The focus of research in the Symbolic and Cultural Development Lab is on examining the processes responsible for the development of symbols that develop early in life.  We primarily investigate visual symbols, but also contrast this development with that in other domains, including language and play. We consider children's drawings to be products of a symbol system, or a means the child has of communicating about his or her world.  Our research is aimed at exploring the emergence and subsequent development of symbolizing about both the physical world and the emotional world.  The experiments typically address these questions by asking how children understand and respond to a range of graphic symbols, from simple line drawings to complex museum art, as well as when children first come to produce intentional graphic symbols as they cross the divide between scribbling and first representational forms.  Because we are interested in the earliest beginnings of graphic symbolic function, and how this may be related to other symbolic systems, our experiments sometimes explore infants' abilities.

Research Program

Overview

I conduct research that explores the development of early cognitive development.  My specialty within this broad field is the development of symbolic functioning.  The long term goals of this research are to explore the precursors of symbolic functioning in infancy, and its origins and refinement in young children.  I am investigating how the various domains of symbolic functioning (i.e., including pictures, 3D models, language, play and gesture) interact or influence each other throughout the developmental process. In this research, infants’ and children’s comprehension as well as their production of symbols is studied. The research focuses on perceptual, cognitive, and social factors that influence symbolic functioning, and explores how the influence of these factors may vary across cultures.

The development of symbolic functioning has an impact beyond cognitive development.  Social interaction is based on the use of culturally-derived symbol systems.  Children’s ability to acquire fluency in the symbols of their culture critically determines their success in becoming a fully contributing member of that culture.  Although the foundation for expertise in symbolic systems is laid down in early childhood, there is significant refinement and development of the process throughout childhood, and even beyond, in cases where the system is complex, as in theoretical mathematics.  A great deal of young children’s education is devoted to developing and enhancing communication skills with symbols, especially language and mathematics.  Without knowledge of the origins of symbolic functioning, educational intervention in the foundation years and subsequent educational efforts, especially those that are remedial, are made more difficult.

My research is grounded on the premise that a full understanding of symbolic development requires investigations of the basic perceptual, cognitive, learning and social cognitive processes the individual organism brings to the situation, as well as the supports that others provide to that development as the organism begins to function in a symbolic cultural context. Thus, I conduct two, related lines of research. My NSERC-funded research focuses on the basic processes involved in preparing an infant for the insight that symbols are representational vehicles for communication, and for the subsequent refinement of this symbolic understanding. My SSHRC-funded research examines the role that cultural supports play in symbolic development.  Here, I examine parental beliefs and practices in contexts that foster visual, play and language symbols, usually within family units. In addition, I have begun to explore more formal educational contexts that engage children in the visual symbol systems of their culture. The cross-cultural work is conducted across diverse cultures, including Grenada, Peru, Samoa, Japan, India, and Canada.

Current NSERC¹ Funded Research: Visual Symbols

The main question explored in my research of visual symbols is one that is general to all domains of symbols, which is the problem of reference.  How does the child come to understand that a picture, or indeed any symbol, stands for a particular referent? In the case of pictures, how does the child come to be able to use a graphic vocabulary to produce a drawing that conveys a meaningful message about the world?  This research has already shown that social factors mediate visual symbol understanding and production (Callaghan, 1999; Callaghan & Rankin, 2002), that language may scaffold visual symbol use (Callaghan, 2001), and that perceptual similarity of choice objects and iconicity between symbol and referent predictably influence visual symbol use (Callaghan, 2001).

Callaghan (1999) identified that social communicative processes can facilitate the understanding of the symbolic status of pictures and that intentional production of representational drawings does not appear until symbolic understanding has occurred (at 3 years).  In this study, children aged 2-4 years were asked to draw pictures of simple objects and then use those pictures to indicate to the experimenter the object to choose in a social-communicative game.  Children also responded to the experimenter’s symbols and drew a second set of drawings.  The results showed that 2-year-olds did not use the experimenter’s pictures as symbols and did not produce effective symbols.  Three- and 4-year-olds produced more effective symbols after the game and improved their drawings even more if they were given feedback that their symbol had not been effective.  That 2-year-olds do not use pictures as symbols is consistent with DeLoache’s claim (e.g., DeLoache et al, 1997) that children of this age are unable to hold two concepts of the picture in mind at once; the picture as an interesting object and the picture as a symbol of another object.  The findings with older children provide a unique insight that an important variable in the emergence of symbolic understanding is social exposure to the communicative use of those symbols. 

