Poster group
Details of individual items:
poster
Infants less than 8 months old appear to lack the concept of object permanence because they fail to search for hidden objects. However, when looking rather than reaching is assessed, researchers conclude that infants have object permanence as early as 3.5 months. Why the discrepancy? One explanation is that young infants lack the means-end motor skill to retrieve objects hidden by covers: The end is to retrieve the object, and the means is to displace the cover. The present research tested the object permanence deficit hypothesis against the means-end deficit hypothesis. Infants were presented with both means-end search tasks and direct-reach search tasks. If the means-end hypothesis is correct, direct-reach search tasks should result in increased search by young infants. In Experiment 1, 72 six- and ten-month-old infants were presented with an object visible in water, partly visible in milk, hidden in milk, or hidden under a cloth. The first three events were direct-reach tasks and the fourth was a means-end task. As predicted by the object permanence deficit hypothesis, 6-month-old infants searched less, both manually and visually, when the object was completely hidden in milk or under the cloth than when it was visible in water or partly visible in milk, but there were no differences at 10 months. The means-end prediction that 6-month-old infants would search less when the object was hidden by a cloth than when it was hidden by milk was not confirmed. (See Figure 1.) Infants in a control group who were merely presented with water, milk, and a cloth did not contact the milk less than the water or cloth. Therefore, the failure of 6-month-old infants in the experimental group to search consistently in the milk was not due to the novelty of the milk. In Experiment 2, 72 six- and ten-month-old infants were presented with an object visible behind a transparent curtain, partly visible through a hole in an opaque curtain, partly visible (a lit flashlight) under a cloth, and hidden behind a completely opaque curtain. As predicted by the object permanence deficit hypothesis, 6-month-old infants searched less, both manually and visually, when the object was hidden than when it was visible or partly visible, but there were almost no differences at 10 months. (See Figure 2.) In a comparison of the hidden cloth event of Experiment 1 and the partly visible cloth-over-flashlight event of Experiment 2 (both means-end tasks), half the measures indicated that 6-month-old infants searched more for the partly visible object than for the completely hidden object, and half indicated no differences. In addition, measures of locomotor experience were not reliably related to 6-month-old infants' search in either Experiment 1 or Experiment 2. In conclusion, the results fail to support the means-end deficit hypothesis for young infants' search limitations and are consistent with the object permanence deficit hypothesis.
poster
Young infants may fail to search for hidden objects because they lack the means-end motor skill to lift occluders from objects. This account was investigated by presenting 5- to 8-month-old infants with objects hidden behind transparent, semitransparent, and opaque curtains. If a means-end deficit explains search failures, then infants should 'search' no more for an object behind a transparent curtain than for objects behind semitransparent or opaque curtains. However, level of occlusion had a significant effect on manual search and visual attention. Infants retrieved and contacted the object more, contacted the curtain more, and looked away less with the transparent curtain than with the semitransparent or opaque curtains. Adding a time delay before allowing search and presenting a distraction after occlusion further depressed infants' behavior. The findings fail to support the means-end explanation, but are consistent with the account that young infants lack object permanence.
poster
Infants first engage in manual search for hidden objects at age 8 months. Many researchers have concluded that the absence of search in younger infants cannot be attributed to a representational deficit because there is considerable evidence that young infants can both construct and access representations of occluded objects. Instead, it has been proposed that the onset of manual search is delayed until infants have developed the capacity to produce means-end behaviour. This claim was examined by Munakata et al. (1997) in a study of 7-month-olds. Infants were tested on object-retrieval tasks for which it was claimed they demonstrated relevant means-end ability. The results showed that infants were less successful at retrieving an object from behind an opaque screen than a transparent screen. Munakata et al. concluded that, because both opaque- and transparent-screen tasks required the same means-end ability, infants' failure in the opaque-screen task could not be due to a means-end deficit. Instead, they argued that infants' failure demonstrated a representational deficit, and that the late onset of manual search may be explained by the gradual development of representations that eventually achieve the capacity to regulate manual behaviour. However, it is uncertain whether infants in this study did demonstrate means-end ability. In one task infants had to pull a long cloth to retrieve the object, but recent work has shown that 7-month-olds do not produce means-end sequences with cloths of the length of those that were used. Additionally, the principal measure of success was object retrieval, but infants lacking means-end ability can also solve means-end problems. Instead, measures of intentionality are needed to show whether infants produce means-end sequences. A new study is reported in which 7-month-old infants were tested on tasks in which a screened object could be retrieved by pulling a short cloth. Previous work has established that infants first produce means-end sequences on such support tasks at age 7 months. Infants were tested in four conditions in which either an object or no object was placed on the far end of the cloth, and either an opaque or transparent screen was moved horizontally to cover the end of the cloth. Scores for intention to bring the far end of cloth within reach were obtained from measures of cloth behaviour and fixation. Intention scores were significantly higher in the object conditions than in the no-object conditions, and this difference was maintained irrespective of whether the screen was opaque or transparent. These results show that 7-month-old infants used intentional means-end behaviour to retrieve both the visible object in the transparent-cover condition, and the hidden object in the opaque-cover condition. These findings suggest that young infants representations of hidden objects are sufficient to regulate manual behaviour, and that the late onset of manual search may be explained by the gradual development of appropriate means-end ability.
