Poster group
Details of individual items:
poster
In the first years of life, infants gain increasing control over two significantand complementary representational systems, one a verbal symbol system manifestfor example in referential language, and the other a nonverbal symbol systemmanifest for example in pretense play. Some controversy has surrounded thequestion of whether children with multilingual heritages are advantaged ordisadvantaged in symbolic development. How do infants' initial representationalabilities unfold under bilingual versus monolingual rearing experience? Thisstudy examined two representational abilities, expressive language and symbolicplay, in 28 infants reared in a monolingual Dutch and 34 infants reared in asimultaneous bilingual Dutch-French environment in Belgium. 'Bilingual' wasdefined as families in which the mother and father each spoke a different nativelanguage (Dutch or French) and where both Dutch and French were spoken at home.Symbolic development was assessed when the children were 20 months old usingmaternal report and observation. For language, mothers completed the EarlyLanguage Inventory (the ELI consists of 624 words divided into 19 categories).Mothers also completed a questionnaire to assess their tendency to make sociallydesirable responses (maternal reports of children's behavior may overestimatechildren's abilities and research has pointed to cultural differences inself-serving bias) and a family sociodemographic questionnaire (mothers who wererearing infants bilingually estimated infants' exposure to the mothers'non-native language and infants' communication in the mothers' non-nativelanguage). All instruments were translated into Dutch and adapted to Belgianculture as needed. For play, infants were videotaped for two consecutive 10-minplay episodes, first alone and then with mother; play was coded using a mutuallyexclusive and exhaustive coding system. Group differences emerged in language,but not symbolic play. Monolingual children said significantly more words thanbilingual children, even when the bilingual children's words in both of theirlanguages were counted (i.e., the total number of different words that bilingualchildren produced in Dutch and French was still fewer than their monolingualcompatriots). Monolingual children said significantly more nouns, verbs, andadjectives than bilingual children. Both groups of children said significantlymore nouns than verbs, suggesting a noun bias in early vocabularies. Bycontrast, monolingual and bilingual children did not differ in nonverbalsymbolic play, either when they played alone or with their mothers. Genderdifferences emerged in symbolic development in both language and play: Girlssaid significantly more words than boys, and girls engaged in significantly moresymbolic play than boys both when they played alone and when they played withtheir mothers. In early cognitive growth, bilingual exposure from infancyappears to be specific and (at least temporarily) affects development of theverbal and not the nonverbal symbol systems.
poster
The current study investigated the role of iconicity in children's acquisition of gestures as names for objects. Previous research has indicated that while 18-month-old infants learn either arbitrary words or arbitrary gestures as names for things, 26-month-old infants learn the words but not the gestures (Namy & Waxman, 1998). While traditional theories of symbol learning would predict that the initial symbols learned by children would be iconic in nature (i.e. resemble their referents), these results raise the possibility that iconicity may not be as important early in the symbol learning process. The goal of the current study was to further investigate the role of iconicity in symbol acquisition. Of particular interest was how the influence of iconicity on symbol acquisition changes with age.Eighteen- and 26-month-old infants were shown 2 objects (e.g. hammer and bunny) and were taught a gesture as a name for one of them. During an initial training session, half of the children learned an iconic gesture (e.g. hopping motion for bunny) while the remaining half learned an arbitrary gesture (e.g. dropping motion for bunny). Following the training session, comprehension of the iconic and arbitrary symbols was assessed using a forced-choice procedure. There were 2 trial types. In the Test trials, the experimenter elicited a choice using the novel gesture (e.g. 'Which one can you get? [gesture]') In the Control trials the experimenter elicited a choice without using the gesture (e.g. 