Built-in modules vs a developmental Built-in modules vs a developmental process of gradual modularisation:process of gradual modularisation:

Insights from genetic disorders Insights from genetic disorders Annette Karmiloff-SmithAnnette Karmiloff-Smith

Developmental Neurocognition Lab, Centre for Brain & Cognitive Development, Developmental Neurocognition Lab, Centre for Brain & Cognitive Development,

Birkbeck, University of LondonBirkbeck, University of London

Paris Cosy - September 2007Paris Cosy - September 2007

Fogarty/NIH

PlanPlan1.1. Where do arguments for innate modularity stem from?Where do arguments for innate modularity stem from?

∙∙ Adult neuropsychologyAdult neuropsychology ∙∙ Evolutionary psychologyEvolutionary psychology ∙∙ Early infant competenciesEarly infant competencies ∙∙ Genetic disorders with Genetic disorders with unevenuneven cognitive profiles cognitive profiles (juxtaposition: preserved/impaired modules)(juxtaposition: preserved/impaired modules)

2.2. Why ‘preserved module’ unlikely:Why ‘preserved module’ unlikely: ∙∙ what we know about normal infant brainwhat we know about normal infant brain

3. The case of Williams syndromeThe case of Williams syndrome ∙∙ Domains of behavioural proficiency:Domains of behavioural proficiency: Face processing and language: preserved modules? Face processing and language: preserved modules?

4. The normal case: gradual localisation/specialisation 4. The normal case: gradual localisation/specialisation

5. WS: failure of brain localisation/specialisation5. WS: failure of brain localisation/specialisation

6. Importance of tracing development back to origins in infancy:6. Importance of tracing development back to origins in infancy: Cross-domain relationsCross-domain relations

Face Processing

Syntax

Number

Social cognition

Face Processing

Syntax

Number

Social cognition

Agrammatism

Prosopagnosia

Adult neuropsychology Adult neuropsychology

Social cognition

Face Processing

Syntax

Number

Infant start stateInfant start state

Face Processing

Syntax

Number

Social cognition

Adult end stateAdult end state

Assumptions about development

Data that seem to support the nativist assumptionData that seem to support the nativist assumption

• Evolutionary psychologyEvolutionary psychology

• Genetic disorders seeming to show similar modular deficits as Genetic disorders seeming to show similar modular deficits as those found in adult neuropsychological patientsthose found in adult neuropsychological patients

• Early competences in normal childrenEarly competences in normal children

““For instance, children with Williams syndrome have a barely For instance, children with Williams syndrome have a barely measurable general intelligence and require constant parental measurable general intelligence and require constant parental care, yet they have an exquisite mastery of syntax and vocabulary. care, yet they have an exquisite mastery of syntax and vocabulary. They are, however, unable to understand even the most immediate They are, however, unable to understand even the most immediate implications of their admirably constructed sentences.”implications of their admirably constructed sentences.” (Piattelli-Palmarini, 2001)(Piattelli-Palmarini, 2001)

““In sum, brain volume, brain anatomy, brain chemistry, In sum, brain volume, brain anatomy, brain chemistry, hemispheric asymmetry, and the temporal patterns of brain hemispheric asymmetry, and the temporal patterns of brain activity are all atypical in people with WS. How could the activity are all atypical in people with WS. How could the resulting system be described as a normal brain with parts intact resulting system be described as a normal brain with parts intact and parts impaired, as the popular view holds? Rather, the brains and parts impaired, as the popular view holds? Rather, the brains of infants with WS develop differently from the outset, which has of infants with WS develop differently from the outset, which has subtle, widespread repercussions…”subtle, widespread repercussions…” (Karmiloff-Smith, 1998)(Karmiloff-Smith, 1998)

““Autism is due to a deficit in an innately-specified module thatAutism is due to a deficit in an innately-specified module that

handles theory-of-mind computations only”handles theory-of-mind computations only” (Leslie, 1992)(Leslie, 1992)

“…“…a module that is localized in the orbito-frontal cortex”a module that is localized in the orbito-frontal cortex” (Baron-Cohen et al., 1999)(Baron-Cohen et al., 1999)

““Autism affects the interconnectivity among and within Autism affects the interconnectivity among and within various cognitive systems…..in autism, functional brain various cognitive systems…..in autism, functional brain development goes awry such that there is increased intra- development goes awry such that there is increased intra- regional specialization and less inter-regional interaction”regional specialization and less inter-regional interaction”

(Carpenter et al., 2001)(Carpenter et al., 2001)

““..examine the crucial role of unbalanced excitatory-inhibitory ..examine the crucial role of unbalanced excitatory-inhibitory networks…complex pathogenetic pathways…leading to ASD networks…complex pathogenetic pathways…leading to ASD through altered neuronal morphology, synaptogenesis and cell through altered neuronal morphology, synaptogenesis and cell migration”.migration”. (Persico & Bourgeron, 2006)(Persico & Bourgeron, 2006)