poster
Studies on rhesus monkeys and human infants have shown that a success on theABtask depends on the integrity of dorsolateral prefrontal cortex. Infantsaged between7.5 and 12 months show clear improvement in performance on the AB task. Thesuccess is reflected in the length of delay imposed from no delay at all to10 or moreseconds necessary to provoke an AB error. Considering the important role oftheprefrontal cortex in the development of the working memory capacity, wewanted toexamine whether milder pre-perinatal brain lesions have any influence ondevelopment of prefrontal functions and if they do, whether these effectshave generalimpact on both components of the working memory - spatial and phonological,thelatter of which is essential for language development.Six subjects with milder pre-perinatal ischemic and/or hemorrhagic (gradeI-II)lesions and five healthy subjects were followed on a monthly basis. Allsubjects withlesions were included in a neuropediatrical follow-up and a detailedneurologicalassessment was performed. None of the mentioned subjects had hemiparesis.TheBayley Scales of Infant Development were administered to assess generalinfantdevelopment. A translated and adapted version of the MacArthur CommunicativeDevelopmental Inventory was used to follow language development from thepreverbal to verbal developmental stages. As for the data analysis, emphasiswas puton the onset time of expressive vocabulary as a reflection of phonologicalworkingmemory development.A measure of success on AB was the length of delaynecessaryto provoke an AB error reflected in perseverative reaching.Although we have dealt with a small group of subjects which puts certainlimitationsto the interpretation of results, our preliminary results indicate thatsubjects with pre-perinatal lesions make the AB error at shorter delays than healthy subjects.A clearconnection between the success at the AB task and development of theexpressivevocabulary has not been found.
poster
The errors made by infants in the AB task were taken by Piaget (1954) as an indication of an inability to represent the spatial relations of hidden objects. More recent research (Smith, Thelen, Titzer and McLin, 1999) has shown that the error arises even when infants are not searching for a hidden object, but merely responding to prompting by the experimenter in a 'lids-only' version of the AB task. Dynamic systems theorists have thus suggested that the error arises not from the infant's incomplete representation of the object but from a dynamic system which 'remembers its own previous activity', suggesting that this effect arises whether or not an object is hidden. This paper attempts to further investigate the role of a 'hidden-object' in the A-not-B error by manipulating the spatial cues to location in the task. Separate groups of infants were tested under different cueing conditions ('same covers' and 'different covers') on 'lids-only' and 'hidden-object' versions of the AB task. The order of task type was counterbalanced across subjects. The group of infants given the spatially disambiguating 'different' covers made fewer errors in the 'hidden-object' task than the infants given the 'same' covers. However this was not found to be the case with the 'lids-only' response task, in which 'different' covers seem to actually strengthen the error. Two measures indicate this dissociation: i) Accuracy compared between A and B trials (see table 1). ii) Number of extra B trials needed before a subject corrected the A-not-B error (see table 2). The above results support Munakata's (1997) claim that infants are able to represent differentially an object hidden by a lid, and a lid on its own. It is further argued that an account of the A-not-B error must consider the motivations of the infants performing these different tasks. The suggestion made here is that: i) In a 'lids-only' task, in which there is a less clearly defined goal, infants may simply be pursuing a perseverative response. ii) In a 'hidden-object' task infants may be falling back on a perseverative response due to their inability to represent the spatial location of the object for which they are searching. Whilst this difference is not always apparent when the spatial locations of the lids are not made salient, the use of 'different' covers displays a fundamental difference between 'lids-only' and 'hidden-object' tasks. The spatial disambiguation seems to reduce error in the 'hidden-object' task and strengthen perseveration in the 'lids-only' task. Returning to the traditional A-not-B error, an explanation in the light of these findings would certainly have to include a deficit in the ability to construct an external frame of reference in which to locate the hidden object. This brings us back to Piaget's original explanation and presents a strong challenge to the dynamic systems account of the error (Smith et al, 1999). Munakata's (1998) PDP model of the A-not-B error is also discussed in relation to these findings.
