohesion

The development of narrative ability is related to the development of character reference: Introduction, Maintenance, Shift (Reintroduction) (Hickman et al 1995; 1996;

Hickman & Hendriks, 1999)

Reference use in autism

Underspecification: Children with autism overused pronouns and

zero-pronouns where full DPs should have been used in shift

contexts ((Norbury & Bishop, 2003; Norbury et al., 2013;Tager- Flusberg, 1995)

Overspecification: children with autism overused full DPs

expressions instead of pronouns in maintenance (Baltaxe, 1977;

Colle, Baron- Cohen, Wheelwright & Van der Lely, 2008; Arnold et al., 2009)

Why?

A. Memory problems (maintaining activation in cases of overload)

B. Taking the listener’s perspective into account (ToM)

B. Global narrative coherence

holistic organization of discourse,

use of causal explanatory frameworks to integrate

narrated events in meaningful ways,

and referential cohesion(Hogan-Brown, Losh, Martin, &. Mueffelmann, 2013 )

Individuals with autism tend to use fewer

evaluatives (i.e. characters’ thoughts & emotions),

fewer causal networks and more irrelevant,

inappropriate utterances while narrating a story (Norbury et al., 2013; Diehl et al., 2006)

20 Greek-speaking children diagnosed with HFA (1

girl, age range: 7;0-12;6, Mean: 9;4 yrs., SD: 1.8)

Diagnostic criteria:

Autism Diagnostic Interview-Revised (ADI-R; Lord,

Rutter, & Le Couteur, 1994)

clinical assessment of the child’s social-adaptive

functioning by a child psychiatrist

20 typically-developing (TD) Greek-speaking

children (1 girl, age range: 7;0-12;5, Mean: 9;4 yrs.,

SD: 1.78)

I. The study - Participants

- Morpho-Syntactic Comprehension (Diagnostic

Verbal IQ (DVIQ); Stavrakaki & Tsimpli, 1999)

Greek Expressive Vocabulary task (Vogindroukas,

Protopapas, & Sideridis, 2009; adaptation from Renfrew

Word Finding test, 1995).

II. The study –Methodology

Screening tasks

Results (WISC-III, DVIQ, Vocabulary), Means (SD)Groups WISC

(general IQ index)

DVIQ score Expressive

Vocabulary

(max. score: 50)

TD control

Children 120.6

Mean verbal IQ: 120.65 (16.4)

Mean performance IQ: 115.7

(15.4)

17.1 (3.9) 45.1 (3.7)

MEVA: 11;7

Children with

HFA

103.9

*Mean verbal IQ: 103.8 (17.3)

**Mean performance IQ: 103.2

(15.4)

15.0 (3.8) 42.4 (5.0)

MEVA: 10;7

*verbal IQ index: 9 children with HFA scored sign. lower

(p<.05) than the mean of the control children

**performance IQ index: 7 children with HFA scored sign.

lower (p<.05) than the mean of the control children

(Singlims_ES software; Crawford et al., 2003)

II. The story

Wordless storybook: ‘Harry the dirty dog’ (Zion, 1956)

Telling: all pictures were presented initially and then

page by page in a book form; the experimenter was

watching

Retelling: the child listens to the story through

headphones; the experimenter was sitting opposite

the child without looking at the book; the child had to

retell the story looking at the pictures again

“Harry the dirty dog” picture-story

II. a.Microstructure

Lexical Diversity of Internal State terms (emotions, beliefs,

attitudes, … expressed through different categories)

Syntactic Complexity Index (Hunt, 1970): No of subordinated sentences

for every c-unit

II.b. Macrostructure

-No of Episodes

-Setting

-Closing events: statements that indicate goal attainment/ failure of the

protagonists actions

-Accuracy: theme & point of the story

-Causal connectivity: the average number of connections between

propositions

Causal Networks (Trabasso & Sperry, 1985)

In text cause-and-effect connections between events or linguistic

units create a network of information.

This network identifies the causal chain, the events that create the

gist of the story (Trabasso & Sperry, 1985)

.

Causal Networks (Trabasso & Sperry, 1985)

Causal connections in the story “Harry the dirtydog” between 5 events:

1.Harry was a white dog with black spots

2. who liked everything, except… getting a (2)bath.

3. So one day when he heard the water running in the tab, (3)

4.he took the scrubbing brush… and buried it in the backyard

5.Then he ran away from home.

Results

I. Frequencies of use of referential forms in

telling and retelling in Maintenance

4,3

14,2

93,95

85

0

10

20

30

40

50

60

70

80

90

100

TD children HFA children

Lex. NPs

Pronoun

14,89 13,81

87,4484,74

0

10

20

30

40

50

60

70

80

90

100

TD children HFA children

Lex. DPs

Pronouns

Telling Retelling

- Pronouns > Lex. DPs in both telling and retelling (p<.05) for both groups

- No significant between-group differences in either telling or retelling

I. Frequencies of use of referential forms in

telling and retelling in Reintroduction

17

8,1

47,83

42,4

0

10

20

30

40

50

60

70

80

90

100

TD children HFA children

Lex. DPs

Pronouns47,54 47,78

43,33

35,55

0

10

20

30

40

50

60

70

80

90

100

TD children HFA children

Lex. DPs

Pronouns

Telling Retelling

- Pronouns > Lex. DPs in telling only (p<.05) for both groups

- Non significant between-group differences in both modes

- Lex. DPs sign. higher in retelling (p<.010) for both groups (retelling effect)

II. Narrative length and

Lexical Diversity of IST

Telling Retelling

- TD children: higher rates for both C-units and lexical diversity in retelling (p<.005)

- Children with HFA: higher rates for lexical diversity in retelling (p=.013)

- Significant between-group differences in both C-units and Lexical Diversity in

Retelling

(TD children > children with HFA)

III. Microstructure:

Syntactic Complexity Index

1,291,23

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

TD children HFA children

Syntactic Complexity Index

SyntacticComplexity

Index

1,491,41

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

TD children HFA children

Syntactic Complexity Index

SyntacticComplexity

Index

Telling Retelling

- TD children: higher Syntactic Complexity Index in Retelling

(p=.013)

- n.s. between-group differences

IV. Macrostructure measures

Telling Retelling

- TD children: narrative mode effect; higher rates for Settings in Retelling (p=.025)

- Children with HFA: narrative mode effect; higher rates for No of Episodes (p=.034) and

No of Peripheral Events (p=.046) in Retelling.