‘…‘…..overall, the ..overall, the genetic double dissociationgenetic double dissociation is striking…..The is striking…..The genes of one group of children [SLI] impair their grammar genes of one group of children [SLI] impair their grammar while while sparingsparing their intelligence; the genes of another group of their intelligence; the genes of another group of children [WS] impair their intelligence while children [WS] impair their intelligence while sparingsparing their their grammar.’ grammar.’ (Pinker, 1999, p. 262, (Pinker, 1999, p. 262, italics addeditalics added))

Claimed Claimed geneticgenetic double dissociation in double dissociation in developmental disordersdevelopmental disorders

Grammarnumber

Social Cognition

Face processing

Specific Language Impairment (SLI)Specific Language Impairment (SLI)

Williams syndrome (WS)Williams syndrome (WS) Face Processing

Grammar

NumberSocial cognition

The case of Williams syndromeThe case of Williams syndrome

D7S489AD7S1870

GT

F2I

RF

C2

CY

LN

2

GT

F2I

RD

1

NC

F1

WB

SC

R1/

E1f

4H

LIM

K1

EL

NC

LD

N4

CL

DN

3

ST

X1A

WB

SC

R14

WB

SC

R21

TB

L2

D7S2476

BC

L7B

BA

Z1B

FZ

D9

D7S489B

WB

SC

R5/

LA

B

WB

SC

R22

FK

BP

6

PO

M12

1

NO

LR

1

GT

F2I

RD

2

C CA BB A B A C

D7S489C

WB

SC

R18

Common WBS Deletion (~1.6Mb)

D7S2472

D7S613

DUPLICONS DUPLICONS

WS Critical Region: hemizygoticWS Critical Region: hemizygoticdeletion of ~ 28 genes on chromosome 7 @ q11.23deletion of ~ 28 genes on chromosome 7 @ q11.23

WS genotypeWS genotype

Claimed WS phenotypeClaimed WS phenotype

Preserved modules: Preserved modules: face processingface processing languagelanguage

Impaired modules: Impaired modules: visuo-spatial cognitionvisuo-spatial cognition numbernumber

SpatialSpatialCognitionCognition

LanguageLanguage Face Face ProcessingProcessing

Spatial Spatial CognitionCognition

Do significantly better scores in one domain Do significantly better scores in one domain indicate a ‘preserved module’?indicate a ‘preserved module’?

• Comparisons across domains are relative, not absolute

• What we know about normal brain

• Good behavioural scores might be reached by different cognitive/brain processes from the normal case

Why “preserved module” unlikely?Why “preserved module” unlikely?

Genetic mutation in WS: present from conception.Genetic mutation in WS: present from conception.Critical genes are expressed throughout cortex, so Critical genes are expressed throughout cortex, so effects of WS mutation will be widespread, not specific effects of WS mutation will be widespread, not specific to a single region of cortex.to a single region of cortex.

What we know about normal brain:What we know about normal brain:

• Cortex: highly interconnected in very young infant Cortex: highly interconnected in very young infant (Conel; Huttenlocher)(Conel; Huttenlocher)

• Ratio of white matter to gray matter: different in infancy Ratio of white matter to gray matter: different in infancy (Geidd)(Geidd)

• Corpus callosum: thickness of fibre bundles different in infancy Corpus callosum: thickness of fibre bundles different in infancy (Geidd)(Geidd)

• Early on, widespread activity in response to faces or to language:Early on, widespread activity in response to faces or to language: across multiple regions of cortex in both hemispheres across multiple regions of cortex in both hemispheres (Casey,Neville, Johnson)(Casey,Neville, Johnson)

• Subsequent pruning in normal development: Subsequent pruning in normal development: gradualgradual specialisation/localisation of function specialisation/localisation of function (Johnson, Rakic)(Johnson, Rakic)

Example of face processingExample of face processing

Progressive Progressive modularisationmodularisation(localisation, specialisation and relative encapsulation of function)(localisation, specialisation and relative encapsulation of function)

andand lack thereof in WSlack thereof in WSdespite good behavioural scoresdespite good behavioural scores

Progressive modularisation Progressive modularisation of face processing in of face processing in normalnormal infants over developmental time infants over developmental time (first 12 months and beyond) (first 12 months and beyond) (2 decades of research by Johnson, de Haan, de Sch(2 decades of research by Johnson, de Haan, de Schoneonen, Simion and others)n, Simion and others)

Typically developing infantsTypically developing infants

6 mo 12 mo adult6 mo 12 mo adult………

Different labs (Benton/Rivermead): Different labs (Benton/Rivermead):

WS good at face processing: WS good at face processing: ‘in the normal range’‘in the normal range’

““Preserved” face processing module in WS?Preserved” face processing module in WS?

