Dynamic-Systems Theories

Like all biological processes, thinking serves an adaptive purpose: it enables people and other animals to devise plans for attaining goals. However, attaining goals also requires the ability to take effective action; without this ability, thinking would be pointless. For example, what purpose would planning serve for an organism that could not implement the plans through action (see Figure 4.11)? Despite this inherent connection between thinking and acting, however, most theories of cognitive development have ignored the development of the skilled actions that allow children to realize the fruits of their mental labor.

FIGURE 4.11 Problem solving often requires motor skills A major insight of dynamic-systems theories is that thinking would be pointless without motor capabilities. In these photos, a 12-month-old is shown knocking a barrier out of the way (left frame) and then grasping the edge of a cloth in order to pull the cloth, string, and toy toward him (right frame). If the infant lacked the motor dexterity to grasp the cloth or the strength to pull in the toy, his problem-solving processes would have been fruitless.

COURTESY OF PETER WILLATTS, UNIVERSITY OF DUNDEE, SCOTLAND

One increasingly influential exception to this generalization is dynamic-systems theories, a class of theories that focuses on how change occurs over time in complex systems. Research that reflects the dynamic-systems perspective indicates that detailed analyses of the development of infants’ basic actions, such as crawling, walking, reaching, and grasping, yield surprising and impressive insights into how development occurs. For example, dynamic-systems research has shown that improved reaching allows infants to play with objects in more advanced ways, such as organizing them into categories or interesting configurations (Spencer et al., 2006; Thelen & Corbetta, 1994). Dynamic-systems research also has shown that the onset of crawling changes infants’ relationships with family members, who may be thrilled to see their baby attain an important motor milestone but also find themselves having to be much more watchful and controlling to avoid harm to the child and to the objects in the child’s path (Campos, Kermoian, & Zumbahlen, 1992).

Another contribution of dynamic-systems research has been to demonstrate that the development of seemingly simple actions is far more complex and interesting than previously realized. For example, such research has overturned the traditional belief that physical maturation leads infants to attain motor milestones in stages, at roughly the same age, in the same way, and in a steady progression. It has shown instead that individual children acquire skills at different ages and in different ways, and that their development entails regressions as well as progress (Adolph & Berger, 2011).

One example of this type of research is a longitudinal study of the development of infants’ reaching conducted by Esther Thelen, who, along with her colleague Linda Smith, was the cofounder of the dynamic-systems approach to cognitive development. In this particular study, Thelen and colleagues (1993) repeatedly observed the reaching efforts of four infants during their first year. Using high-speed motion-capture systems and computer analysis of the infants’ muscle movements, they found that because of individual differences in such factors as the infants’ physiology, activity level, arousal, motivation, and experience, each child faced different challenges in his or her attempts to master reaching. The following observations illustrate some of the complexities these researchers discovered, including variability in the ages at which infants reach developmental milestones, their patterns of change, and the differing challenges they must overcome:

These descriptions help to convey what is meant by the label “dynamic systems.” As suggested by the term dynamic, dynamic-systems theories depict development as a process in which change is the only constant. Whereas some approaches to cognitive development hypothesize that development entails long periods of relatively stable stages or ways of thinking separated by relatively brief transition periods, dynamic-systems theories propose that at all points in development, thought and action change from moment to moment in response to the current situation, the child’s immediate past history, and the child’s longer-term history of actions in related situations. Thus, Thelen and Smith (1998) noted that the development of reaching included regressions as well as improvements, and Thelen (2001) described how differences in Hannah’s and Gabriel’s early reaches influenced their later paths to skilled reaching.

As suggested by the second term in the label, dynamic-systems theories depict each child as a well-integrated system, in which many subsystems—perception, action, attention, memory, language, social interaction, and so on—work together to determine behavior. For example, success on tasks viewed as measures of conceptual understanding, such as object permanence, is influenced by perception, attention, motor skills, and a host of other factors (Smith et al., 1999). The assumptions that development is dynamic and that it functions as an organized system are central to the theory’s perspective on children’s nature.

View of Children’s Nature

Dynamic-systems theories are the newest of the four types of theories discussed in this chapter, and their view of children’s nature incorporates influences from each of the others. Like Piaget’s theory, dynamic-systems theories emphasize children’s innate motivation to explore the environment; like information-processing theories, they emphasize precise analyses of problem-solving activity; and like sociocultural theories, they emphasize the formative influence of other people. These similarities to other theories, as well as the differences from them, are evident in dynamic-systems theories’ emphasis on motivation and the role of action.

Motivators of Development

To a greater extent than any of the other theories except Piaget’s, dynamic-systems theories emphasize that from infancy onward, children are strongly motivated to learn about the world around them and to explore and expand their own capabilities (von Hofsten, 2007). This motivation to explore and learn is clearly apparent in the fact that children persist in practicing new skills even when they possess well-practiced skills that are more efficient. Thus, toddlers persist in their first unsteady efforts to walk, despite the fact that crawling would get them where they want to go more quickly and without the risk of falling (Adolph & Berger, 2011).

Like sociocultural approaches, but unlike Piagetian theory, dynamic-systems theories emphasize infants’ interest in the social world as a crucial motivator of development. As noted in our discussion of the active child in Chapter 1, even newborns prefer attending to the sounds, movements, and features of the human face over almost any alternative stimuli. By 10 to 12 months of age, infants’ interest in the social world is readily apparent in the emergence of intersubjectivity (page 159), as infants look to where the people interacting with them are looking and direct the attention of others to things they themselves find interesting (Deák, Flom, & Pick, 2000; von Hofsten, Dahlström, & Fredricksson, 2005). Dynamic-systems theorists have emphasized that observing other people, imitating their actions, and attracting their attention are all potent motivators of development (Fischer & Bidell, 2006; von Hofsten, 2007).

The Centrality of Action

Reaching with Velcro-covered mittens for Velcro-covered objects improved infants’ later ability to grab and explore ordinary objects without the mittens.

© KLAUS LIBERTUS

Dynamic-systems theories are unique in their pervasive emphasis on how children’s specific actions shape their development. Piaget’s theory asserts the role of actions during infancy, but dynamic-systems theories emphasize that actions contribute to development throughout life. This focus on the developmental role of action has led to a number of interesting discoveries. For example, infants’ own reaching for objects helps them infer the goals of other people’s reaches; infants who can skillfully reach are more likely to look at the probable target of another person’s reaching just after the other person’s reach begins (von Hofsten, 2007). Another example of infants’ learning from actions comes from research in which infants were outfitted with Velcro mittens that enabled them to “grab” and explore Velcro-covered objects that they otherwise could not have picked up. After 2 weeks of grabbing the Velcro-covered objects with the Velcro-covered mittens, infants showed greater ability to grab and explore ordinary objects without the mittens than did other infants of the same ages (Needham, Barrett, & Peterman, 2002).

The ways in which children’s actions shape their development extend well beyond reaching and grasping in infancy. Actions influence categorization: in one study, encouraging children to move an object up and down led to their categorizing it as one of a group of objects that were easiest to move in that way, whereas encouraging children to move the same object from side to side led them to categorize it as one of a group of objects that were easiest to move in that way (Smith, 2005). Actions also affect vocabulary acquisition and generalization (Gershkoff-Stowe, Connell, & Smith, 2006; Samuelson & Horst, 2008): for example, experimental manipulations that lead children to state an incorrect name for an object impair the child’s future attempts to learn the object’s correct name. In addition, actions shape memory, as demonstrated by research in which children’s past attempts to locate and dig up objects they had earlier seen being hidden in a sandbox altered their recall of the objects’ new location after they had seen them being re-hidden. That is, the child’s new searches tend to be in-between the past and present locations, as if the new searches involved a compromise between memories of the new hiding place and of the location where he or she had originally looked (Schutte, Spencer, & Schöner, 2003; Zelazo, Reznick, & Spinazzola, 1998). Thus, just as thinking shapes actions, actions shape thinking.

Central Development Issues

Two developmental issues that are especially prominent in dynamic-systems theories are how the cognitive system organizes itself and how it changes.

Self-Organization

Dynamic-systems theories view development as a process of self-organization that involves bringing together and integrating attention, memory, emotions, and actions as needed to adapt to a continuously changing environment (Spencer et al., 2006). The organizational process is sometimes called soft assembly, because the components and their organization change from moment to moment and situation to situation, rather than being governed by rigid stages that are consistently applied across time and situations.

The types of research to which this perspective leads are illustrated particularly well by certain studies of the A-not-B error that 8- to 12-month-olds typically make in Piaget’s classic object-permanence task. As noted earlier, this error involves infants’ searching for a toy where they had previously found it (location A), rather than where they last saw it being hidden (location B). Piaget (1954) explained the A-not-B error by hypothesizing that before their 1st birthday, infants lack a clear concept of the permanent existence of objects.

In contrast, viewing the A-not-B error from a dynamic-systems perspective suggested that many factors other than conceptual understanding influence performance on the object-permanence task. In particular, Smith and colleagues (1999) argued that babies’ previous reaching toward location A produces a habit of reaching there, which influences their behavior when the object is subsequently hidden at location B. On the basis of this premise, the researchers made several predictions that were later borne out. One was that the more often babies had found an object by reaching to one location, the more likely they would be to reach there again when the object was hidden at a different location. Also supported was the prediction that increasing the memory demands of the task by not allowing infants to search for the object for 3 seconds after it was hidden at the B location would increase the likelihood of infants’ reaching to location A (Clearfield et al., 2009). The reasoning here was that the strength of the new memory would fade rapidly relative to the fading of the habit of reaching to the previous hiding place. The dynamic-systems perspective also suggested that infants’ attention would influence their object-permanence performance. Consistent with this view, manipulating infants’ attention by tapping one of the locations just as the infants were about to reach usually resulted in their reaching to the tapped location, regardless of where the object was actually hidden.

In perhaps the most striking test of such predictions, researchers demonstrated that putting small weights on infants’ wrists after the infants had reached to location A but before the object was hidden at location B improved object-permanence performance (Diedrich et al., 2000). The researchers had reasoned that the addition of the wrist weights would require the infants to use different muscle tensions and forces than they had previously used to reach for the object and consequently would disrupt the infants’ habit of reaching to location A. Thus, rather than providing a pure measure of conceptual understanding, infants’ performance on the object-permanence task appears to reflect the combined influence of the strength of the habit of reaching to location A, the memory demands of the current task, the infant’s current focus of attention, and the match between the muscular forces required to reach in the old and new situations.

How Change Occurs

Dynamic-systems theories posit that changes occur through mechanisms of variation and selection that are analogous to those that produce biological evolution (Fischer & Biddell, 2006; Steenbeek & Van Geert, 2008). In this context, variation refers to the use of different behaviors to pursue the same goal. For example, to descend a ramp, a toddler will sometimes walk, sometimes crawl, sometimes do a belly slide, sometimes do a sitting slide feet first, and so on (Adolph, 1997; Adolph & Berger, 2011). Selection involves increasingly frequent choice of behaviors that are effective in meeting goals and decreasing reliance on less effective behaviors. For example, when children first learn to walk, they are too optimistic about being able to walk down ramps, and often fall, but when they gain a few more months of walking experience, they more accurately judge the steepness of ramps and whether they can descend them while remaining upright.

Children’s selection among alternative approaches reflects several influences. Most important is the relative success of each approach in meeting a particular goal: as children gain experience, they increasingly rely on approaches that produce desired outcomes. Another important consideration is efficiency: children increasingly choose approaches that meet goals more quickly or with less effort than do other approaches. A third consideration is novelty, the lure and challenge of trying something new. Children sometimes choose new approaches that are no more efficient, or even less efficient, than an established alternative but that have the potential to become more efficient. For example, when they first learn the memory strategy of rehearsal, it does not improve their memory, but they use it anyway, and eventually it does improve their recall of rehearsed information (Miller & Seier, 1994). Such a novelty preference tends to be adaptive, because with practice, a strategy that is initially less efficient than existing approaches often becomes more efficient (Wittmann et al., 2008). As discussed in Box 4.4, the insights of dynamic-systems theories have led to useful applications as well as theoretical progress.