I have developed a variant of the standard joint attention paradigm (Tomasello & Farrar, 1986) to label novel objects visually in an effort to determine whether graphic labeling of an attended object facilitates comprehension and production in children aged 2 to 3.5 years.  In this task the experimenter made a quick graphic sketch of either the child’s (attended) or the experimenter’s (non-attended) object.  The experimenter ensured that the child watched as the drawing was made. Results from pilot research (Callaghan & Pencer, 1999) suggested that directing attention away from the child’s object of focus to label the object did not hamper performance for children at this age, however, 2-year-olds show better comprehension for pictures they saw the experimenter draw and 3.5-year-olds show better production for objects they saw the experimenter draw. 

A variant of this graphic labeling procedure was used in a longitudinal training study (Callaghan & Rankin, 2002).  In this study, graphic labeling, using a joint attention procedure, was administered as training to 8 children at weekly intervals for a 4 month period, beginning at 28 months.  A control group of 8 children received placebo training (no drawing, but manipulation of the toys).  There was no sign of symbolic understanding of pictures in any children at the outset of the study and the study’s aim was to determine whether training would facilitate this understanding.  A second aim was to determine whether graphic symbol development was related to symbolic development in other domains, thus understanding and production of language, visual and play symbols was measured at monthly intervals for all children from the onset of the study until their third birthday.  Results indicated that training was effective in promoting visual symbol comprehension and production and there were strong positive correlations in symbolic functioning across domains. 

We are following up this research with a number of studies which aim to investigate the transition from perceptual into symbolic levels of functioning. In one study (Depicted Action) we (Callaghan, Rochat, MacGillivray & MacLellan) are contrasting infants’ imitative responses to live, videotaped, and pictured action (through a series of 3 still photos, or a single photo of the goal of the action).  Our reasoning is that infants can model the action directly from the perception of action in live and video conditions but must infer action from the still photos.  Our preliminary results suggest that infants are not successful in either the video or the sequence photo condition until much later than they are able to model in the live condition. In a second study (Categorical perception of symbol/referent) we are using the technique of Quinn & Eimas (1996) to show that infants categorize pictures together, separate from their referents, in spite of the high similarity between pictures and their referents.  We are also using the habituation paradigm in a third study (Expectations of Symbol/Referent Match) to explore infants’ expectations (e.g., Baillargeon, 1991) of the match between pictures used as symbols on boxes that contain referents.

Another recent study explored the refinement in older children’s understanding of the symbolic function of pictures (Callaghan & Rochat, 2003). In this research we asked when children (aged 2-7 years) come to consider the attributes of the artist when making judgments of pictures, that is, when do children understand that in addition to information about the referent, a picture may also contain information about the person who created the picture?  We were also interested in exploring whether children’s developing theory of pictures was related to their developing theory of mind.  In a series of 3 studies we found that children were sensitive to a range of attributes (age, sentience, affective style, emotion) of the artist when making judgments of pictures, and that picture judgments were correlated to performance on a standard false belief task (Perner, Leekham & Wimmer, 1987), but not to emotion (Gross & Harris, 1988) and interpretive (Pillow & Henrichon, 1996) theory of mind tasks.

