Constructivist theories (child development)



Constructivism is the philosophical and scientific position that knowledge arises through a process of active construction. From this view, knowledge structures are neither innate properties of the mind nor are they passively transmitted to individuals by experience. In this entry, we outline recent advances in constructivist models of cognitive development, beginning by analyzing the origins of constructivist developmental theory in the seminal writings of Piaget. We then examine the ways in which theoretical and empirical challenges to his theory have resulted in the elaboration of a more powerful constructivism in the form of neo-Piagetian and systems models of human development.

Piagetian foundations of constructivist theory

Piaget’s theory of cognitive development is simultaneously a structuralist and constructivist theory. For Piaget, psychological structures are constructed in development. The basic unit of cognitive analysis is the psychological structure, which is an organized system of action or thought. All psychological activities are organized, whether they consist of a 6-month-old’s reach for a rattle, an 8-year-old’s logical solution to a conservation problem, or a 15-year-old’s systematic manipulation of variables in a science experiment.

Psychological structures, or schemes, operate through the dual processes of assimilation and accommodation. Piaget appropriated these notions from the prior work of James Mark Baldwin. Drawn from the biological metaphor of digestion, assimilation refers to the process by which objects are broken down and incorporated into existing structures, while accommodation reflects complementary processes of modifying or adapting an existing structure to accept or incorporate an object.

Any psychological act requires the assimilation of an object into an existing structure and the simultaneous accommodation of that structure to the incorporated object. For example, to perform the sensorimotor act of grasping a rattle, an infant incorporates (assimilates) the rattle into her grasping scheme. However, to grasp the rattle, the infant must modify her scheme to the particular contours of the incorporated object.

Piaget maintained that psychological structures undergo successive transformations over time in a series of four stages. Within his theory, stages exhibit several important properties. Firstly, each stage corresponds to a particular type or quality of thinking or psychological organization. From this view, infants are not simply small adults – they think in fundamentally different ways from older children and adults. Secondly, the stages form a hierarchical progression with later stages building upon earlier ones. Thirdly, the stages form a single, universal, and unidirectional sequence. Regardless of the culture in which a child resides, thinking develops in stages toward the common endpoint of formal operations. Fourthly, Piagetian stages form structures d’ensemble (i.e., ‘structures of the whole’). Piaget’s position on the organization of thinking within stages was complex. On the one hand, the concept of stage implies homogeneity of organization. Within a given stage, Piaget held that schemes are general and have wide application to broad ranges of cognitive tasks. On the other hand, he also invoked the concept of decalage – the idea that cognitive abilities within a stage develop at different times. Despite such deecalage, Piaget held that as children resolve the conflicts that exist between cognitive sub-systems, psychological structures develop into increasingly broad and integrated wholes.

According to Piaget, for the first two years of life, infant schemes function within the sensorimotor stage of development. Sensorimotor schemes consist of organized systems of action on objects. Piaget held that infants cannot form representations (images) of events in the absence of direct sensory input. As such, sensorimotor schemes reflect integrations of the sensory and motor aspects of action. Thinking emerges between 18 and 24 months of age with the onset of the semiotic function during the pre-operational stage of development. During this stage, children are capable of forming representations of events (e.g., words, images), but are incapable of manipulating these images in logical or systematic ways. Pre-operational intelligence is marked by the emergence of symbolic play, deferred imitation, and the use of words to refer to present and absent objects. During the concrete operational stage, thinking becomes systematic and logical. Children are able to operate logically on concrete representations of events. The capacity for concrete operations underlies a child’s ability to perform various logical tasks, including conservation, class inclusion, seriation, transitivity judgments, etc. It is not until the formal operational stage (adolescence onward) that individuals are able to free their logical thinking from concrete content. In formal operations, adolescents become capable of operating using abstract forms. In so doing, thinking becomes abstract, and adolescents and adults can conceptualize hypothetical and systematic solutions to logical, mathematical, and scientific problems.

The concept of equilibration provides the backbone of Piaget’s constructivist theory of development. Equilibration refers to an inherent, self-regulating, compensatory process that balances assimilation and accommodation and prompts stage transition. Piaget elaborated upon several forms of equilibration. The first involves the detection of a conflict or discrepancy between an existing scheme and a novel object. He held that a state of equilibrium results when an object is successfully incorporated into a given scheme, and thus when assimilation and accommodation are in a state of balance. A state of disequilibrium results when there is a failure to incorporate an object into a given scheme. A child who only has schemes for cats and dogs will have little difficulty identifying common instances of these two classes, but his schemes would be in disequilibrium when first encountering a rabbit. Disequilibrium, in turn, motivates successive acts of accommodation that result in a significant modification of the existing schemes. A new scheme thus emerges from the failure of existing schemes. Where there were initially only schemes for cats and dogs, there are now schemes for cats, dogs, and bunnies.

Piaget discussed additional forms of equilibration, which involve the resolution of conflict between two competing cognitive schemes (e.g., when conservation of length and conservation of number come into conflict), and between individual schemes and the larger systems of which they are a part (e.g., integrating conservation of length, number, and mass into an abstract understanding of conservation). Piaget also acknowledged other processes that contribute to development. For example, in order for a stage transition to occur, there must be a requisite level of neurological maturation; a child must actively experience the world by acting on objects and people; a child must receive cultural knowledge in the form of socially transmitted and linguistically mediated rule systems (e.g., mathematics, science). Nonetheless, disequilibrium engendered by cognitive conflict provides the driving force of development.

Questions about Piaget’s structuralism

Table 1 describes five basic problems and criticisms that emerged with regard to central principles in Piaget’s theory of development. The first four critiques concern the Piagetian notion of cognitive structure or stage. Each critique is a variant on the idea that there exists more variability in children’s cognitive functioning than would be predicted by a strong notion of stage. Research has indicated that the developmental level of even a single child’s cognitive actions can change with variations in the level of contextual support provided to the child, the specific nature of the task, the conceptual domain in which the task occurs, and the child’s emotional disposition. For example, Western European and North American children generally conserve number by 6-7 years of age, mass by 8 years, and weight by 10 years, but generally do not solve tasks about inclusion of sub-classes within classes until 9 or 10 years of age.