poster
Perseveration is the repetition of a behavior that was just performed,despite knowing that it is incorrect. Typically, perseveration in infantsis tested with the classic Piagetian A-not-B task. Infants manuallysearch for an object hidden repeatedly at location A. Then they watch anexperimenter move the object to new location B. Infants under 12 monthsof age incorrectly continue to search at A. Researchers agree as to therobustness of the A-not-B perseverative error, but disagree as to why thiserror occurs. In contrast to previous explanations based on memory demands, motorhabits, visual attention or conscious awareness, the current studysuggests that task difficulty elicits perseverative errors. I designed twolocomotor versions of the A-not-B task that varied on task difficulty.Instead of manually searching, infants walked over flat ground (easy floorcondition) or climbed down staircases (difficult stair condition) to reachtheir goal. The goal was in full view during all trials, thus eliminatingall memory demands.Fifteen 13-month-olds (7 girls, 8 boys) participated. In both easy anddifficult conditions, infants goal was to locomote to their parents. Inthe easy floor condition, trials took place entirely on flat ground. Thestarting place and locations A and B (91 cm from starting place) weremarked with yellow mats. The experimenter started each trial with infantsfacing their parents. On five consecutive trials, babies were encouragedto walk to their parents at location A. On the sixth trial, in full viewof babies, parents moved to location B, 90 from location A. Theexperimenter faced babies in the new direction and encouraged them to walkto location B from the old starting place. In the difficult stairscondition, infants began each trial on a 42 cm2 platform, 54 cm high. Twoidentical (124 cm long) staircases were attached to the platform 90 toeach other. As in the floor condition, on five consecutive trials, babiesstarted on the platform facing their parents, and were encouraged todescend the stairs and go to their parents at location A. On the sixthtrial, in full view of babies, parents moved to the foot of theperpendicular staircase at location B. The experimenter faced babies inthe new direction and encouraged them to go to location B.Perseverative responses depended on task difficulty. On every measure,infants displayed more perseveration in the difficult stairs conditionafter a switch in goal location than in the easy floor condition. A thirdof the sample used the old A staircase to go to location B, despite havingto perform an unnecessary detour away from the goal and an additional 27 %looked toward the A staircase while descending B. Many children who didnot perseverate on these strict criteria displayed more subtleperseverative behaviors in the stair condition (switching direction,touching more stairs, and taking longer to descend). Results indicatethat an increase in task demands taxes infants cognitive resources,rendering them unable to perform a difficult task successfully, whilesimultaneously inhibiting a compelling behavior.
poster
This experiment examined the integration of visual input and information arising as a result of movement through the environment (e.g., body movement) on the spatial orientation of 13- to 15-month-old infants. Previous work has shown that infants respond more successfully on a search task after self movement, as opposed to after object movement, in a lit environment. As well, infants perform best when they have access to both visual and body movement information simultaneously. However, the form of self movement employed in these studies was passive infant displacement and other work has shown that infants may benefit more from active movement in a search situation. Our study, therefore, investigated the usefulness of active versus passive body movement information in conjunction with the presence or absence of visual cues specifying changes in location.Infants completed a search task that required them to retrieve a toy hidden in one of two containers. Before being allowed to search for the hidden toy on a given trial, each infant received one of four different types of movement transformations. The first of these movements involved rotating the table 180( relative to a stationary infant (e.g., object movement). The second and third transformations involved moving the infant 180( relative to a stationary table, with these infant displacement trials involving either passive movement (the infant was rolled on a chair) or active movement (the infant was induced to walk) through the world. The fourth movement type involved a no movement control. Thus, this variable systematically manipulated the type of body movement information available for spatial updating. The presence versus absence of visual information for spatial orientation was manipulated by having the different movement conditions occur in either a light or a da!rk environment.Search was best following no movement and overall accuracy with visual information groups combined was significantly better in this condition than after object and active self-movement. No differences were evident between performance after no movement and after passive self-movement or between any of the other movement conditions .Contrary to our prediction, as well as to previous findings, the results do not indicate any facilitation of active movement on search. In fact, the results seem to support the opposite notion that passive movement provides more beneficial information for infants to update their knowledge of their body position in space and to subsequently make a correct search response. These findings lead to the suggestion that it is not body movement, per se, that facilitates search but rather the opportunity to visually track the correct hiding location.