In addition to social processes, I have explored stimulus and task factors that may constrain the development of graphic symbol comprehension.   In a series of three experiments (Callaghan, 2001), I used a delayed matching object-to-symbol task with 2.5- and 3-year-olds and varied iconicity, word availability and perceptual similarity of choice items.  The results from the first experiment in this series replicate the earlier finding (Callaghan, 1999) that 2.5-year-olds do not use pictures to stand for objects.  Three-year-olds do use pictures to guide their choice and also benefit from greater iconicity between symbol and referent. Although iconicity has been predicted by others to be an important factor in development, this is the first experiment to manipulate it directly with picture stimuli (DeLoache et al, 1997).  In the second study of the series we found that if verbal labels are available to scaffold performance, then 2.5-year-olds do as well on the task as 3-year-olds. If verbal labels are not available, then even 3-year-olds’ performance deteriorates.  In the third study of the series we manipulated perceptual similarity along with word availability and replicated the word availability effect and found that at both ages children’s use of pictorial symbols was hampered when choices were perceptually similar.  In fact 3-year-olds performed at chance levels if the choice objects were perceptually similar and they were prevented from using verbal labels to scaffold their performance.  Taken together, the findings suggest a greater flexibility in accessing information from the representation for 3- as compared to 2.5-year-olds and provide some support for the notion that language may mediate early picture symbol functioning.  Greater flexibility of processing is a trend that has often been noted in attention and perception research for children of this age (e.g., Callaghan, 1993).

I am currently collecting data in a study that will establish the relative effects of iconicity/abstractness on the use of a variety of types of symbols.  Children (2.5 and 3 years) are being presented with iconic and abstract symbols in verbal, graphic, and gestural domains.   The joint attention paradigm has once again been modified to allow for presentation of the symbol while the child is exploring the object.  After presentation of each set of four symbols (6 sets; iconic and abstract for verbal, gestural, and graphic) comprehension and production of symbols is being assessed.   This within-subjects design contrasts the typical form of graphic symbols (iconic) with the typical form of verbal symbols (abstract).

I am also collecting data in an experiment that modifies the task to assess whether response inhibition may be a factor in the inability of 2.5-year-olds to choose the correct object in the matching tasks used above, as it has been found to be for children this age in object permanence tasks (Diamond, 1991).  Specifically, the requirement to choose between two highly desirable objects in the matching tasks may constrain the symbolic functioning of 2-year-olds who may prefer to reach in and explore any object regardless of what they have seen in the picture.   In this study I am presenting 18- and 24-month-olds with an object to explore and then, out of view, hiding it in one of two identical boxes.  The location of the object is cued by placing its picture on the lid of the box. Children are instructed to use the picture to help them find the object.  If this task is successful in producing symbolic behavior with pictures prior to 3 years, then we will adapt it to use in a fourth study that first presents the hiding task for the child to engage in, and then a false belief task for the child to predict where others will look.  Sequencing of tasks in this way will help to establish what develops as the child comes to have a more sophisticated understanding of the symbolic function of pictures (Perner, 1991).

My most recent research on the basic processes involved in visual symbolic development have been investigations of children’s intentional symbolic productions and their ability to read referential intentions in others’ actions.  In one study (Callaghan & Rochat, submitted), we explored the stimulus factors that children use to identify intentions in their own, and others’, drawings. They primarily use form cues, but color cues can also be utilized if form cues are not present.  Children also benefit greatly in the visual symbols domain by having a trace of their intention as they produce it.  This does not happen for language and play domains.  When that trace is eliminated, as in having children draw with chopsticks on tracing paper, they have great difficulty finding the drawing that they made, especially if it is embedded in a background of drawings that other children made of the same item. In a second study (Callaghan & Anton, submitted), children were presented with one of a variety of demonstrations involving referential intent with visual symbols.  In one, an adult drew pictures to label the different contents of boxes that looked identical when closed.  In a second, she used three other sorts of visual symbols (photos, replicas, play dough models) to label the boxes on three trials.  In a third, the child had to infer referential intent from a failed attempt of the adult to make a picture. In a control condition, children where simply given the test trial (common to all conditions), which involved giving the child a pencil and paper to determine whether they would spontaneously draw pictures to label the boxes. The findings indicate that 3-year-olds imitate adults’ referential intent when that has been directly modeled, but not when they have to infer it from multiple symbols or failed attempts.  Five- and 7-year-old children did well in all cases, except control where no children of any age spontaneously labeled the boxes with pictures.  These results support our claim that children can read referential intentions in the actions of others, and will imitate those intentions.  We believe that imitation is a powerful social learning mechanism that brings children into symbolic worlds.  In the final study of this area (Callaghan, Rochat, Lerikos, MacDougall & Corbit, submitted), we have developed a visual symbols version of the false belief location task.  In this task, children and two experimenters sort toys into two boxes and label them with a picture.  One experimenter leaves the room and the second asks if the child wants to trick her by switching the pictures on the box.  When the pictures are switched, the child is then asked to predict where the experimenter will look for her favorite toys when she returns.  When we used small plastic toys of typical categories like fish, bugs, cows, dogs, etc we found that it was not until 7 years that children correctly predicted that the experimenter would search for toys in the box labeled with a picture of that type of toy, not the box where those toys actually were. When we used icons that children are highly familiar with (MacDonald’s, Tim Horton’s), 5-year-olds were successful at the task.  In follow-up studies, we showed that children’s performance on symbolic tasks using icons that are highly favored (e.g., Disney characters) was significantly better than tasks using familiar, but less engaging, categories of objects. In the picture version of the false belief task, we propose that we are tapping children’s understanding of the communicative intent that underlies the use of pictorial symbols.