Research also suggests that providing training and contextual support for concrete operational tasks lowers the age at which children succeed in performing such tasks. For example, Peter Bryant and Thomas Trabasso demonstrated that providing young children with memory training (e.g., having them memorize which of each pair of adjacent sticks was larger or smaller) lowered the age at which they were able to perform a transitivity task, determining which stick in a pair is larger by inferring from comparisons of other pairs of sticks. Studies like these challenge the idea that children’s thinking develops in broadly integrated and homogeneous structures (i.e., stages). Instead, they suggest that thinking is organized in terms of partially independent cognitive skills that develop along different pathways. Researchers have also criticized Piagetian concepts such as equilibration, assimilation, and accommodation as difficult to translate into clear and testable hypotheses. Finally, others have noted that Piaget did not pay enough attention to the ways in which social processes contribute to development. This last issue requires additional elaboration.

Table 1. Moving toward the new constructivism.

Piagetian construct

Source of problem

Analysis of developing skills

Structural principles I. Inner competence as property of individual child. Individual cognitive structures function as basic units of cognitive activity. Cognitive structures are seen as properties of individual children.

Social context and affective state play a direct role in modulating level of functioning. Evidence suggests that performance on similar tasks in the same children vary dramatically with changes in contextual support and affective state.

Skill as property of individual in social context. Skills reflect actions performed on physical and social objects in particular social contexts. Child and social context collaborate in the joint construction of skills.

II. Limited number of broad stages. Piaget postulated four broad stages of cognitive development with a series of sub-stages.

Variability in performances as a result of task complexity. Differences in the complexity of tasks used to test children’s stage acquisition produce different assessments of operative ability.

Precise developmental yardsticks. Skill analyses allow both broad and fine-grained analysis of development across a total of thirteen levels with a large number of smaller steps between levels.

III. Stage as structure d’ensemble. Piaget held that cognitive structures entail broad abilities having wide application to multiple tasks.

Decalage. Uneveness in the development of skills is the rule rather than the exception in ontogenesis, even for abilities presumed to be at the same developmental level.

Skills develop within particular tasks, domains, and social contexts. Rejecting the notion of globally consistent stages, skill analyses assess skill development within particular conceptual domains, tasks, and social contexts.

IV. Development as unidirectional ladder. Piaget proposed a unidirectional model of stage progression in which cognitive capacities in all cultures follow the same abstract progression of stages.

Varied sequences of development.

Evidence suggests variation in developmental sequence in different children, tasks, and cultures, as well as failures to observe predicted Piagetian sequences.

Development as multidirectional web.

Different skills develop along different trajectories for different tasks, domains, persons, contexts, and cultures. As such, development proceeds as a web of trajectories rather than as a ladder of fixed or universal steps.

Process principles

V. Individual action as primary source of developmental change. Piaget viewed cognitive disequilibria as the primary mover of development, suggesting a central role for the individual child as the main mover of development.

Limited focus on social, cultural, biological, and emotional organizers of developmental change. Evidence suggests that social interaction, language, culture, genetics, and emotion play important roles in the constitution of psychological structures.

Developmental change occurs as a product of relations between biological, psychological, and sociocultural processes. Biological, psychological, social, and cultural processes necessarily coact in the formation of novel psychological structures.

Sociocultural challenges to the primacy of individual action

Piagetian constructivism relies heavily, but not exclusively, on the notion that children’s own actions are primary movers of development (equilibration). According to Piaget, thinking emerges in the pre-operational stage as children abbreviate and internalize sensorimotor actions to form mental images (inner abbreviated action). Constructing an image of one’s mother involves the abbreviated and internal reconstruction of actions that one performs when one actually looks at one’s mother. Thus, thinking becomes a matter of internally manipulating images that have their origins in the actions of individual children.

Sociocultural psychologists, especially those inspired by Vygotsky, noted that Piaget’s constructivism neglected the role of social interaction, language, and culture in development. From a Vygotskian perspective, children are not solitary actors. They work with adults and peers in the creation of any higher-order developmental process. In social interaction, partners direct each other’s actions and thoughts using language and signs. Signs function as important vehicles of enculturation. Unlike symbol systems, such as mental images or pictures, signs are used to represent relatively arbitrary meanings that are shared within a linguistic community. For example, understanding the meaning of words such as ‘good’ or ‘democracy’ involves learning a relatively arbitrary cultural meaning that is shared and understood among individuals who comprehend a certain language, such as English. Vygotsky maintained that all higher-order psychological processes are mediated by signs. Development of higher-order mental functions occurs as children internalize the results of sign-mediated interactions that they have with others. As children come to use signs to mediate their thinking, they think in culturally not merely personally organized ways.

In his explanation of the social origins of higher-order functions, Vygotsky (1978) invoked his general genetic law of cultural development: “Any function in children’s cultural development occurs twice, or on two planes. First, it appears on the social plane and then on the psychological plane. First it appears between people as an interpsychological category and then within the individual child as an intrapsychological category” (p. 57). The concept of internalization explains how sign-mediated activity that initially occurs between people comes to be produced within individuals in development. For example, to help his 6-year-old remember where she put her soccer ball, a father may ask, “Where did you last play with it?” In so doing, the father and daughter use signs to regulate the mental retracing of the girl’s actions. As the girl internalizes these sign-mediated interactions, she acquires a higher-order memory strategy – ‘retracing one’s steps.’

This vignette illustrates the Vygotskian principle of the zone of proximal development (ZPD). The ZPD refers to the distance between a child’s level of functioning when working alone and her developmental level working with a more accomplished individual. In the above example, the father’s questions raise his child’s remembering to a level beyond that which she can sustain alone. The child’s remembering strategy is formed as she internalizes the verbal strategy that originated in joint action. In this way, the research spawned by sociocultural theory challenges the primacy of children’s individual actions as main movers of development.