poster
One of the major tasks in cognitive development is acquiring inhibitorycontrol. With weak inhibitory control, behavior is dominated byprepotent responses. Inhibitory control is not mature until well intoadolescence. In late infancy, a major milestone in the development ofinhibition is reached when the infant can reach around a transparentbarrier to retrieve an object rather than attempting to reach throughit, and the infant stops making the AB error. However, from this pointuntil preschool, when inhibitory control is studied with tasks like theWisconsin Card Sorting Task, little is known about the developmentaltrajectory of inhibition. Recently, Bauer, Schwade, Wewerka, and Delaney (1999) tested childrenaged 20 and 27 months on a problem-solving task. Children had to solvefour three-step problems after seeing only one of the three steps,either the first or the third. They found differences in performancerelated to age as well as which step was seen. Particularly in the thirdstep condition, to succeed, children had to suppress a tendency tosimply imitate what they saw in order to produce the whole sequence. The data in Bauer et al. present an opportunity to examine thedevelopment of inhibitory control in late infancy and early childhood. Twenty-four 20-month-olds and twenty-four 27-month-olds were given fourproblems to solve. Each problem involved a set of unique props that hadto be assembled in three steps, in the correct order, to achieve adesired end-state. After exploring the props during a baseline period,children saw one of the three steps modeled, either the first step orthe third step. After children attempted to solve the problems, thewhole solution was modeled and children were allowed to imitate as acontrol. Children who saw the third step were more likely to solve the problemsand were more efficient in their problem solving than children who sawthe first step. To see if inhibitory control was related to thesedifferences, we looked at children's perseveration on individual stepsby counting the maximum number of times each child repeated each stepconsecutively. On most steps, repetition increased over baseline whenchildren saw the third step but not when they saw the first step. Patterns of perseveration on the first step were especially interestingacross ages. Older children repeated the first step more when they hadseen the first step than when they had seen the third step, as one mightexpect. However, the younger children's results seem counterintuitive:they perseverated more on the first step when they saw the third stepmodeled. One hypothesis that might explain the younger children'sbehavior is that seeing the third step made them more motivated to solvethe problem, but they still got 'stuck' on the first step. Olderchildren, conversely, were able to inhibit this tendency to get to theend-state, resulting in more efficient performance overall. Analysis of children's performance on a problem-solving task thatrequires suppression of a prepotent response presents itself as a viablemeans of exploring the development of inhibitory control in the periodof transition from infancy to early childhood.