'Which one can you get?') to ensure that the child was not consistently choosing the target toy regardless of the experimenter's instructions. We examined whether children more consistently selected the target object on Test trials than on Control trials to determine whether they successfully mapped the symbol to the referent. Preliminary results indicate that while the 18-month-olds learned both arbitrary and iconic gestures as names for things, 26-month-olds only learned the iconic gestures. These preliminary results replicate the findings of Namy and Waxman (1998) in that the 18-month-olds learned the arbitrary gestures as symbols while the 26-month-olds did not. However, this study additionally demonstrated the counterintuitive finding that iconicity facilitates symbolic acquisition later in development but is not necessary for early symbolic acquisition. While the 26-month-olds' success at learning the gestures was influenced by the iconicity of the symbol, 18-month-olds learned symbols regardless of their iconicity. These findings suggest that iconicity might be especially helpful for 26-month-old infants acquiring symbols as names for things. Table 1Mean Percentage of Target Toys Chosen by Age and Symbol Type 18-month-olds 26-month-olds Test Trials Control Trials Test Trials Control TrialsGesture Type n n Iconic 7 .69 .52 5 .68 .54Arbitrary 8 .71 .50 6 .53 .46
poster
Standard assessments of children's cognitive and language abilitiestypically depend on motor skills, used in object manipulation and pointing, andexpressive language. However, these assessments are not appropriate forchildren with severely limited motor abilities and no expressive language,who may have receptive language in the normal range. An assessment batteryto evaluate fundamental cognitive skills in nonverbal, motor-impairedchildren is being developed. The battery is being pilot-tested on 30typically developing toddlers and a clinical population of 5 children with earlybrain malformations (holoprosencephaly). We report on the efficacy of twotasks, the Pictorial Representation Scale (PRS) and the Forced ChoiceVocabulary Task (FCV), which use eye-gaze to assess symbolic skills andreceptive language, and we correlate performance with parental languageassessments.Symbolic skills are highly correlated with and may be pre-requisites todevelopment and are typically assessed during pretend play in toddlers(Bates et al, 1979, Sugarman, 1978). The PRS is a visual match-to-sample taskdesigned to determine symbolic skills in children whose pretend play islimited by motor skills. Children are presented with a target object at themidline and instructed to 'match' by looking at the target in a pair ofitems presented to either side, to receive verbal reinforcement. The pairsare objects (identity match), or colored photographs or line drawings(non-identity matches). Ten pairs of each modality are presented semi-randomly.Previous studies have shown that children of all developmental levels aremore accurate with identity than non-identity matches, although childrenwithout functional language have difficulty with both (Guerette et al.,1998; Mirenda & Locke, 1989; Sevcik & Romski, 1986). Standardized tests of toddlers' receptive vocabulary typically use apointing response to pictures, thus, receptive language can be under-estimatedbecause of low symbolic skills (Evans & Wodar, 1997) or the inability orreluctance to point. The FCV is a verbal match-to-sample task in whichchildren are instructed to indicate (by looking or pointing) which of a pairof objects corresponds with a spoken word (nonidentity match). Objectsrepresent 18 monosyllabic words normally emerging in children's receptivevocabularies between 12 and 15 months of age (Fenson et al, 1993).Understanding is also assessed by parental report. To assess a finer level ofphonological discrimination, performance on words differing in the wholesyllable is compared with performance on words differing in onset only. Children's first looks and overall responses were scored by two blindjudges. Preliminary analyses of children in the non-clinical group revealthat group performance was greater than chance on both tasks and moreaccurate when scored by overall response than first looks. Performance on the PRSimproved with age as expected; however, in contrast to previous researchcited here, performance was not dependent upon modality. Previous studiespresented blocks in order of increasing difficulty confounding modalityeffects with fatigue. Performance on the FCV did not vary with phonologicaldiscrimination. Efforts to evaluate the utility and validity of thesetasks is ongoing.
