PHOTOS

EARLY COGNITIVE DEVELOPMENT  AND ITS IMPAIRMENTS       Workshop at Collegium Budapest  December 21, 2001         Budapest, I. Szentháromság u. 2.

All talks are 30 minutes plus 10 minutes discussion.

Morning chair: Ilona Kovács

9:00 Opening

9:10 Csaba Pléh, Ágnes Lukács and Mihály Racsmány: Language in Williams syndrome: morphology and the mental lexicon in acquisition

9:50 Ildikó Király, Mihály Racsmány, Ágnes Lukács and Csaba Pléh: Spatial cognition and spatial language in Hungarian normal children and in WMS subjects

10:30 coffee break

11:00 György Gergely and Gergely Csibra: The Teleological Stance in infancy and its relation to naïve theory of mind

11:40 Miklós Gyõri and Krisztina Gy. Stefanik: Roundabouts of ‘theory of mind’ in autism: possible pathways, actual mechanisms

12:20 Zvia Breznitz: A Comparison between Young and Adult Dyslexics

13:00 lunch

Afternoon chair: Csaba Pléh

14:00 József Fiser:  The statistical nature of visual feature development in adults and infants

14:40 Zsuzsa Káldy and Alan Leslie: Identification of objects in 9-month-old infants: integrating what and where information.

15:20 coffee break

15:50 Daphne Bavelier: Cortical Reorganization of Visual and Language
functions after Early Auditory Deprivation

16:30 Gergely Csibra: Perception of illusory contours in infancy

17:10 Ilona Kovács and Peter Gerhardstein: Detection of contour continuity and closure in 3-month-old human infants
 
 

ABSTRACTS
 

CORTICAL REORGANIZATION IOF VISUAL AND LANGUAGE FUNCTIONS AFTER EARLY AUDITORY DEPRIVATION

Daphne Bavelier
Brain and Cognitive Sciences, Radiology, and Center for Visual Science, University of Rochester, USA, daphne@bcs.rochester.edu
 

Studies of adults who have altered sensory and language experience, such as congenitally deaf individuals, suggest that different brain systems within vision and within language display different degrees of experience-dependent modification. Within vision, the organization of systems involved in processing peripheral space and in sustaining spatial attention is most altered following auditory deprivation. Within language, altered language experience such as the exposure to a visuo-manual language (American Sign Language) does not alter bias of the left hemisphere to process natural languages. However, in contrast to English, ASL strongly recruits the right hemisphere indicating that the specific processing requirements of the language can also influence the organization of the language systems of the brain. Hypotheses concerning the origins of the differential effects of early experience will be discussed.

ASYNCHRONY BETWEEN COMPONENTS OF INFORMATION PROCESSING AS A POSSIBLE EXPLANATION OF DYSLEXIA

Zvia Breznitz
Neurocognitive Laboratory, University of Haifa, Israel
breznitz@colbud.hu
 

Recent studies argue that the dyslexia phenomenon stems from lower level processing. However, research in this
line has been carried out mainly on the young population, and has raised the question whether the characteristics
of the young dyslexic are similar to those of the adult dyslexic.  This research examined the differences of lower
level visual and auditory as well orthographic and phonological processing among young and adults dyslexic
readers and their age-matched controls by utilizing behavioral and electro-physiological measures. Results
indicate that young subjects where significantly less accurate and slower in most of the tasks as compared to
adults. Young dyslexics were significantly slower and less accurate than their age-matched group in the
visual-orthographic and the auditory-phonological tasks.  No significant differences were found within the adult
groups in the accuracy of task experimental tasks. However, at all levels the adult dyslexics continue to process
information in significantly slower manners than do than the controls. This slowness appeared in reaction time as
well as in later latencies of P2 and P3 ERP's components. A potential theory with regard to the underlying factor
of dyslexia will be presented.
 

PERCEPTION OF ILLUSORY CONTOURS IN INFANCY

Gergely Csibra
Centre for Brain and Cognitive Development, Birkbeck College, London
g.csibra@bbk.ac.uk

The Kanizsa illusion is one of the most extensively studied visual illusions, perhaps because, unlike other illusions, it not just distorts an image but actually adds a further a visual object to the scene. Earlier studies with infants yielded controversial results on whether they can perceive the Kanizsa illusion. Recent findings suggest that although young infants can see the "good form" specified by the illusory contour, they do not see the Kanizsa figure as a spatially separated object in depth until they are 7 to 8 months of age. Our electrophysiological measures showed that, around the same age, the infant brain starts to respond to Kanizsa stimuli with a characteristic oscillatory burst in the gamma frequency band (40 Hz). Some aspects of this activity (e.g., latency) resemble to what was reported in comparable studies in adults, others (e.g. brain region), however, differ. Our finding suggests that the neural development around 6 to 8 months of age that allows infants to perceive static illusory objects in depth involves a decrease in variability of gamma range oscillatory activity in the frontal cortex.
 