- Telling: TD > children with HFA in Causal Connectivity, Accuracy, No of Episodes and

Peripheral Events

- Retelling: TD > children with HFA in all 5 categories

Syntactic Complexity Index and causal connectivity

• HFA children

Syntactic Comprehension (DVIQ) was positively correlated with the Syntactic Complexity index in retelling (F(1, 19)=4.003, p=.051, r=.182).

Expressive Vocab was positively correlated with Causal connectivity in the telling condition (F(1, 19)=4.113, p=.058, r=.186).

Summary

Referential Cohesion: no significant between-group

differences in the use of appropriate referential forms per

referential function

Microstructure: children with HFA produced shorter

narratives and less lexical diversity in the use of internal

state terms

Macrostructure: children with HFA had less structured causal

networks than the TD children-->story structure and

coherence were problematic for the children with autism

Narrative mode effects: retelling had more marked effects

on TD children than children with HFA in terms of number of

C-units, Syntactic Complexity and Settings as part of causal

networks

Remaining questions

To what extent can differences in language ability in children with

HFA vs. TD account for problems in encoding macrostructure?

Many thanks to the children participating in the study and their parents

And thank you for your attention!

Selected References

American Psychiatric Association. (2013). Diagnostic and statistical manual ofmental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Arnold, J.E., Bennetto, L.,& Diehl, J.J.(2009).Reference production in youngspeakers with and without autism: Effects of discourse status and processing constraints.Cognition, 110, 131–146.

Baltaxe, C.A.M.,&D’Angiola, N.(1992). Cohesion in the discourse interaction ofautistic, specifically language-impaired, and normal children. Journal of Autism andDevelopmental Disorders,22, 1–21.

Bamberg,M.,&Damrad Frye,R.(1991).On the ability to provide evaluativecomments: Further explorations of children’s narrative competencies. Journal of ChildLanguage,18,689–710.

Baron-Cohen, S.(1995).Mindblindness: An essay on autism and theory of mind.Cambridge, UK: The MIT Press.

Diehl, J. J., Bennetto, L., & Young, E. (2006). Story recall and narrative coherenceof high-functioning children with autism spectrum disorders. Journal of Abnormal ChildPsychology, 34(1), 87–102.

Selected References (ctd.)

Hunt, K. W. (1970). Recent measures in syntactic development. In M. Lester (ed.)

Reading in applied transformational grammar (179-192). Nueva York: Holt, Rinehart and Wiston.

Frith, U. (2003). Autism: Explaining the enigma. Oxford: Blackwell.

Joliffe, T., & Baron-Cohen, S. (2000). Linguistic processing in high-functioningadults with autism or Asperger syndrome: Can global coherence be achieved? A further testof central coherence theory. Psychological Medicine, 30, 1169- 1187.

Norbury, C.F., Gemmel,T., &Paul, R.(2013). Pragmatic abilities in narrative production:across-disorder comparison. Journal of Child Language Advance online publication.

Stavrakaki, S. & Τsimpli, I. M. (2000). Diagnostic verbal IQ test for school andpreschool children: Standardization, statistical analysis, and psychometric properties.Proceedings of the 8thconference of the Panhellenic Association of Speech and LanguageTherapists (pp. 95-106). Athens: Ellinika Grammata.

Schuh, J. M., & Eigsti, I. M. (2012) .Working Memory, Language Skills, andAutism Symptomatology. Behavioral Sciences, 2 (4), 207-218.

Trabasso, T., & Sperry, L. L. (1985). Causal relatedness and importance of story events. Journal of Memory and Language, 24, 595–611.

Vogindroukas, I., Protopapas, A. X., & Sideridis, G. (2009). Greek version of the

Word Finding Vocabulary Test (Renfrew, 1995). Athens: Glauki.

Peripheral events of the story

1.Harry was a white dog with black spots

2. who liked everything, except… getting a bath. Setting

1st. Harry is avoiding the bath

2nd. Harry’s adventure w/ sub- episodes

3rd. Harry’s trouble back home

4rth. Finding the brush No of episodes

5fth. Towards the bath tub

6th. Taking a bath

7th. Bach home

1. Harry’s bath was the soapiest one he’d ever had. It worked like magic.

As soon as the children started to scrub, they began shouting, “Mummy! Daddy! Closing

Look ! Come quick! It’s Harry! It’s Harry! It’s Harry!” they cried. events

2. Harry wagged his tail and was very, very happy.

3. It was wonderful to be home

Accuracy

“Harry the dirty dog” story is that it is about a white dog

with black spots who loves everything , except baths.

So one day before bath time, Harry runs away. THEME

He plays outside all day long, digging and sliding

in everything from garden soil to pavement tar.

By the time he returns home, Harry is so dirty he looks like

a black dog with white spots. His family doesn't even

recognize him!

Of all things, Harry has to ask for a bath and when finally

gets it POINT

then his family recognizes him.