poster
The second year of life is the period in which transition occurs fromexploration to symbolic play. To capture this transition, a well known12-step sequence of play development was defined (Belsky & Most, 1980). Itwas found that simple manipulation decreases linearly, whereas pretence playincreases linearly. Behaviours that involve relating materials andapproximation to pretence display curvilinear functions, suggesting thatthey are forms of activity that link early exploration with more advancedplay.The present study aims at examining these transitional behaviours moreclosely and at reconsidering them. In the above research, some of theresponses coded as symbolic do not necessarily involve a symbolic component(e.g. the child may not have been pretending to drink imaginary liquid butmay have put a cup to his/her mouth because of the learned associationbetween cup and mouth). On the other side, some of the responses coded asfunctional-relational relate two objects in a manner that implies more thantheir functional properties (e.g. set a cup on a saucer instead on a glass).We propose to consider these kinds of behaviours as conventional forms ofactivity, learned by the child in everyday social practice. They are tied toan arbitrary use of the objects and can be enacted at procedural levelwithout implying pretence. Therefore they could properly function astransitional forms from literal to symbolic level of acting.Six infants (two females and four males) were videotaped for 25 minutesevery week from 10 to 24 months of life when interacting at home with theirmothers around a set of age-appropriate toys. About sixty sessions for eachinfant were coded by a system which distinguishes literal object usage (i.e.acting on an object on the basis of its intrinsic properties) fromconventional (i.e. acting on the basis of a learned socio-cultural model)from symbolic (i.e. acting on the basis of a mental, imaginaryobject/event). Given our interest in conventional level of the activity,more analytic coding was applied to this level and conventional behaviourwas divided into poor-conventional (approximations, attempts, incomplete),conventional and rich-conventional (association of a conventional actionwith gestures and/or typical words). Multilevel modelling will be applied tothe resulting data with the aim of tracing the average developmental trendand individual paths. We hypothesise that the main transition from literalto symbolic level of the activity, which occurs in the second year of life,is mediated by conventional level. Therefore, we expect that conventionalbehaviour will display curvilinear functions, whereas literal and symbolicbehaviour will display linear functions. Also interesting will be theinspection of the supposed 3-step conventional sequence in order to discussthis transition in terms of continuous/discontinuous process.Finally, we plan to apply to our sessions an observational instrument whichis aimed at capturing the interactive aspects of individual behaviour (i.e.Relational Coding System, developed by Alan Fogel). We expect to findrelationships between co-regulation in mother-infant joint activity and thedevelopment of infant's object usage.
poster
Previous research has established that very young children oftenhave difficulty using a symbol (e.g., a scale model or picture of a room)as a source of information about a real situation (the location of anobject hidden in the room itself). We recently demonstrated that the sameis true for video: In several studies, 24-month-olds who watched on a videomonitor as an experimenter hid a toy in the room next door retrieved thetoy less than half of the time. Children who watched identical hidingevents directly through a window had perfect performance--100%. One factor that may contribute to the children's problem is theirprior experience with video. Hours spent in front of a TV set may teachthem that events on television have little to do with ongoing reality,leading them to discount the video presentation in the lab as irrelevant tothe problem they are trying to solve. Perhaps exposure to live video wouldclarify the video-reality connection. In two recent studies, parents gave their 24-month-olds five10-minute sessions of live home video experience during which the childreninteracted with their own live image 'on TV.' The children who receivedthis relevant, prior exposure to video were significantly more successful(63% to 75% correct) than the children in the earlier video studies. Thiswas the first reported use of a symbolic medium by 24-month-olds to solvean object-retrieval task. Given the importance of these results, the first study to bereported here is a replication using essentially the same procedures.Results from half of the proposed number of subjects (N 16) reveal thatchildren exposed to live home video have been very successful at theobject-retrieval task with video (71% correct), whereas children withoutthis experience have performed very poorly (17%). Further results involve the effect of live video exposure onchildren's use of other symbolic media. We report two exciting 'transfereffects.' The 24-month-old children in the present study return to the labto participate in an object-retrieval task employing still pictures (theexperimenter points on a picture to indicate where the toy is hidden in thenext room). The children with live home video experience have been muchmore successful at this task (54% correct) than the control group (only8%). Results of a long-term transfer study are also very promising: 9 ofthe 11 children who succeeded in the preliminary home video studies a yearago were contacted and returned to the lab (as 36-month-olds) toparticipate in a difficult symbolic task involving a scale model. Thechildren did much better at this task (67%) than is typical at this age(around 30%). We will add an appropriate control group to complete thisstudy. Results found thus far suggest that relevant experience with livehome video may have remarkably persistent effects on young children's useof symbolic media.