WS behavioural data

Kate Humphries

Dagmara Annaz

Michael Thomas

Karmiloff-Smith, et al., 2004Karmiloff-Smith, et al., 2004

Inversion effect (hallmark of Inversion effect (hallmark of configural configural processing) processing)

doesn’tdoesn’t emerge developmentally in WS. emerge developmentally in WS.

CognitiveCognitive processes underlying good processes underlying good

behaviouralbehavioural scores: same as normal? scores: same as normal?

Grice et al., 2001, 2003Grice et al., 2001, 2003

What about WS brain?What about WS brain?Mark Johnson & Gergo Csibra

Sarah GriceGergo CsibraMark Johnson

Michelle de Haan

WS adolescent inWS adolescent inGeodesic HD-ERP netGeodesic HD-ERP net

Healthy controls:Healthy controls:

Progressive restriction of input typeProgressive restriction of input type

WS: lack of modularisation of function over timeWS: lack of modularisation of function over time

WS: failure to specialiseWS: failure to specialise

Healthy controls:Healthy controls:Progressive restriction of brain localisationProgressive restriction of brain localisation

WS: failure to localiseWS: failure to localiseWS

Controls

Controls

WS adults

Face processing:Face processing:Space processing:Space processing:Sound processing:Sound processing:

))) all processed more featurally ) all processed more featurally ) than configurally) than configurally

Williams syndrome: not only facesWilliams syndrome: not only faces

Y

Y

Y Y Y Y Y Y

Y

Y

Y Y Y

Y

ModelModel WS copyWS copy

Mayada Mayada ElsabbaghElsabbagh

Note change-YNote change-YContour change-NContour change-N

Gamma-band bursts:Gamma-band bursts:integration/binding of featuresintegration/binding of features

Sarah GriceGergo CsibraMark JohnsonMike Spratling

Atypical brain Atypical brain function in both function in both syndromes,syndromes,but cross-syndrome but cross-syndrome brain difference brain difference

Rethink notion of Rethink notion of “featural” at cognitive“featural” at cognitivelevel…..level…..

Not: impaired spatial module versus preserved face processing module

Spatial CognitionFace Processing

Behavioural scores “in normal range”: underpinned by different cognitive/brain processes

Common featural processing bias for processing: spatial stimuli facial stimuli auditory stimuli

Lack of modularisation in developmental Lack of modularisation in developmental

disorders disorders common initial processes common initial processes

before, in normal development, these would before, in normal development, these would

have become increasingly segregated have become increasingly segregated

What about the other domain ofproficiency in WS?

Claims of preserved language module in WS?

Why WS language so delayed?(often not until 5th/6th year…. and as delayed as DS)

Late maturing module?

Or developmental explanation?

WS infants, toddlers and children: WS infants, toddlers and children: -extremely delayed in onset of babbling-extremely delayed in onset of babbling

-extremely delayed in segmenting speech stream -extremely delayed in segmenting speech stream

-rely more on perceptual cues than linguistic labels -rely more on perceptual cues than linguistic labels

-production precedes pointing-production precedes pointing

-comprehension doesn’t show normal advance over production -comprehension doesn’t show normal advance over production

-comprehension in WS infants/toddlers as delayed as in DS-comprehension in WS infants/toddlers as delayed as in DS

-don’t use or follow eye gaze for referential communication, -don’t use or follow eye gaze for referential communication,

despite fascination with faces (dyadic vs triadic joint attention)despite fascination with faces (dyadic vs triadic joint attention)

-don’t understand referential function of pointing -don’t understand referential function of pointing

-auditory perception follows atypical developmental pathway -auditory perception follows atypical developmental pathway

No single explanation: No single explanation: allall contribute, in contribute, in complex interactions, to late onset and complex interactions, to late onset and atypical trajectory of WS languageatypical trajectory of WS language

Julia Grant

Sarah Paterson

Thierry Nazzi

Emma Laing

Gaia Scerif Mayada Elsabbagh

Is Is developmentaldevelopmental explanation of WS late language explanation of WS late language

confined to speech/language/communication?confined to speech/language/communication?

confined toconfined to same same domaindomain??

Target 1

Target 2

FixationJanice Brown

Mark Johnson & Rick Gilmore

Target 1

Target 2

Fixation

Body centered Retinocentric

Vector summation

Brown et al., 2003

DS/WS = different causes for similar language delayDS/WS = different causes for similar language delay

Visuo-spatial precursors to socio-communication innormal/DS infants and toddlers, impaired in WS

Basic deficits in visual system, Basic deficits in visual system,

early in WS developmental trajectory early in WS developmental trajectory

focus on features:focus on features:

cascading developmental effectscascading developmental effects over timeover time

across several across several emergingemerging higher-level higher-level

linguistic/cognitive systems.linguistic/cognitive systems.