Reinventing constructivist theory: trajectories of skill development

In what follows, we will elaborate the major tenets of dynamic skill theory, a neo-Piagetian constructivist theory of psychological development. We describe how skill development can explain cognitive development and address key challenges to Piaget’s theory elaborated in Table 1. Rather than speaking of broad logico-mathematical competences, according to dynamic skill theory the main unit of acting and thinking is the developing skill.

The concept of skill

The concept of skill provides a useful way to think about psychological structures. A skill refers to an individual’s capacity to control her behavior, thinking, and feeling within specified contexts and within particular task domains. As such, a skill is a type of control structure. It refers to the organization of action that an individual can bring under her own control within a given context. The concept of skill differs from the Piagetian notion of scheme or cognitive structure in several important ways. To begin with, a skill is not simply an attribute of an individual; instead, it is a property of an individual-in-a-context. The production of any instance of skilled action is a joint product of person and context (physical and social). As such, a change in the context in which a given act is performed can result in changes in the form and developmental level of the skill in question. In this way, context plays a direct role in the construction of skilled activity.

Contexts differ in the extent to which they support an individual’s attempt to produce skilled activity. Contexts involving high support provide assistance that supports an individual’s actions (e.g., modeling desired behavior; providing cues, prompts, or questions that prompt key components to help structure children’s actions). Contexts involving low support provide no such assistance. Level of contextual support contributes directly to the level of performance a person is able to sustain in deploying a given skill. A person’s optimal level refers to the highest level of performance one is capable of achieving, usually in contexts offering high support. A person’s functional level consists of his or her everyday level of functioning in low support contexts. In general, a person’s optimal level of performance under conditions of high support is several steps higher than his functional level in low support contexts.

Figure 1 depicts developmental variation in a child’s story telling in a variety of high and low support conditions. In the context of elicited imitation, a child is asked to imitate a complex story modeled by an adult. In elicited imitation, the child’s story functions at a level that is several steps higher than when he or she tells stories in free play, or is asked to tell his or her best story – both conditions of low support. Minutes later, when an adult prompts the child by stating the key components of the story, the child again functions at optimal level. Then after a few more minutes low support conditions result in reduction of the child’s performance to functional level again. These fluctuations in skill level occur in the same child on the same task across varying conditions of contextual support separated by mere minutes.

Variation in skill level for stories as a function of social-contextual support. In the high-support assessments, the interviewer either modeled a story to a child (elicited imitation) or described the gist of a story and provided cues (prompt); the child then acted out the story. In low support assessments, the interviewer provided no such support but either asked for the child's best story or simply observed story telling in free play.

Figure 1. Variation in skill level for stories as a function of social-contextual support. In the high-support assessments, the interviewer either modeled a story to a child (elicited imitation) or described the gist of a story and provided cues (prompt); the child then acted out the story. In low support assessments, the interviewer provided no such support but either asked for the child’s best story or simply observed story telling in free play.  

Contexts involving high and low support differ from contexts involving scaffolded support. In contexts involving high or low support, the child alone is responsible for coordinating the elements of a given skill. For example, an adult may model a complex story for a child who then produces the story without further assistance. In scaffolded contexts, an adult assists the child by performing part of the task or otherwise structures the child’s actions during the course of skill deployment. Scaffolding allows adult-child dyads to function at levels that surpass a child’s optimal level. When a mother helps her 6-year-old tell a story by intermittently providing story parts and asking the child leading questions, the dyad can produce a more complex story than the child could tell alone, even with high support. As a result of contextual support and scaffolding, children do not function at a single developmental level in any given skilled activity. Instead, they function within a developmental range of possible skill levels.

A second way in which the concept of skill departs from Piagetian theory is that skills are not general structures. There are no general, de-contextualized, or all-purpose skills. Skills are tied to specific tasks and task domains. Skills in different conceptual domains (e.g., conservation, classification, reading words, social interaction, etc.) develop relatively independently of each other at different rates and toward different developmental endpoints. Assessments of the developmental level of one skill in one conceptual domain (e.g., conservation) will not necessarily predict the developmental level of skills in a different domain (e.g., classification), or even in conceptually similar tasks (conservation of number versus conservation of volume). One can chart developmental sequences only for skills within a given domain and within particular social contexts and assessment conditions.

Levels of skill development

Skills develop through the hierarchical coordination of lower level action systems into higher-order structures. Table 2 presents the levels of hierarchical organization of a developing skill based on Fischer’s dynamic skill theory (Fischer, 1980; Fischer & Bidell, 1998). In this model, skills develop through four broad tiers: reflexes refer to innate action elements (e.g., sucking; closing fingers around an object placed in the hand); sensorimotor actions refer to smoothly controlled actions on objects (e.g., reaching for an object); representations consist of symbolic meanings about concrete aspects of objects, events, and persons (e.g., “Mommy eat candy”); abstractions consist of higher-order representations about intangible and generalized aspects of objects and events (e.g., “Conservation refers to no change in the quantity of something despite a change in its appearance”). Within each broad tier, skills develop through four levels. A single set refers to a single organized reflex, action, representation or abstraction. Mappings refer to coordinations between two or more single sets, whereas systems consist of coordinations of two or more mappings. A system of systems reflects the intercoordination of at least two systems and constitutes the first level of the next broad tier of skills. For example, a system of sensorimotor systems constitutes a single representational set.

In this way, dynamic skill theory specifies four broad qualitatively distinct tiers of development comprising a total of thirteen specific levels. It also provides a set of tools for identifying a variable number of steps between any two developmental levels. These levels have been documented in scores of studies in a variety of different developmental domains. In the following sub-sections, we illustrate dynamics of skill development through an analysis of how sample skills move through the levels and tiers specified in Table 2.

Development in infancy

Here, we examine the development of visually guided reaching as an example of skill development. Like all skills, reaching does not emerge at any single point in time. Instead, like all skills, it develops gradually over the course of infancy and takes a series of different forms over time. In addition, at any given point in development, an infant’s capacity to reach for seen objects varies dramatically depending upon the task at hand, the trajectory of an object’s movement, degree of postural support, and other variables.