poster
The multistep multilocation search task (Zelazo, Reznick, &Spinazzola, 1998) is an age- appropriate analog of the A-not-B task thathas been successfully used with 2-year-olds. To retrieve a conspicuouslyhidden reward (e.g., a candy), the child must: (a) remove a foam barrier,(b) pull out a tray, and (c) choose the stimulus (from an array of 5) thatleads to the reward. On 'B' trials, the location of the reward ischanged. Interference on B trials could arise either from a tendency toreach to the initially correct location or from a tendency to choose theinitially correct stimulus, or some combination of the two. The current study was designed to determine whether stimulus cuesaffect 2-year-olds' performance on the multistep multilocation search taskabove and beyond interference associated with the location of a previouslycorrect stimulus. Two sets of stimuli were created. One set consisted ofconcrete, familiar, and easily identifiable pictures (i.e., cup, hat). Theother set included abstract and unfamiliar line drawings. In bothconditions, all 5 stimuli were visually distinct from one another. Preliminary data analyses indicate that performance on the first'B' trial was independent of condition, both for the number of childrenwho performed correctly (concrete: 4/12, abstract: 3/10, 2(1) .03, p .87 ) and for the number who perseverated (concrete: 6/12, abstract: 3/10, 2(1) .90, p .34). However, children in the concrete condition mademore errors than those in the abstract condition both in terms of thetotal number of errors (Cohen's d .42) and the number of perseverativeerrors (Cohen's d .47) before the first correct 'B' trial. Thissuggests that experience with the concrete stimuli increased interferenceon the task beyond what was expected based on location alone. Based on a final trial in which the reward was hiddensurreptitiously and the location of the previously correct stimulus wasswitched with the location of a neutral stimulus, children werecategorized either as focussing on the location or on stimulus cues.Children focussing on location made more total errors (Cohen's d .45)and perseverative errors (Cohen's d .75) before the first 'B' trialthan children focussing on stimulus cues. This suggests that experience atthe previously correct location increases interference on the task beyondthat of the stimulus cues. Although preliminary, these seemingly inconsistent findings can beexplained under an additive model that is consistent with proposalsregarding dissociable systems of learning and memory (e.g., Marcovitch &Zelazo, in press; McDonald & White, 1993). According to McDonald andWhite (1993), for example, a dorsal striatal system that mediatesstimulus- response learning may be distinguished from a hippocampal systemthat is involved in spatial learning and memory. In the current findings,both the concrete stimulus cues and a tendency to focus on locationinterfere with performance on the task. When these sources of interferenceoccur simultaneously, the interference is combined, but there is nointeraction.
poster
Since Piaget's (1954) seminal work on the development of the object concept, a variety of explanations for the failure to search correctly for hidden objects have been offered, including memory (Cummings & Bjork, 1981), response inhibition (Diamond, 1985), attention (Harris, 1983), problem-solving (Baillargeon, DeVos, Graber & Black, 1990), and accounts based on the differential ability of representations to control behavioral systems (Munakata et al., 199?). A factor that has been ignored in the literature is the infant's developing self-regulatory skill. Self-regulatory has been shown to predict cognitive performance in infancy (Kopp, 1989), specifically infant's performance on an A-not-B task (MacLean, Miller, Keenan & Keating, 1995). As Lewis (1993) has argued, infants who are unable to modulate negative affect tend to withdraw their attention from the external world. As MacLean et al. argue, failures of self-regulation tie in with explanations of infant search errors in the A-not-B task based on failures in attention and memory. Infants who become dysregulated by negative affect and who fail attend to the critical events (i.e., the hidden object's location) in the course of the A-not-B task should show a higher incidence of search errors on the B trial (the reversal trial) then do infants who remain positive and engaged in the task. In this study, the role of developing self-regulatory ability was studied by examining infant's ability to modulate their level of negative affect and attention over the course of a standard, two location A-not-B task. The A-not-B task affords a unique opportunity to study affect regulation because the infant is required to interact with a stranger, is restrained by their parent before searching, is delayed before being freed to search, and has the toy repeatedly taken from them upon successful search. In the present study, 36 nine-month-old infants and their parents participated in a two location object permanence task. After a brief warm-up, infants were given three A trials and a single B trial. Infant's level of negative affect and level of task engagement were measured across the four trials. Correct search on the B trial was also measured. The prediction was that infant's who demonstrated high levels of negative affect would show low levels of task engagement and poor search performance on the B trial whereas infants who demonstrated low levels of negative affect would show higher levels of task engagement and correct search on the B trial. The results supported both hypotheses. Infants who searched correctly at B were significantly more likely to show high levels of task engagement and low levels of negative affect than infants who searched incorrectly at B. These data highlight the importance of self-regulatory ability in children's developing cognitive competence. The ability to search for a hidden object requires the infant to remain engaged in the task and to modulate their level of negative affect. Infants who fail to remain engaged and positive are significantly less likely to search correctly for the hidden object. These data attest to the importance of examining interactions between self-regulatory ability and cognitive development. Developing self-regulatory ability allows the infant to maintain a representation of the location of the hidden object in working memory over the delay period and thus, significantly predicts infant's search performance on the critical reversal trial.