THE STATISTICAL NATURE OF VISUAL FEATURE DEVELOPMENT IN ADULTS AND INFANTS

Jozsef Fiser
Center for Visual Science, Department of Brain and Cognitive Sciences, University of Rochester, USA

The question of how humans develop useful descriptions of complex visual environment surrounding them lies at the heart of intelligent behavior. Although the task can be formulated as a statistical pattern recognition problem, a longstanding criticism against such approach to the issue has been that the statistical complexity of the task rules out any plausible explanation based on the principles of statistical learning. I will present results from a number of experiments conducted with adults and infants showing that humans from an early age on are highly sensitive to statistical aspects of visual scenes. Moreover, the different statistics we are involuntarily extract from our visual environment provide a plausible setup for describing natural scenes without being overwhelmed by the complexity of the task.
 

THE TELEOLOGICAL STANCE IN INFANCY AND ITS RELATION TO NAIVE THEORY OF MIND

György Gergely
Research Institute for Psychology, Hungarian Academy of Sciences, Budapest, Hungary
gergely@cogpsyphy.hu

Gergely Csibra
Centre for Brain and Cognitive Development, Birkbeck College, London, UK
g.csibra@bbk.ac.uk

In this talk we shall summarize evidence from a series of visual habituation studies that indicate that 9- and 12-month-old (but not yet 6-month-old) infants can interpret the behaviour of computer animated figures as goal-directed action and can evaluate the rationality of the goal approach. We propose that in doing so they apply an interpretative representational system that we call the ’teleological stance’ or the infant’s ’naive theory of rational action’. This – as yet, nonmentalistic - interpretative system parses and represents intentional actions in terms of three representational elements: 1. Goal (the outcome state of the behaviour), 2. Reality Constraints (that constrain the set of possible actions leading to the goal in the particular situation), and 3. Action (the behaviour leading to the outcome state). This system relies on the abstract core ’principle of rational action’ that provide the well-formedness conditions of teleological representations. We demonstrate that the teleological stance is a productive interpretative system that allows infants to infer and predict novel and unseen aspects of intentional actions: in particular, it can predict any one of the three representational elements of goal-directed action given available evidence about the other two. We propose that this early and as yet non-mentalistic system provides the developmental core out of which naive theory of mind is constracted from the second year onwards. We consider our hypothesis from an evolutionary and comparative point of view and discuss dissociative evidence that shows the existence of teleological reasoning, on the one hand, and the lack of theory of mind, on the other, in children with autism and in non-human primates.

ROUNDABOUTS OF ‘THEORY OF MIND’ IN AUTISM: POSSIBLE PATHWAYS, ACTUAL MECHANISMS

Miklós Gyõri
Institute of Psychology, ELTE University, Budapest, Hungary
Autism Research Group, Budapest, Hungary
GyoriMiklos@axelero.hu

Krisztina Gy. Stefanik
Autism Research Group, Budapest, Hungary

Although theory of mind impairment is well documented in autism and explains many of the behavioural manifestations, several key issues are still open. Three interrelated of them are: (1) the primacy of this impairment in autism, (2) its universality within the syndrome, and (3) the problem of those individuals with autism who pass standard theory of mind tasks (‘passers’). In the present contribution we show how these issues are tied together and explore the idea that in the case of passers compensatory mechanisms may substitute the biologically default theory of mind mechanism. We review the various conceptual possibilities how atypical cognitive development may lead to such compensatory mechanisms and briefly show that current evidence available does not suffice to exclude any of them. We also present our empirical evidence for a non-mentalizing algorithmic strategy applied by some individuals with autism to solve theory of mind problems, and suggest a more subtle model on the development of impaired theory of mind in autism.

IDENTIFICATION OF OBJECTS IN 9-MONTH-OLD INFANTS: INTEGRATING WHAT AND WHERE INFORMATION.

Zsuzsa Káldy
Rutgers Center for Cognitive Science, NJ
kaldy@ruccs.rutgers.edu

Alan M. Leslie
Rutgers Center for Cognitive Science, NJ

Following Leslie et al (1998), we distinguish between individuation (the establishment of an object representation) and identification (the use of information stored in the object representation to decide which previously individuated object is being encountered). Recently, there has been controversy over how and when infants begin using property or featural information to individuate objects (Xu and Carey, 1996, Wilcox, 1999). However, there is relatively little work on the question of when and how property information is used to identify objects. Our studies show that 9-month-old infants use shape, but apparently not color, information in identifying objects that are each moved behind spatially separated screens. We relate our behavioral findings to recent neuroscientific studies of object recognition and object localization and conclude that ventral and dorsal pathways may be functionally integrated by nine months.

DETECTION OF CONTOUR CONTINUITY AND CLOSURE IN 3-MONTH-OLD HUMAN INFANTS

Ilona Kovács
Department of Psychology, Rutgers University, NJ
kovacs@cyclops.rutgers.edu

Peter Gerhardstein
Binghamton University, NY

The process by which the human visual system distinguishes objects, often termed figure-ground segmentation, depends on the ability to detect continuous contours and closed regions in the visual field. The importance of “good continuation” and “closure” was first noted by the Gestalt school in the early part of the 20th century. More recently, good continuation and closure has been investigated in a contour integration paradigm employing stimuli appropriate for low-level level cortical processing. The perceptual superiority of closed contours over open contours was found in adult human observers with this method, substantiating the gestalt school theory that closed contours are not merely lines: closed contours form a shape or surface area. This crucial step towards object-oriented processing is initiated at an early cortical processing level by the intricate interactions of orientation-selective neurons in the primary visual cortex. To what extent is the human infant’s visual system equipped with the highly specific neural connectivity-pattern giving rise to well-defined perceptual shapes? In other words, when does the visual system develop to the point at which relationships among local features in a scene are exploited, and how much fine-tuning by visual experience is necessary to acquire the basics of object perception? We used an operant conditioning procedure employing contour integration stimuli to test 3-month-olds’ sensitivity to both contour continuity and contour closure. The data demonstrate the immaturity of continuity detection and the lack of closure detection at that age. This finding modifies the general view of infant visual perception that has been more focused on the quantitative development of contrast sensitivity. Our results indicate qualitative change in terms of the organization of visual information during development.
 

SPATIAL COGNITION AND SPATIAL LANGUAGE IN HUNGARIAN NORMAL CHILDREN AND IN WMS SUBJECTS

Ildikó Király
Research Institute for Psychology, Hungarian Academy of Sciences, Budapest, Hungary
kiralyi@mtapi.hu

Mihály Racsmány
Institute of Psychology, University of Szeged, Hungary
University of Bristol, UK
racsmany@edpsy.u-szeged.hu

Ágnes Lukács
Research Institute of Linguistics, Hungarian Academy of Sciences, Budapest, Hungary
alukacs@nytud.hu

Csaba Pléh
Institute of Psychology, University of Szeged, Hungary
Budapest University Technology and Economics, Hungary
pleh@itm.bme.hu

In Hungarian spatial language encodes obligatory distinctions along the path as well, giving for all relations three differential codings for GOAL, SOURCE, and LOCATION. Beside this, suffixes, postpositions and object part names are all used to express space relations. Based on earlier studies showing earlier acquisition of suffixes compared to postpositions, and a universal preference for coding GOALS, an artificial spatial language-learning paradigm was designed for preschool children. Children had to learn new suffixes like par/per, postpositions like bagott, or object part names like a füzet revett to code for relations such as vertical, diagonal, and UNDER. Acquisition was tested following three trials in a forced choice task. Results show that the acquisition system is flexible: children picked up the markers relatively easily. There were interesting differences between the spatial relations, vertical being the easiest. However, the two most interesting results for acquisition theory were that suffixes were much easier to learn than postpositions or part names, and there was a strong GOAL preference even with artificial expressions that are not confounded with frequency. Thus, even after learning most of the existing system, the child acquiring an agglutinative system still shows her preferences based on tuning up to the system, and also general cognitive preferences like the GOAL primacy.
Williams syndrome is a genetic condition characterized by impairments of spatial cognition. In our studies on WMS subjects we observed interesting interactions between their performance on an elicited spatial description task and their general level of spatial orientation. Performance on two tasks of visual working memory (Corsi blocks and pattern matching) had an R value of 0.93 in a multiple regression equation towards performance on spatial markers. Thus, in line with the double dissociation theories, learning spatial language is related to spatial information storage systems rather than to verbal working memory. On the other hand, morphological and lexical performance was a function of verbal working memory. These findings support the general idea proposed by Landau and Jackendoff that spatial language relies on different cognitive and brain systems than the ones supporting the lexicon of nouns.
 

LANGUAGE IN WILLIAMS SYNDROME: MORPHOLOGY AND THE MENTAL LEXICON IN ACQUISITION

Csaba Pléh
Institute of Psychology, University of Szeged, Hungary
Budapest University of Technology and Economics, Hungary
pleh@itm.bme.hu

Ágnes Lukács
Research Institute of Linguistics, Hungarian Academy of Sciences, Budapest, Hungary
alukacs@nytud.hu

Mihály Racsmány
Institute of Psychology, University of Szeged
University of Bristol
racsmany@edpsy.u-szeged.hu

Williams syndrome (WMS), a rare neurogenetic is characterized by a distinctive cognitive profile: mild to moderate mental retardation with relatively good linguistic abilities, and extremely poor performance on spatial tasks. Concentrating on the linguistic abilities of children and adolescents with WMS, studies of lexical and grammatical development in 14 Hungarian WMS children are presented: children were examined on tasks testing picture naming, regular and irregular suffixation and on standard measures of verbal short-term memory. In contrast to several previous results, we observed frequency effects in both the vocabulary and the morphology task in WMS. Results on the production of accusative and plural forms confirmed for Hungarian as well that regardless of the frequency of the item, inflected forms of irregulars are harder to produce, and often overregularized in WMS, revealing a dissociation between the rules of grammar vs. the mental lexicon. Overall performance on the vocabulary and morphology tasks shows a stronger correlation with phonological short-term memory in WMS than in normals: subjects with higher span perform better on both tasks. We speculate that verbal short term memory as a relatively intact system might play a larger role in language acquisition in WMS than in normals The specification of the surprisingly close relation of phonological short-term memory with the linguistic measures awaits further study.