Table 2. Tiers and levels of skill development.

Ages are modal times that a level first emerges according to research with middle-class American and European children. They may differ across social groups.

1 Ages are modal times that a level first emerges according to research with middle-class American and European children. They may differ across social groups.

Note: In skill diagrams, each letter denotes a component of a skill. A large letter = a main component (set), and a subscript or superscript = a subset of the main component. Plain letter = component that is a reflex, in the sense of innate action-component. Bold letter = sensorimotor action; italic letter = representation; and script letter = abstraction. Line connecting sets = relation that forms a mapping, single-line arrow = relation that forms a system, and double-line arrow = relation that forms a system of systems.

1 Ages are modal times that a level first emerges according to research with middle-class American and European children. They may differ across social groups.

The first tier of skill development consists of reflexes. In skill theory, reflexes refer to simple elements of controlled action, present at birth, that are activated by environmental events. Reflexes do not simply consist of encapsulated reactions such as eye blinks or knee jerks. Instead, they consist of more molar elements of action over which infants exert limited control in contexts that activate them. These include action elements such as simple acts of looking at an object held in front of the face, sucking objects placed in the mouth, as well as emotional acts such as cooing or smiling. At the level of single reflexes (Rf1), infants in the first month of life are capable of exerting limited control over several action elements that function as precursors to visually guided reaching. Soon after birth, infants engage in pre-reaching, which involves making arm movements in the direction of objects. In addition, at this level, infants actively look at an object placed in front of the face. An infant is capable of making wobbly adjustments of the head to track an object that moves slowly within his line of vision. In addition, a baby is also capable of various reflex actions, such as the palmar reflex, in which pressure on the hand prompts the infant to close her fingers together. In each of these reflex actions, a situation or object must be made available to the child (e.g., an object in front of the eyes; a situation placing the body in a particular position).

Beginning at about 7 weeks of age, infants gain the capacity to construct reflex mappings (Rf2), which consist of active coordinations of two or more single reflexes. At this level, for example, an infant coordinates two or more simple movements into short swiping movements toward a seen object. Such swipes at this age tend to be rigid and ballistic motions that are poorly coordinated for grasping or touching objects. Over the next month, infants gain the capacity to coordinate multiple reflex mappings into reflex systems (Rf3). As a result, arm movements are smoother but still poorly coordinated. In highly supportive contexts, depending upon the precise placement of the target and the child’s posture, infants sometimes hit their targets.

At around 15-17 weeks of age, infants gain the capacity to coordinate two reflex systems into a system of reflex systems (Rf4/Sm1). Systems of reflex systems are also the first level of the next broad tier of development, because they engender sensorimotor actions (Sm1). Sensorimotor actions consist of single more highly controlled actions on objects in the environment. Unlike reflex acts, sensorimotor actions are more modulated by the child, with less need for activation by environmental events. Using sensorimotor actions, in high support contexts, an infant can produce the first successful reaches. When an infant’s posture is supported, she is capable of directing arm movements toward a seen object such as an object moving toward her. Initial movements are generally jerky, consisting of multiple action segments that do not follow a straight line. At this level, although an infant can reach for an object while looking at it, looking and reaching are not yet fully differentiated. The child must already be in the process of looking at the object to reach for it. In so doing, she looks mainly at the object and not at her arm. As such, looking operates primarily to trigger movement. The infant does not yet map variations in arm extension to variations in looking in a controlled and coordinated way.

Over the next several months, within the level of single sensorimotor acts, looking and reaching become increasingly differentiated and coordinated. In one study, after watching an object moving back and forth in a slow and constant motion, 5-month-old infants were able to reach for and intercept the moving object. In so doing, the infant exhibits a degree of coordination of looking and reaching in order to predict the trajectory of the object. In another study, 6-month-old infants were presented with an object that moved toward them from one of two corners of a stimulus display. As the object moved within reach, it either continued to move along its linear trajectory, or else it turned in a 90-degree angle and continued movement. On trials when the object switched directions, infants moved their heads and reaches in the direction of the anticipated path of the object. This study indicates further visual-motor planning in the reaching of 6-month-olds.

Although reaching in 6-month-olds involves the simultaneous use of looking and reaching within the same object-directed action, it is not until the onset of sensorimotor mappings (Sm2), beginning around 7-8 months of age, that infants are fully able to coordinate distinct acts of looking and reaching in relation to each other. At this level, infants can begin to map variations in looking with variations in reaching. For example, an infant can reach for an object in order to bring it in front of the face and look at some aspect of the object. Similarly, a child can look at a moving object and map changes in the movement of his or her hand to changes in the movement of the object. In addition, infants can detour their reaches around barriers placed between themselves and target objects. For example, if an adult attempts to block a child’s reach toward an object, the infant can redirect his action around the obstacle. From this point onward, visually guided reaching becomes increasingly smooth and deployed for more complex purposes. Beginning around 11-12 months, infants become capable of coordinating two or more sensorimotor mappings into a sensorimotor system (Sm3). At this point, infants are capable of using multiple coordinated acts of looking and reaching in order to explore an object from a variety of angles, as in Piaget’s descriptions of a 1-year-old’s systematic variation of the position and orientation of a toy in order to see how to get it through the bars of a crib or how to make it fall in different ways.

Nativism and the question of innate abilities in infancy

Challenges to a constructivist model of infant development have also come in the form of neo-nativist claims that infants are born with innate rules or abilities for particular conceptual domains. For example, with regard to language development, Noam Chomsky has claimed that deep structured rules of syntax are innate. More recently, such claims have also been made for concepts such as gravity and inertia, space, numerical addition. From this view, certain foundational abilities are not constructed gradually in development, but are instead present at birth.