poster
This research addresses two distinct types of knowledgerepresentations --- active and latent --- and their potential role ininfants' task-dependent memory for hidden objects. Infants as youngas 3.5 months show an apparent sensitivity to hidden objects in visualhabituation studies, yet infants fail to manually retrieve hiddenobjects until around 9 months, independent of any motor orproblem-solving deficits. This research explored the possibility thatinfants may use latent representations to recognize unusual events invisual habituation studies, but require later developing activerepresentations to retrieve occluded objects. The current studies were adapted from tasks used with non-humanprimates (Miller & Desimone, 1994), in which monkeys were presentedwith sequences of visual stimuli and trained to pull a lever when thefirst stimulus of a sequence was repeated. Inferotemporal neuronsshowed no sign of maintained firing for the target stimulus during theintervening stimuli; however, roughly half of the neurons recordedresponded differently to the target stimulus on its secondpresentation. Monkeys appeared to simply observe the first stimulusand as a result, formed a latent memory trace for this stimulus,resulting in facilitated processing (usually suppressed activity) toits second presentation. Monkeys' behavioral responses in a second experiment supported thefacilitated processing interpretation of suppressed neuronalresponses. The stimuli were slightly altered so that monkeys couldnot solve the task on the basis of familiarity. Half of the trialsnow contained a repeat of a non-first stimulus, and the same monkeyswere trained to respond to only the match to the first stimulus in thetrial. Early in training, monkeys often mistakenly responded to therepeat of the non-first stimulus, supporting the idea that monkeyswere not responding based on active maintenance of the target stimulusbut instead based on latent facilitated processing that occurred forany repeat. In contrast, when monkeys eventually learned to respondto the match-to-first only, sustained neuronal firing to the targetwas observed in prefrontal cortex, indicating a more active form ofrepresentation. In our infant-adapted paradigm, 72 8-month-old infants werefamiliarized with sequences of visual stimuli followed by a dancingflower lighting up and moving to music after the first stimulusrepeated. They were then tested with three types of sequences: 1)first-stimulus-repeats, 2) no-stimulus-repeats, and 3)non-first-stimulus-repeats. Infants' head-turning reaction times tothe flower following these sequences were used as an indication oftheir ability to remember the stimuli and recognize repeats. Infantsturned to the flower more quickly after first-stimulus-repeats thanafter no-stimulus-repeats (just as monkeys pulled the lever afterfirst-stimulus-repeats but not after no-stimulus-repeats), suggestingthat they were able to learn the rule predicting the dancing flowerevent and to remember the first stimulus. Further, infants turned tothe flower similarly fast following first-stimulus-repeats andnon-first-stimulus repeats (just as monkeys at first pulled the leverafter both first- and non-first-stimulus repeats), suggesting thatthey recognized a repeated stimulus using latent, rather than activerepresentations. Implications of these findings for infants'task-dependent behavior are discussed.
poster
In order to contact a moving object, infants must prospectivelycontrol arm movements in anticipation of the object's future movement.Berthier, Bertenthal, Seaks, Sylvia, Johnson, and Clifton (1999) increasedthe difficulty of this task by occluding a portion of a ball's path withan opaque screen. The ball rolled behind a screen directly in front ofthe infant, disappeared at its left edge and emerged to its right. Onsome trials, the ball's path was blocked by a barrier directly behind thescreen and did not emerge to its right. In order to contact the ball onno-barrier trials, infants needed to begin their reach before the ball wasoccluded. Even though the barrier protruded above the screen and infantshad experience seeing the ball stopped by the barrier during training, 9month-olds reached to the right of the screen on barrier trials. In asubsequent free search period, infants were unable to locate the ball,even though it was available by an indirect reach around the left edge ofthe screen. In order to determine whether infants were unable to encode orremember the barrier's presence due to the difficulty of the task, or ifthey simply had developed a tendency over repeated training trials toreach to the right side of the occluder, Johnson, Sylvia, Berthier, andClifton (1999) replaced the opaque screen with a transparent one. Thiseliminated the need for infants to remember the configuration of thebarrier and the screen. In support of a memory load hypothesis, infantsdid not reach to the right on barrier trials; they reached to the right inanticipation of the ball on no-barrier trials only. The current study sought to reduce the memory load required inthe original task by providing infants with some visual cues about thepresence or absence of the barrier. The barrier was visible