SSHRC² Funded Cultural Research

The first research project in this area was an interdisciplinary work on developing, and piloting, methods to examine early infant and child development across 6 cultures.  The mandate of SSHRC-RDI research program that funded this research is to make a contribution that enables the field to move in a new direction.  Our particular goal in this research was to work collaboratively with scholars from other cultures under study, and across disciplines, to develop culturally sensitive methods to investigate a variety of central issues of early development. 

SSHRC Funded Research: Artistic Development and Visual Symbols

In this research I have explored how the understanding and production of expressive symbols is linked to emotional understanding and communication in general, and to expression in visual art in particular.   Even infants are sensitive to emotional tone in their social interactions with caregivers (e.g., Stern, 1985).  Is the understanding of emotional tone in human products such as artworks built upon this existing ability?  Looking across cultures one is impressed by the multitude of ways emotional tone is expressed in art.  How does the child come to judge emotional tone and other stylistic attributes in the art of their culture?  Does the child’s ability to communicate about their own emotional world in drawings rest upon sensitivity to emotion portrayed in the visual symbols of others?

In the first of the studies in this area I showed that even very young children are sensitive to emotional tone portrayed in museum art (Callaghan, 1997), contrary to the then dominant view in the literature that it is not until adolescence that children can judge expressiveness (e.g., Gardner, 1970).  In this study a nonverbal task was used to avoid the confounding effect that language level may have on the ability to indicate which emotion is portrayed.  Children aged 5 to 14 years were shown reprints of museum art and asked to match individual paintings with one of 4 photos showing an actress portraying emotional expressions.  Although there were developmental differences, all of the children were performing better than chance and were consistent with adult artist’s judgments of emotional tone. 

I have collected data in a second study that looks at preschoolers’ sensitivity to emotional tone in art (Callaghan, 2000).  To do this, I modified the task and used the same paintings.  In the modified task children were asked to help four teddies, costumed to exemplify 4 emotions, find paintings for their houses.  First the teddy ‘chose’ 3 paintings and then the child was asked to choose between two, one of which matched and the other which differed in emotional tone.  When 3-year-olds were given this reference set they performed equivalently to 5-year-olds (approximately 80 percent).  However, when given two choice trials without a reference set 3-year-olds were at chance and 5-year-olds were well above chance (approximately 80 percent).  These results suggest that referencing is a process that facilitates acquisition of the aesthetic norms of the culture.

In addition to asking whether children comprehend emotional symbols in art, I have also explored the variables that have an impact on children’s ability to portray emotion in their own drawings.  Callaghan, MacDonald & MacGregor (unpublished) asked children to make drawings of 5 emotions (happy, sad, angry, calm, frightened), both when they were experiencing the emotion in question (parents collected these at home), as well as when they were out of the context of that emotional state (drawings collected by parents when they were in a relatively calm state).  Naive adults made judgments of which emotion was being portrayed and success was deemed to be the degree of consistency between the judgments made by adults and the intention of the child.  We found that drawings were better when drawn out of the context of the emotional state, with the exception of calm drawings, which were equivalent in and out of context.  This suggests that heightened emotional state interferes with the ability to render an effective graphic symbol of emotion.