An illustrative case in point concerns recent work on the development of object permanence. Piaget maintained that infants gradually construct an understanding that objects continue to exist even while absent from sight, sound, and touch. He argued that this ability undergoes transformation as infants coordinate early grasping and looking schemes into more complex schemes for understanding how objects can behave, as described above. For example, at about 6 months, infants are able to retrieve a partially hidden object, but fail to retrieve a fully hidden object, even when they have seen the object being hidden. A key transition occurs a few months later, when infants begin to coordinate two schemes to enable search for fully hidden objects. At this point, an infant removes a cover in order to retrieve a rattle that she has seen hidden under it, making a major advance in object permanence. Children’s understanding of object permanence continues to develop throughout the sensorimotor stage and beyond.

A well-known challenge to this constructivist view comes from work by Baillargeon (1987). Infants from 3 to 5 months of age were habituated to the sight of a small door that swung upward from a flat position (the top of the door facing the child) in a 180-degree arc to lie flat again on a solid surface (the top now facing away from the child). They were then shown two scenes involving objects placed behind the rotating door. In one scene, called the possible event, the door swung up but stopped at the object. In the impossible event, the object was removed surreptitiously and the door appeared to swing right through the space occupied by the object. Infants as young as 3 to 4 months dishabituated to the impossible event to a greater degree than they did to the possible event. Baillargeon interpreted this behavior as evidence for object permanence and concluded that infants achieve it four to five months earlier than Piaget had reported.

Based on such studies, neo-nativists often offer interpretations that proceed from what might be called an argument from precocity. They argue that if one can demonstrate behaviors that are an index of a given concept (such as object permanence) at an age much earlier than reported in previous work, then the concept in question is likely to be innate. Such arguments are seriously flawed by the failure of researchers to assess the full range of variability involved in the developmental phenomenon. Evidence that infants look longer at the impossible versus possible display is interesting – even fascinating! An interpretation that such evidence indicates object permanence, however, suffers major flaws. Firstly, it implies that object permanence consists of a singular, monolithic, or abstracted ability, rather than a capacity that, like all developing skills, takes different forms over time. Secondly, the action taken to indicate object permanence – looking time – is very simple and requires very little responding by the infant. What the infant actually can do with the stimulus display is at best unclear. Thirdly, differences in looking are assessed in a complex task that richly supports the baby’s action (viz., a visuospatial field involving moving objects exhibiting Gestalt properties of good or bad form). Such rich perceptual information raises the question of whether such tasks measure perceptual processing rather than cognitive concepts.

Researchers are beginning to conduct studies that address these issues. A series of studies has examined the question of whether dishabituation to the impossible display in Baillargeon’s paradigm occurs because infants understand the possibility or impossibility of the events observed, or because of the relative novelty or familiarity of the stimulus display (Bogartz, 2000). In a series of studies with infants ranging from 4 to 8 months of age, researchers varied the extent to which infants were familiarized with the various possible and impossible displays used in their study. They reported evidence that infant looking time reflected preferences for novelty and familiarity, and did not indicate understanding of the possibility or impossibility of a given display. Regardless of whose findings are ultimately supported by future research, the main point is that skills emerge over time, not at a single point in development: for a reliable sense of the developmental course of any given skill, children’s actions must be assessed across a broad range of ages, behaviors, tasks, and assessment conditions.

Extending the paths: webs of representational development

As Table 1 (points III and IV) indicates, research conducted over the past three decades underscores the idea that unevenness in the emergence of skills is the rule rather than the exception in development. Skills from different conceptual domains develop relatively independent of each other, moving through their own developmental trajectories. Development takes place in a multidirectional web of pathways rather than a unidirectional ladder – a metaphor depicted in Figure 2. Developing skills do not move in a fixed order of steps in a single direction, but they develop in multiple directions along multiple strands that weave in and out of each other in ontogenesis. The developmental web portrays variability in developing skills within individuals, not only between them. For development in an individual child, different strands represent divergent pathways in the development of skills for different tasks or conceptual domains. For example, the development of addition and subtraction skills might occupy one strand, skills for producing stories another, and skills for reading words still another. As such, the developmental web provides a metaphor for understanding how different skills develop through diverging and converging pathways toward or away from different endpoints.

The developmental web, which provides a metaphor for understanding development in terms of multiple divergent and converging paths. Strands of development represent different skills in individual children, or different trajectories of the same skills in different children.

Figure 2. The developmental web, which provides a metaphor for understanding development in terms of multiple divergent and converging paths. Strands of development represent different skills in individual children, or different trajectories of the same skills in different children.

According to dynamic skill theory, there are no generalized or ‘all-purpose’ skills. As a result, a precise analysis of developmental changes in a skill can be performed only within particular conceptual domains, tasks, and social contexts. Figure 3 depicts pathways through which many American and Western European children pass as they construct representational skills for telling stories about nice and mean social interactions. The pathways depicted are meant to illustrate the types of trajectories through which skills at the representational level develop within a given task domain. While the pathways depicted may generalize to the development of related skills in similar children (e.g., stories involving similar content), no assumption is made that the sequences necessarily generalize beyond the specified task domain, social group, and assessment contexts. Typically, skills do not spontaneously generalize to different domains and contexts. Generalization of skills can and does occur, but it requires actions linking existing skills to new content, domains, and/or contexts.

Figure 3 was derived from a series of studies conducted by Fischer and his colleagues (Fischer & Bidell, 1998; Lamborn, Fischer, & Pipp, 1994) who assessed children’s story-telling skills in highly supportive contexts. To support children’s capacity to function at their optimal levels, using dolls and props, adults modeled stories of different levels of complexity about characters who acted nice or mean toward one another. Children then told or acted out the stories modeled by the adult. In Figure 3, each diagram of YOU or ME acting NICE or MEAN represents a story exhibiting a certain structure. In skill theory, beginning around 18-24 months of age, children gain the capacity to coordinate two sensorimotor systems into a system of sensorimotor systems, which is equivalent to a single representation (Sm4/Rp1). In a single representation, a child uses one sensorimotor system (e.g., moving a doll or uttering a word) to stand for or represent a single concrete meaning (e.g., the movement of a doll represents the act of walking). At the level of single representations, two-year-old children tell a story about a character who exhibits a single NICE or MEAN action. For example, a child makes a doll representing YOU or ME act nice (e.g., by offering candy to another doll) or mean (e.g., by knocking another doll down).

