using hungarian language to clarify language-thought relations in impaired populations csaba pléh...
Post on 16-Dec-2015
224 Views
Preview:
TRANSCRIPT
Using Hungarian language to clarify language-thought relations in impaired
populations Csaba Pléh
and Ágnes Lukács Department of Cognitive Science
Budapest University of Technology and Economics and HAS-BME Research Group of Neuropsychology and Psycholinguistics
Talk at the symposium on Williams Syndrome
Fonyód, Hungary, June 25th 2005
Outline
• The logic of studying genetic disorders of cognition
• Williams syndrome as a favorit dissociative disorder
• Possible role of interlanguage comparions• Studies on Hungarian in WS subjects• Supporting spatial language problems, but
rejecting a specific disorder, questioning the simple intact grammar versus impaired lexicon dissocation
Disordered populations and cross-linguistic comparisons
• If there is a genetic disorder, it should be manifested in the same way in all languages and cultures
• Crosslingustic comparisons are still relevant:I. They allow to separate factors that are tied in one
language. E.g. irregularity and frequency in English morphology.
II. They may help to support epigenetic theories that emphasize the complex pathways leading to disturbed cognition
The interest towards Williams syndrome
• Promises to help understand the genes-brain-cognition chain
• Clear dissociations proposed in cognition:
WS social - autistic asocial
WS localistic - Down syndrome holistic
WS good language – SLI weak language
WS cognition: Strengths and weaknesses
• Strengths• Good social skills• Relative good
language• Grammar good in
language • Musicality
• Weaknesses• Low IQ 60-80• Weak visuo-spatial
cognition• Lexicon and
knowledge weak• Hyperacusia
Language in Williams syndrome: Theories and the Hungarian data
• Studies and theories
• Frequency is not relevant in the lexicon
• The mental lexicon is atypically organized
• Grammatical rules seem to be intact (Pinker, Clahsen)
• The language of space is especially weak
• Hungarian „test”
• Frequencies and individual differences
• More categories and category fluencies
• More stem classes and frequencies used
• Weakness, but no qualitative differences
Language profile in the Hungarian WS project (Lukács és Pléh)
WS level WS pattern
1. Vocabulary < +
2. Semantic fluency + +
3. Pragmatic cues in lexical learning szótanulásban
+ +
4. TROG < +
5.Rules and exceptions in morphology
< +
6.Grammaticality judgement
< -
7. Anaphora + +
8. Spatial expressions production and comprehension
< +
9. Sentence completion + +
Lexical frequency effect: Present
010
2030
4050
6070
8090
100
Frequent Rare
WS
Age control
Voc Control
0
100
Frequent Rare All
Large individual differences: Knowledge of rare items in WS
children is a function of digit span
Low
High
Effects of verbal STM spanFrequent: F(1,13)=1.35, n.s.Rare:(1,13)=13.13, p<.005
Threshold effects: in controls, verbal working memory correlations nin significant in controls
Vocabulary measure WS Control
N, Freq ,51 * ,23
N, rare ,63 * ,27
V, freq ,43 ,22
V, rare ,48 * ,16
Comp, Freq ,61 * ,25
Comp, rare ,47 * ,14
Semantic fluency and supposed iddyosynchretic organization
WS subjects and controls matched on verbal age in a category fluency task
0
2
4
6
8
10
12
14
anim
als food
drinks
music
al in
strum
ents
cloth
es
furn
iture
thin
gs to
read jo
bs
WSVC
`
Comparisons based on category norms of Kónya & Pintér (1986)
No systematic differences between the two groups in the frequency of items they produced
WS produced less frequent musical instrument names
No systematic differences between the two groups in the average rank of items they produced
WS subjects had higher scores (i.e. pruduced items appearing later in the original lists) in clothes and musical instruments
Morphological irregularity Stimuli in the morphology task
Regularcipő-cipők ‘shoe’-’shoes’
Irregularbagoly-baglyok‘owl’-’owls’
Presupposed dissociation is missing: Rules vs. Items
0
10
20
30
40
50
60
70
80
90
100
Reg Freq Reg Rare Except Freq Except Rare
Vcontrol
WS
Qualitative comparison of errors
• If anything, more overgeneralization in controls
0
20
40
60
80
100
120
WSExcep
WSReg
ContExcep
ContReg
Overgener
Other
Spatial versus nonspatial morphology in WS: As expected, spatial language is
impaired
0
20
40
60
80
100
spatialsuffix
spatialpostposition
nonspatialsuffix
VCSCWMS
The language of space in Hungarian
allows for qualitative comparisons • Obligatory distinctions
along the path
• Three markers for GOAL, SOURCE, LOCATION. Is there a difference?
WS is weaker in postposition production.
SOURCES are difficult for both groups
0
10
20
30
40
50
60
70
80
90
100
Static Source Goal
WS
VC
Less difference in comprehension than production, suffixes
0
0,5
1
1,5
2
WSprod WScomp VCprod VCcomp
Static
Source
Goal
Source difficulty
• Our data support Landau and Zukowski’s hypothesis: the difficulty with retaining information in memory can account for special difficulty with SOURCE paths.
• The pattern is similar to what we observe in typical development at earlier stages.
Comparing spatial and non-spatial uses in a repetition task
SPATIAL Az oroszlán megszökött a ketrecből.
The lion escaped the cage-ELA.The lion escaped from the cage.
NON SPATIAL Pisti tanult a balesetbőlPisti learnt the accident-ELA.Pisti learnt from the accident.
Possibilities here
• Non-spatial is relatively easier for WS subjects
• Non-spatial is more difficult for both groups
• Differences diminish since there is no need for referntial coding of space
Non spatial is weaker for both groups
0
10
20
30
40
50
60
70
80
90
100
STATIC spatial SOURCE spatial GOAL spatial STATICnonspatial
SOURCEnonspatial
GOALnonspatial
% c
orre
ct WS
VC
Conclusions: Why was it relevant to do studies in Hungarian?
• Frequencv is a factor in WS language, and memory is an important mediating variable
• WS shows no clear support for the intact rules – impaired lexicon model
• Spatial language is impaired in WS, but the patterns is the same as in typical development
• Difficulties with spatial language in WS reflect their problems in spatial cognition
top related