In another set of experiments, I have been examining when children come to be able to make judgments of artistic style.  This research brings the study of children’s sensitivity to artworks closer to body of research that investigates children’s developing theory of mind (c.f., Astington, Harris & Olson, 1988) in the sense that children are asked to make judgments about the impact the style of a person will have on the product that person makes.  In the first of these studies children made stylistic judgments of museum art using a matching-to-sample task (Callaghan & MacFarlane, 1998).  Artistic style varied alone (control) or in conflict with subject matter (experimental), and both varied in level of discriminability (low, high).  We found that 5-year-olds can make style judgments when subject matter conflicts if the styles are highly discriminable (e.g., VanGogh, Magritte).  10-year-olds were better able than 5-year-olds to ignore conflicting subject matter cues and chose style even when styles were hard to discriminate (e.g., VanGogh, Gaugin).  The impact of attentional processes on performance was expected given findings from perception research with children this age (Shepp & Swartz, 1976).   Another of my honors students, Susan MacLean, is extending this investigation to preschoolers using the referencing task (see description above) I have previously used to examine preschoolers judgments of emotional tone (Callaghan, 2000).  In the current study 3- and 5-year-olds are being asked to judge style alone or in conflict with subject matter cues and under conditions where style is hard or easy to discriminate.  It is expected that 3-year-olds will perform poorly in the two-choice task but will benefit from the referencing task, and 5-year-olds will do well under both task conditions.

One of the benefits of using museum art in the studies described above is ecological validity of the stimuli.  One of the problems is that the stimulus factors that mediate performance are not controlled in the stimuli.  We addressed this issue in one of our Callaghan & Rochat (in press) studies when asking children to make judgments of how attributes of the artist would affect the products of those artists.   In one study of 3 to 5 year old children that asked children to judge which of two pictures was drawn by an agitated and calm person, we found that it is not until 5 years that children consistently made distinctions between these styles.  Children used the stimulus factors of line quality, closure, overlap, and symmetry when making these judgments. 

Future Research Goals

            To expand the significance of this research my aim is to collaborate on research in three areas. First, convergence on the issue of domain specificity in the graphic symbol domain - or lack thereof - and its development would benefit from an examination of neural processing using tasks that either permit processing in two domains (e.g., language and visual) or restrict it to a single domain (e.g., visual).  The strategy would be to examine adults using both fMRI and ERP methods, and to collect developmental data using the noninvasive ERP method. Second, gaining a complete picture of early symbolic development requires more extensive investigations of cases where symbolic development is compromised due to illness or disability.  Studies of autistic individuals would be a natural path for this research to take, given that one of the primary symptoms of this disorder is a compromised ability to engage in symbolic communication with others. Convergence on the particular issue of how language dominates other symbol systems is also a future goal of mine.  This could be achieved by studying the development of visual symbol understanding and production in deaf children who do not have access to verbal language, as well as in clinical populations of adults who have experienced isolated damage to language processing areas.  Third, cross-cultural research is needed to help us to understand the relative resilience of developmental process across varieties of cultural support systems. I am currently working out the details of two cross cultural projects related to this goal.  One will investigate both the biological underpinnings (i.e., the individual perceptual, cognitive and social cognitive skills available from infancy) and the social supports (i.e., the characteristic ways that caregivers interact with infants and young children when using symbols) for symbolic understanding in the domains of visual symbols, play and language across different cultural settings.  The goal is to identify any universal milestones, as well as cultural diversity, as may exist in the development of symbolic functioning.  A second project will explore at the level of cultural institutions how children are enculturated into the world of art. In addition to these planned projects, I will continue to collaborate on a cross-cultural project of youth resilience to negative life experiences.

Natural Sciences and Engineering Research Council of Canada¹

Social Science and Humanities Research Council of Canada²