Nature versus Nurture = Nature versus Nurture = false dichotomy!false dichotomy!

What about role of environment in dynamically What about role of environment in dynamically shaping developmental outcomes (genetic, brain,shaping developmental outcomes (genetic, brain, behavioural)? behavioural)?

But is environment same for typical/atypical But is environment same for typical/atypical development?development?

Need for detailed study of how having Need for detailed study of how having developmental disorder subtly developmental disorder subtly changes the environmentchanges the environment (social, e.g. language; motor, e.g. exploration)(social, e.g. language; motor, e.g. exploration) in which atypical infant/child develops.in which atypical infant/child develops.

Domain-specific approaches:Domain-specific approaches:Start state = domain-specific modules; core knowledge/domainsStart state = domain-specific modules; core knowledge/domains(e.g. Pinker, 1999; Spelke, 2005; Gelman, 2005)(e.g. Pinker, 1999; Spelke, 2005; Gelman, 2005)

Domain-general approaches:Domain-general approaches:Start state = single learning mechanism Start state = single learning mechanism (e.g. McClelland, 2005)(e.g. McClelland, 2005)

Domain-relevant approachesDomain-relevant approaches(Neuroconstructivism):(Neuroconstructivism):Start state = limited no. of domain-relevant biases (slight Start state = limited no. of domain-relevant biases (slight differences across cortex in brain chemistry, neuronal differences across cortex in brain chemistry, neuronal density, type/orientation of neurons, etc.). These initial biases density, type/orientation of neurons, etc.). These initial biases becomebecome domain-specific over developmental time; modules domain-specific over developmental time; modules emergeemerge developmentally developmentally from competition during ontogenetic from competition during ontogenetic process of gradual modularisationprocess of gradual modularisation (e.g. Karmiloff-Smith, 1992, 1998; (e.g. Karmiloff-Smith, 1992, 1998;

Elman, Bates, Johnson, Karmiloff-Smith, Parisi & Plunkett, Elman, Bates, Johnson, Karmiloff-Smith, Parisi & Plunkett, 1996)1996)

2 developmental processes2 developmental processes

1.1. Progressive modularization Progressive modularization

(specialisation/localisation and relative (specialisation/localisation and relative encapsulation of function)encapsulation of function)

2.2. Progressive explicitationProgressive explicitation

(RR – increasing role of language)(RR – increasing role of language)

Cognition Cognition not not built-in; gradual change over built-in; gradual change over developemntal time -> PLASTICITY FOR developemntal time -> PLASTICITY FOR

LEARNINGLEARNING

Both deficient in WS and manyBoth deficient in WS and manydevelopmental disordersdevelopmental disorders

Concluding thoughts…Concluding thoughts…

• Trace Trace full developmental trajectories full developmental trajectories back to origins in early infancy: developmentalback to origins in early infancy: developmental changes in brain activitychanges in brain activity• Focus also on domains of Focus also on domains of proficiency (“in the normal range”)proficiency (“in the normal range”): same behavioural : same behavioural scores -> scores -> differentdifferent brain and cognitive processes: not ‘preserved modules’ brain and cognitive processes: not ‘preserved modules’

• Adult neuropsychological models too static for developmental disorders; Adult neuropsychological models too static for developmental disorders; need to need to thinkthink developmentally + importance of timing across domainsdevelopmentally + importance of timing across domains: : adult endstate, if modular -> adult endstate, if modular -> emergent emergent from developmental process of from developmental process of specialisation, localisation and relative encapsulation: specialisation, localisation and relative encapsulation: not a not a statestate, but a developmental , but a developmental processprocess: modularisation: modularisation

• Infant cortex starts out Infant cortex starts out highly interconnectedhighly interconnected->progressive modularisation; ->progressive modularisation; atypical brains less pruning so atypical brains less pruning so remainremain moremore interconnectedinterconnected with widespread with widespread cortical activitycortical activity

• Don’t only seek dissociations; Don’t only seek dissociations; cross-domain associationscross-domain associations outside domain of outside domain of overt deficit;overt deficit;

• Choice is Choice is notnot between multiple innately-specified modules vs single domain-general between multiple innately-specified modules vs single domain-general learning mechanism.learning mechanism. Small number of Small number of domain-relevant biasesdomain-relevant biases compete (activity across whole brain) compete (activity across whole brain) until one wins out ->until one wins out ->becomesbecomes domain specific over timedomain specific over time Progreessive modularisation and explicitation of function Progreessive modularisation and explicitation of function

a.karmiloff-smith@bbk.ac.uka.karmiloff-smith@bbk.ac.uk

http://www.psyc.bbk.ac.uk/research/DNL/personalpages/annette.htmlhttp://www.psyc.bbk.ac.uk/research/DNL/personalpages/annette.html