through a vertical transparent window down thecenter of an otherwise opaque screen. This gave infants a cue to thebarrier's presence, while occluding the same portion of the ball's path. Even with the added cue of the barrier's presence, infantscontinued to anticipate the ball's emergence on barrier trials by reachingto the right of the screen. In contrast to Berthier et al., however, oncethis initial reach was completed, infants were far more successful atfinding the ball on barrier trials. We conclude that the visual cue of the barrier's presence does notsufficiently reduce the memory load while infants are reaching to enablethem to perform appropriately on barrier trials. Although they have avisual reminder of the barrier, this information appears not to beavailable to guide infants' reaches for the moving ball. When the task issimply to locate the ball without the time pressure of catching it, thevisual cue becomes useful. These results reflect infants' capabilities tointegrate knowledge about the configuration of the apparatus and toremember the consequences of the barrier's presence under relaxed taskconstraints.References Berthier, N.E., Bertenthal, B.I., Seaks, J.D., Sylvia, M.R.,Johnson, R.L., & Clifton, R.K. (1999). Using object knowledge in visualtracking and reaching. Manuscript submitted for publication. Johnson, R.L., Sylvia, M.R., Berthier, N.E., & Clifton, R.K.(1999, April). A transparent screen can facilitate infants' planning areach. Poster session presented at the biennial meeting of the Societyfor Research in Child Development, Albuquerque, NM.
poster
Because they challenge the Piagetian claim that infants are notable to represent unseen objects until they have achieved objectpermanence at 8-10 months of age, studies investigating infants reachingin the dark for unseen, sounding objects have received much attention. These studies have found that infants as young as 6 1/2 months reach forsounding objects at consistently high rates in the dark. Based on theseresults, researchers have concluded that infants at this age can formassociations between objects and sounds and then rely on theserepresentations to guide their reaching behavior in the dark (Clifton,1998). Despite these repeated findings, the question of whether infantsreally 'know' what they are reaching for when they do reach for soundingobjects in the dark has plagued this line of research. In many of thestudies, infants also had reached for the objects in the light. Stack,Muir, Sherriff, and Roman (1988) and Clifton, Perris, and McCall (1999)found that infants who were not given the opportunity to associate arattle sound with a graspable object in the light did reach for thatobject in the dark. These studies, however, did not control for infants'experiences with the common rattle sound outside of the laboratory.Equally problematic has been the suggestion that by presenting infantswith sounding objects in the light prior to their reaches for thoseobjects, researchers have classically conditioned them to reach towardthese sounds in the dark (Clifton). The current study sought to examine whether 6 1/2 month olds canform and utilize representations to guide their behavior by controllingfor the problems just described. Infants were divided into two groups, asilent-object and a sound-experience group, and were allowed to reach fora toy during several familiarization trials in the light. For infants inthe silent-object group, the toy remained completely silent, while forinfants in the sound experience group, the toy produced an unfamiliardoorbell sound only after these infants had reached for and obtained thetoy. Following their respective light experiences, infants in both groupswere presented with the doorbell sound in the dark and their reachingbehavior was observed. The frequency rate of infants' reaches in the dark was affected bytheir experience with the sounding toy in the light. Infants who weregiven the opportunity to associate the doorbell sound with the toy in thelight (i.e. the sound-experience group) reached during more of the darktrials than infants who were not given this opportunity (i.e. thesilent-object group). This result provides evidence that infants willreach toward novel sounds in the dark only when they have been given theopportunity to associate them with graspable objects; they will not reachfor just any sound in the dark. Furthermore, because the sound neverpreceded the reach, this action could not have been conditioned. Weconclude that infants at this age can form and utilize representations toguide their actions toward objects when those objects can only be heard,but not seen.ReferencesClifton, R. K. (1998, April). Representation expressed by action onunseen objects. In A. Needham (Chair), Effects of experience onperception. Symposium conducted at the International Conference on InfantStudies, Atlanta, GA. Clifton, R.K., Perris, E.E., & McCall, D.D. (1999). Does reaching in thedark for unseen objects reflect representations in infants? Manuscriptsubmitted for publication.Stack, D.M., Muir, D.W., Sherriff, F., & Roman, J. (1989). Development ofinfant reaching in the dark to luminous objects and 'invisible sounds'.Perception, 18, 69-82.