Stories about nice and mean interactions function as starting points for two trajectories in different (albeit related) interpersonal domains (nice and mean, respectively). At the second row, these two strands of development begin to come together as children gain the capacity to shift the focus of their attention between a single NICE representation and a single MEAN representation. At this step, a 2 – to 3-year-old tells part of a story about a nice interaction and another separate part of a story about a mean interaction. Note that separate NICE and MEAN representations are juxtaposed and not yet fully coordinated. A child, for example, makes one doll give candy to a second doll, and then, later in the story, makes one of the dolls act in a mean way, without connecting the two events.

With the stories in the third row, children can move in three different directions along the developmental web. All three directions involve skills at the level of representational mappings (Rp2), beginning around 3.5 to 4.5 years of age, in which children fully coordinate two representations in terms of a relationship such as reciprocity, causality, temporality, etc. Moving along the NICE strand, a child tells a story involving reciprocal nice actions (e.g., one character gives candy to another, who returns the favor with a hug). Similarly, moving along the MEAN strand, a child tells a story involving reciprocal mean actions. In the central strand, bringing NICE and MEAN together, a child tells a story in which a MEAN act is opposed by a NICE act, or vice versa. Within the level of representational mappings, children tell increasingly complex stories by stringing together multiple NICE or MEAN representations in different ways. This is represented by the story structures depicted at Steps 4, 5, and 6.

At Step 7, which arises around 6-7 years of age, children coordinate two representational mappings into a higher order representational system (Rp3). A child is able to construct a fully coordinated understanding of how one concrete relationship between NICE and MEAN maps onto another such concrete relationship. For example, a child acts out a story in which one doll invites another doll to play (NICE), while playfully slapping the other doll on the arm (MEAN); the other doll responds by accepting the invitation to play (NICE), but by sternly saying that the hitting must stop (MEAN). In this example, a child understands how one concrete relation (e.g., asking me to play while you teasingly hit me) is related to another (i.e., I’ll play but only if you stop hitting). The ability to construct representational systems underlies children’s capacities to perform various Piagetian tasks at the level of concrete operations (e.g., conservation, seriation, transitivity, etc.). Figure 3 specifies the structure of three types of stories involving NICE and MEAN interactions at the level of representational systems.

Developmental web for nice and mean social interactions. The numbers to the left of each set of brackets indicate the complexity ordering of the skill structures. The words inside each set of brackets indicate a skill structure.

Figure 3. Developmental web for nice and mean social interactions. The numbers to the left of each set of brackets indicate the complexity ordering of the skill structures. The words inside each set of brackets indicate a skill structure.  

As indicated in Table 2, around 10-11 years of age, pre-adolescents enter a new tier of development marked by the emergence of abstractions. Abstractions refer to generalized and intangible aspects of events, people, or things. Beginning around 10-11 years, in high support contexts, children gain the capacity to intercoordinate two concrete representational systems into a system of representational systems, which entails a single abstraction (Ab1). A child ‘abstracts’ across or generalizes what is common about at least two concrete descriptions of an event. For example, given two concrete stories depicting separate acts of kindness from one person to another (e.g., one child gives his lunch to another who forgot his; a student helps another study for a test when she wanted to go out and play), an 11-year-old can ‘abstract’ across these two stories and identify what is common to them: “Kindness is helping somebody in need, even if they can’t help in return.” Prior to this age, children have difficulty separating intangible from concrete aspects of events.

Beginning around 14-15 years of age, adolescents gain the capacity to coordinate two abstractions into an abstract mapping (Ab2). At this level, an individual can represent the relation between two abstract concepts. For example, an adolescent can understand the complex concept of a ‘social lie’ in terms of the contrast between ‘honesty’ and ‘kindness,’ or he might offer an explanation like “Honesty is telling the truth about something even when it is easier to lie. Kindness is helping someone in need. In a social lie, a person gives up honesty in order to be kind.” With further development, beginning around 18-20 years of age, young adults become capable of coordinating two abstract mappings into a single abstract system (Ab3). At this level, an individual can form an abstract conception of ‘constructive criticism’ in terms of the intercoordination of two abstract aspects of’honesty’ and two abstract elements of’kindness’: “Constructive criticism combines two types of honesty and two types of kindness. It involves being honest about praising another person’s accomplishments while criticizing his shortfalls. This expresses kindness as one helps to improve another person’s skills while being compassionate in not hurting his feelings.” At the final level, beginning at about 25 years of age, a person interprets multiple systems of abstractions in terms of general principles, such as a broad moral principle of treating others justly.

The epigenesis of psychological structures

As indicated in Table 1 (point V), theorists and researchers have criticized several aspects of Piaget’s constructivist view of developmental change. In this section, we describe an epigenetic model of cognitive change that builds upon Piaget’s model and addresses these issues. Traditionally, many have assumed that development proceeds as a result of two independent processes: heredity and learning. Recent theory and research has called into question the separability of genes and environments as causes of development. Researchers have examined development as an epigenetic process in which cognitive skills emerge over time rather than being predetermined by genes or transmitted from social experience. From an epi-genetic view, organisms function together with their environments as multi-leveled systems, the elements of which necessarily coact (affect each other) in the production of any set of actions, thoughts, or feelings.

One can differentiate among three broad levels of individual-environment processes. ‘Biogenetic systems’ refers to all biological systems below the level of the experiencing person. Biogenetic systems are multi-leveled with lower-order systems embedded in higher-order systems. For example, DNA is located in chromosomes, which themselves are hierarchically nested within cell nuclei, the cell matrix, cells, tissues, organs, organ systems, and the organism as a whole. ‘Individual-agentic systems’ refers to processes at the psychological level of action and experience – individual agents controlling actions. Such processes correspond to psychological systems of acting, thinking, and feeling and the control structures (skills) that regulate them. Although individual-agentic processes are themselves biological, they have emergent properties (e.g., meaning, control) that are absent in their lower-level constituents. Such systems function within larger sociocultural systems (the third level), which consist of patterns of interaction between persons and the shared symbolic meanings distributed among members of a given community. Symbol systems, particularly words, represent conventionalized meanings common to a linguistic community.

The main proposition of an epigenetic view holds that anatomical and psychological structures emerge as a result of coactions that occur over time both within and between these broad sets of systems. Coaction refers to the ways in which component systems act together mutually to regulate and influence each other’s functioning. Component systems coact both horizontally (e.g., gene with gene; cell-cell; organ-organ; organism-organism) and vertically (e.g., biogenetic with individual-agentic). For example, while gene action affects the functioning of the components of the cell, changes in the cell matrix also influence gene action. The direction of influence of component systems is dynamical and bidirectional rather than linear or unidirectional. The following contains a brief outline of how biogenetic, individual-agentic, and sociocultural systems necessarily coact in the development of cognitive skills.

Biogenetic processes

Developmental changes in biogenetic systems are necessary for the emergence of new levels of cognitive skill. Recent research suggests that brain development exhibits discontinuities that are related to the emergence of new psychological skills. Research on the development of cortical (electroencephalogram or EEG) activity, synaptic density, and head growth provides evidence for discontinuities in brain development for at least twelve of the thirteen levels of skill development listed in Table 2. Little research exists to test hypothesized brain-behavior relations for the thirteenth level. For example, in a series of studies, American, Swedish, and Japanese infants demonstrated spurts in brain growth (EEG and head circumference) at approximately 3-4, 7-8, 10-11, and 15-18 weeks of development and at 2-4, 6-8, 11-13, and 24 months. As indicated in Figure 4, studies measuring EEG activity in various cortical regions show discontinuities at approximately 2, 4, 8, 12, 15, and 19 years of age. There are many different brain systems and different classes of behavior that develop relatively independently of each other, but research suggests patterns of concurrent growth over time in related systems, especially during rapid growth of new skills. For example, Ann-Marie Bell and Nathan Fox assessed the relations between growth functions for EEG activity and the development of object search, vocal imitation, and crawling skills in infancy. They found that for many individual infants between 8 and 12 months of age, connections between specific cortical regions involving planning, vision, and control of movement exhibited a surge while the infants were mastering crawling. The surge disappeared after they had become skilled crawlers.

Within the representational and abstract tiers of development, transformation from one level of skill to another (e.g., from single sets to mappings, etc.) seems to be supported by the production of new systems of neural networks that link different brain regions. Matousek & Petersen (1973) examined changes in EEG activity for each of four cortical brain regions (viz., frontal, occipital/parietal, temporal, and central) in children and adolescents. Their results suggested that for the representational (2-10 years of age) and abstract tiers (10-20 years), transitions to different levels within a developmental tier are marked by cyclic changes in brain activity in different cortical regions. Within this cycle, a new tier emerges with a maximal spurt in the frontal cortex: the first level is marked by a maximal spurt in the occipital-parietal region, one in the temporal region marks the second level, and one in the central region marks the third. Another maximal surge in the frontal region marks the onset of the next broad tier of development. These changes illustrate the systematic relations between movement through skill levels and cyclic changes in brain activity.

Development of relative power in alpha EEG in occipital-parietal area in Swedish children and adolescents. Relative power is the amplitude in microvolts of absolute energy in the alpha band divided by the sum of amplitudes in all bands.

Figure 4. Development of relative power in alpha EEG in occipital-parietal area in Swedish children and adolescents. Relative power is the amplitude in microvolts of absolute energy in the alpha band divided by the sum of amplitudes in all bands.

Individual-agentic processes

Biogenetic changes are necessary but not sufficient for the emergence of new skill levels. While any given level of skill requires a requisite level of brain development, skill development also requires action at the level of individual children. For novel skills to develop, individual children must perform controlled acts that coordinate lower-level skill components into higher-order structures. Fischer and his colleagues have identified a series of transformation rules that describe the active processes by which children coordinate skill elements into higher-order structures. Substitution occurs as children perform a previously constructed skill on a novel object. For example, an infant who has acquired the skill of grasping a rattle might use this new skill to grasp a small teddy bear. Shift-of-focus occurs when children redirect their attention from one component of skilled activity to another. Children often use shift-of-focus to reduce task demands. In such contexts, children break down complex tasks into simpler sub-tasks and then shift the focus of their attention from one task element to another. For example, 3-year-olds, who are generally unable to imitate a story modeled for them at the level of representational mappings (e.g., Bobby acts mean to Sally because Sally was mean to him), can simplify the story by breaking it into two separate parts and directing attention first to one part (Sally is mean to Bobby), and then the other (Bobby is mean to Sally). Such shifts function as initial attempts to integrate skill components beyond a child’s immediate grasp.

Children can use compounding to construct more complex skills within a given developmental level. A child links a series of skill elements at the same developmental level. For example, 5-year-olds often tell stories by combining in a single story characters that act out distinct roles, such as doctor, patient, and nurse. Compounding involves integration among skill elements without re-organizing them to form a higher-level skill. The only change mechanism that produces movement to a new level or developmental tier is intercoordination. Using intercoordination, children go beyond merely compounding together skill elements. Instead, skill elements become reciprocally coordinated to form a single higher-level skill. For example, as indicated in Figure 3, when telling stories about nice and mean interactions, children move from row 2 (shift-of-focus between representations) to row 3 (representational mappings) by connecting two representations in terms of some type of relation (e.g., “Bobby gave Sally a kiss because Sally gave him her lunch”). In this way, higher-order skills emerge from the intercoordination of lower-order components.

Another important process by which individuals create new knowledge involves a process called bridging (Granott, Fischer, & Parziale, 2002). Bridging arises from the capacity to function simultaneously at two developmental levels in a one-skill domain. It occurs when individuals establish a target level of skill and then direct their knowledge construction toward that target. In so doing, the to-be-constructed level of understanding functions as a shell for constructing a new level of knowledge. The shell helps to bridge one level to a higher level of understanding. For example, in a study in which a pair of adults observed the operation of self-moving robots in order to figure out how they worked, one person experimented with a robot by putting his hand around it in different positions. His partner noted: “Looks like we got a reaction there.” In using the word reaction, the partner made a vague reference to cause and effect, but did not provide specifics about the cause or effect. By speaking of a reaction, he created a bridging shell postulating a link between two unknown variables, X and Y, related to each other:


Further observations allowed the observers to establish a causal connection that filled in the shell to create a mapping (i.e., when the robot moves under a shadow, its behavior changes):


As such, shells function as a kind of self-scaffold that helps to explain how individuals bootstrap their knowledge to new developmental levels based upon existing lower-level knowledge.

Sociocultural processes

While individual children must actively coordinate lower-level components to produce new skills, they act in a rich sociocultural environment, not in a vacuum. Interactions with others play a direct role in the formation of psychological structures in at least two ways. Firstly, in face-to-face social interaction, partners engage in continuous reciprocal communication. In so doing, both partners are simultaneously active as senders and receivers of meanings. As a result, they continuously adjust their ongoing actions, thoughts, and feelings to each other. In so doing, neither partner exerts complete control over his behavior, but instead they co-regulate each other’s actions. In this way, social partners function as actual parts of each other’s behavior.

Secondly, in constructing new knowledge, children work with others using cultural meanings, tools, and artifacts. As children’s thinking becomes mediated by cultural tools – particularly language and other symbolic vehicles that represent shared cultural meanings – their thinking develops in directions defined by cultural meanings and practices. In this way, sociocultural systems play an active role in the constitution of cognitive skills.

Co-regulated interaction with children, especially when it involves sign-mediated guidance by adults or more accomplished peers, raises children’s thinking to levels that they would be incapable of sustaining alone. It is from this social matrix that children construct novel skills. Rogoff (1998) has offered the concept of participatory appropriation to refer to the ways in which individuals seize novel meanings from their participation in social interaction. Appropriation occurs as children coordinate lower-level meanings into higher-order skills in ways that are structured by their interactions with social and cultural agents. As such, appropriation involves more than simply incorporating other people’s meanings. When a child appropriates meanings from her interactions with others, she transforms those meanings in ways that are biased by her existing skills and meanings.

This process can be illustrated by a study on the development of a child’s causal understanding in the context of parent to child story telling. The study traced the production and resolution of question-answer sequences as a parent read a 4-year-old boy the same children’s topic over the course of six evenings. In four of the six sessions, the boy asked questions about a part of the story in which a character (Pig Won’t) caused soap suds to splash in his eye and sting it. In the second session, the following dialogue occurred: child: Why does it sting? parent: Well because when soap gets on your, in your eyes, it stings.

In this reading, the parent responds to the child’s question by drawing a causal relationship (a representational mapping) between the action of the soap (“When soap gets in your eyes”) and a consequential result (“it stings”). The question of stings was next raised in the third reading session: child: Why does it sting? parent: Why DOES it sting? child: ‘Cause it hurts. parent: Soap hurts your eyes.

In this sequence, we again see the child trying to represent the cause of the stinging. When the parent prompted the child to elaborate his own thinking, the child responded at the level of single representations (“‘Cause it hurts”). The child has not yet differentiated the cause from the effect in this concrete incident. By the sixth session, the following interchange occurred:

child: That’s sad.

parent: Why is that sad?

child: It’s sad that Pig Will has stings.

parent: Yeah? Why? Why is it sad that it stings?

child: Because when soap gets in your eyes it stings. parent: Right.

child: When bumble bee stings, me. Does that stings?

parent: When bumble bees sting do they sting? You betcha’ it does. child: Why?

parent: Well when bumble bees put their little stinger inside of you, it pierces the skin, and it stings.

Here, the child responds to the adult’s question with virtually the same representational mapping that was produced by the adult in earlier sessions. These exchanges illustrate a series of points about the role of sociocultural systems in development. Firstly, the exchanges involve co-regulated interactions that function to raise the child’s understanding to levels that he could not sustain alone. Secondly, the interaction is mediated by cultural tools, practices, and artifacts (e.g., words, causal story content, bedtime reading rituals, use of topics). Thirdly, through the verbal exchanges, the child is able to appropriate a novel meaning. He does so by coordinating lower-level skill elements (single representations) into a higher-order meaning (representational mapping). Thereafter, the child initiates an act of substitution involving the application of his new causal knowledge about ‘stinging’ to new content (i.e., bee stings). These exchanges illustrate how a child’s novel meanings are jointly created but individually appropriated though individual acts of hierarchic integration.


The new constructivism in cognitive development builds upon central tenets of Piaget’s thinking. Cognitive development involves qualitative and quantitative changes in psychological structures. However, the new constructivism maintains that transformations in psychological structures are tied to specific tasks, domains, and social contexts. While retaining the principle that individual action functions as a central organizer of cognitive change, the new constructivism more fully embraces the joining of biological, psychological, and sociocultural processes as coacting causes of cognitive change.

Future research is needed to address several questions. Firstly, to the extent that skills are defined in terms of tasks and domains, what are the boundaries of developing skills? Is it possible to specify a relatively distinct set of psychological skills that cluster together in development for a particular domain? Secondly, given that change processes involve bidirectional coactions among vertical and horizontal dimensions of organism-environment systems, how do these coactions move development? How do different component systems affect each other? Thirdly, although not discussed at length in this entry, the new constructivism is built around the premise that cognitive development is inextricably linked to socioemotional development. Emotion is a central organizer of all behavior, intentional or otherwise. Thus, a major contemporary question is, “How do emotional and cognitive processes coact in the creation of developmental pathways?”

Next post:

Previous post: