History of Perception-Action Approach

Theoretical Foundation

The study of theories that support treatment methods is useful for clinicians because theory informs not only why the treatment works but how to modify treatment for optimal results.

The Perception-Action (P-A) Approach is based upon three modern theories, perception-action theory1, dynamic systems theory,and theory of neuronal group selection.3,4 These theories are consistent with one another and complement each other in their explanation of the developmental process and formation of new movement.5

The Perception-Action Theory came to light in the 1960’s with the work of James and Eleanor Gibson, psychologists studying perception.1,6-8 This ecological perspective describes the “changing relationship between a habitat (an animal’s immediate environment) and a niche (‘making a living’ in a habitat)”.9

Perception and action are intricately tied to one another and are driven by a process of spontaneous exploration or motor activity toward a goal.1,10,11 Perception is an active process of seeking information about the environment through all available sensory channels. For this reason, James Gibson looked at human senses as perceptual systems.1 Perception guides one’s actions. Actions are necessary for gathering the perceptual information that, in turn, guides action. In addition, actions provide information about one’s own body, movement and capabilities.1,10,11 This has been termed a perception-action loop,12 with both parts seeming to occur simultaneously.5,12

For example, perception and action are occurring simultaneously in support of postural control.5 It is known that both adults and children have continuous postural sway even when standing still.13 It is postulated that the background swaying motion is providing continuous perceptual information that may be particularly useful in infancy.13 This information and the associated movements or actions allow the individual to maintain the same postural reference, to change it, or to prepare to move for another purpose.

If, again, we use the example of postural control, information from multiple perceptual systems is needed. The sensitivity of the somatosensory system (touch and pressure) is thought to be greater than that of the visual or vestibular systems.14,15 Small changes in position and rate of postural sway are easily detected. The P-A Approach takes advantage of this sensitivity.16 Therapists trained and experienced in this approach are able to provide manual guidance that alters the information entering the somatosensory system. This therapeutic guidance does not conflict with the patient’s current postural control, but may add perceptual information that aids the patient in creating better posture or a new movement. In this way, the patient’s ideas and motivations for moving are respected while he or she is provided an opportunity to change a movement pattern.16

This approach also employs the concept of affordances developed by Gibson.1,12 Affordances are simply aspects of the environment that provide a possibility for action. Affordances are considered with reference to the attributes of the individual.12 For instance, a dining room chair affords sitting down to dinner for a parent, but it may only be a place to stand or to crawl under for an infant. Thus, selections of furniture or other supports are carefully made during a treatment session.

The Dynamic Systems Theory (DST) was introduced by Edward Lorenz.17 In 1972 he titled a talk: Predictability: Does the Flap of a Butterfly's Wings in Brazil set off a Tornado in Texas?18 The Lorenz attractor modeled atmospheric conditions, specifically convection currents.17 The DST is applied to the chaotic or non-linear behavior of complex systems across many domains, such as social systems, physiologic processes, and movement.19 Esther Thelen and colleagues brought a large body of work to the understanding of child development by viewing development itself as a dynamic system.20 In this approach to child development, they challenged the traditional view of motor development in which the maturation of the central nervous system was seen as the primary agent of change. Instead, from a DST perspective, they proposed that development was multi-causal, loosely assembled and driven by many components or subsystems.20

A clever longitudinal study conducted by Thelen and Ulrich21 helps to illustrate several significant concepts of DST as applied to motor development.20 Infants were studied monthly from the age of 1 month, for 7-10 months.21 These infants did not demonstrate stepping behavior when supported with their feet on a stationary surface but did initiate steps when held upright on a moving treadmill. Over time, as the study continued, the treadmill stepping behavior changed, with a pronounced increase in stepping observed at about 3 months of age and the exact timing of the increase varying between infants. In this longitudinal study, Thelen and Ulrich21 demonstrated several constructs of the DST which are discussed below.

Spontaneous Self-Organization

Self-organization is a principle of the DST that refers to multiple subsystems spontaneously coming together as equal contributors to a behavioral change of the entire system.2,5,21 In the study described above,21 as a result of providing a context or task that brought together multiple components (subsystems) needed for stepping, infants that did not take steps on a stationary surface were able to organize their stepping behavior on the moving treadmill belt. Previously “hidden skills” were “uncovered” by the treadmill.21 It was the context of the treadmill that allowed the system to self-organize and produce a behavior it had not previously demonstrated.

Nonlinearity and Control Parameters

Thelen and Ulrich21 showed that stepping behaviors occur at different times and rates, varying in different infants. The developmental process is nonlinear, which means that a dramatic change in the behavior of the system can be observed with a disproportionately small but critical change in a component of one of its subsystems (musculoskeletal, neurological, cardiovascular and pulmonary, vestibular, and somatosensory).22,23 Dynamic systems are sensitive to control parameters, which are the factors or variables that produce a significant change in the system’s behavior.22-24 Changes in body weight, body proportions, muscle strength, etc., serve as control parameters at different points in time, leading to a change in behavior that is not the result of a direct cause and effect relationship.21 As Thelen and Smith2 stated, “everything matters.” In the study discussed above,21 the treadmill was the context and its speed was the control parameter that drove the system to self-organize. The stepping behavior not present on a stationary surface emerged. An increase in muscle strength was another control parameter. Once it achieved a critical value over time, it changed the infants’ stepping behavior on the treadmill.21


Complex systems are multidimensional, comprised of any number of subsystems.2,20 Interaction of these multiple subsystems can lead to a change in pattern or emergence of a new behavior of the system.2,20 In the study described above,21 body weight, dimensions and muscle strength, the components of the musculoskeletal subsystem, were the control parameters that influenced postural control and changed the infants’ stepping behavior in the context of a moving treadmill. Early patterns of stepping were unstable and variable. Over time, with increased strength, control over multiple body segments, and improved postural control, a reorganization of the pattern of stepping occurred through self-organization of multiple subsystems of the body around the task at hand. Although a variety of stepping patterns and the organization of the patterns changed over time, months of exploration were needed for walking to emerge from earlier stepping behavior.21

Time Scales

The DST looks at time in several different ways, all of which are important: a single observed behavior happens “in the moment”, but the child makes many behavioral decisions over a longer period of time.20 A chain of these decisions results in a developmental change, and every change serves as a foundation for the new changes that emerge later.20 In their longitudinal study, Thelen and Ulrich21 demonstrated the importance of the time scales. Within the same context of being held upright on a moving treadmill, the infants’ stepping behavior changed over time from unstable patterns of fewer, less organized, irregular, “single, parallel and double steps” to a more organized, stable, regular pattern of stepping. Thelen and Ulrich21 “concluded that stepping and, ultimately, walking are not innate or prescribed. Rather, they are self-organized and emergent, reflecting the assembly of multiple subsystems within the infant’s history of activity in context.”20

The Theory of Neuronal Group Selection (TNGS) or neural Darwinism was described by Gerald Edelman who argued against traditional concepts of a fixed nervous system that operates by computing, feedback loops and information processing.3,4,25 Instead, he viewed the nervous system as continuously modifying itself according to the adaptive value of the activity to the individual.3,4,25 Early in life, one forms a primary neuronal repertoire of structurally diverse groups of neurons that are widely distributed and variably connected.3,4 The groups are formed by a combination of genetic information, physiologic activity, and motor activity.3,4

The selection of neuronal groups for adaptive activity is made possible by movement exploration and forms the secondary neuronal repertoire.3,4,25 When the infant explores the possibilities for action in relation to a particular context, many neuronal groups respond, in varied combinations. The combination that provides the infant’s best available solution at that time would be selected and strengthened, and would be likely to be used again. Further exploration, as well as growth and changing skills, yield more variations in motor behavior, building the repertoire of movement.3,4,25

The perception-action theory, the DST and the TNGS were described by Esther Thelen as complimentary theories relating to the developmental process.2,5 The perception-action theory describes the functional interactions between the environment and the individual’s adaptive activity.1,10,11 The DST describes how movement patterns change with respect to their spatial and temporal arrangement.2,5 Variable and complex patterns are desirable.2,5 The TNGS captures the essence of the contributions of the nervous system with regard to changing motor behaviors.3,4 Movement variability and complexity are hallmarks of movement for typically developing individuals.26-28 This notion can be explained using all three discussed theories by highlighting the process of finding variable solutions to motor tasks related to constant changes in the dynamic interaction between the individual and his or her environment, changing parameters and constraints of a task at hand, and changes in proportions of the body of a growing and developing child.1-5 To illustrate, having multiple ways of approaching a task gives one many choices or variations when the task changes slightly or some factor within the individual changes. The ability of the individual to select not only the most efficient motor pattern for the current condition, but also a different pattern when the situation changes, indicates presence of typical movement components of flexibility and adaptability.28 Complexity in movement comes with experience as a skill develops. For instance, a child first learning to drink from a cup may vary a great deal the degree of tipping. Many errors occur. The milk has a wide splatter pattern. With experience, the tipping is refined, spills of the sort described above decrease, but more variations are introduced. The maximum tip may be prolonged while gurgling sounds are made, repetitive tips can be made for sipping, or tipping can be performed with one hand. Spills and splatters of other sorts are likely. Advancements in cup handling such as these represent complexity in motor behavior and serve to enrich the child's experience.

Another common feature of the perception-action theory, the DST and the TNGS is the emphasis on the important role cognition plays in developmental change as cognition is inseparably linked to perception and action.1,4,5,11,20,25,29 The significance of this connection was first acknowledged by the perception-action theory;1,11 the DST described it using the concept of embodiment (the existence of cognition is inseparable from perception and action or movement);5,20,29 and the TNGS used the term “adaptive value” to explain the process of selection of a new movement pattern during spontaneous exploration.25 The P-Action Approach to intervention used with children with movement disorders represents a clinical application of the perception-action theory, the DST and the TNGS. This approach employs skillful manual guidance and environmental set-up techniques grounded in an integrated view of these theories, promotes spontaneous exploration, and emphasizes movement variability, complexity and adaptability as the goal of intervention.16,30
  1. Gibson J. The Senses Considered as Perceptual Systems. Boston, MA: Houghton Mifflin; 1966.
  2. Thelen E, Smith LB, eds. A Dynamic Systems Approach to the Development of Cognition and Action. Cambridge, MA: MIT Press; 1994.
  3. Edelman GM. Neural Darwinism. New York, NY: Basic Books; 1987.
  4. Edelman GM. Neural Darwinism: selection and reentrant signaling in higher brain function. Neuron.1993;10(2):115-125.
  5. Thelen E. Motor development. A new synthesis. Am Psychol. 1995;50(2):79-95.
  6. Gibson JJ, Pick AD, Osser H. A developmental study of the discrimination of letter-like forms. J Comp Physiol Psychol. 1962;55:897-906.
  7. Gibson EJ. Perceptual learning. Annu Rev Psychol. 1963;14:29-56.
  8. Gibson EJ. Development of perception: discrimination of depth compared with discrimination of graphic symbols. Monogr Soc Res Child Dev. 1963;28(2):5-32.
  9. Goldfield E. Emergent Forms: Origins and Early Development of Human Action and Perception. Oxford, NY: Oxford University Press; 1995.
  10. Gibson JJ. Principles of Perceptual Learning and Development. New York, NY: Appleton-Century-Crofts, 1979.
  11. Gibson EJ. Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. Annu Rev Psychol. 1988;39:1-41.
  12. Gibson JJ. The Ecological Approach to Visual Perception. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.; 1979.
  13. Chen LC. The Development of Adaptive Sensorymotor Control in Infant Upright Posture [dissertation]. College Park: University of Maryland; 2007.
  14. Jeka J, Oie K, Schöner G, Dijkstra T, Henson E. Position and velocity coupling of postural sway to somatosensory drive. J Neurophysiol. 1998;79(4):1661-1674.
  15. Fitzpatrick R, McCloskey DI. Proprioceptive, visual and vestibular thresholds for the perception of sway during standing in humans. J Physiol. 1994;478(Pt 1):173-186.
  16. Tscharnuter I. Clinical application of dynamic theory concepts according to Tscharnuter Akademie for Movement Organization (TAMO) therapy. Pediatr Phys Ther. 2002;14:29-37.
  17. Lorenz EN. Deterministic nonperiodic flow. J Atmos Sci. 1963;20(2):130–141.
  18. Lorenz EN. Predictability: Does the Flap of a Butterfly's Wings in Brazil set off a Tornado in Texas? Talk presented at: American Association for the Advancement of Science, 139th Meeting; December 29, 1972; Boston, MA. Accessed September 23, 2012.
  19. Gleick J. Chaos Making a New Science. New York, NY: Viking Penguin, Inc.; 1987.
  20. Spencer JP, Clearfield M, Corbetta D, Ulrich B, Buchanan P, Schöner G. Moving toward a grand theory of development: in memory of Esther Thelen. Child Dev. 2006;77(6):1521-1538.
  21. Thelen E, Ulrich BD. Hidden skills: a dynamic systems analysis of treadmill stepping during the first year. Monogr Soc Res Child Dev. 1991;56(1):1-98.
  22. Latash ML. The Bernstein problem: how does the central nervous system make its choices? In: Latash ML, Turvey MT, eds. Dexterity and Its Development. Mahwah, NJ: Lawrence Erlbaum Associates, Publishers; 1996:277-303.
  23. Harbourne RT, Stergiou N. Movement variability and the use of nonlinear tools: principles to guide physical therapist practice. Phys Ther. 2009;89(3):267-282.
  24. Kamm K, Thelen E, Jensen JL. A dynamical systems approach to motor development. Phys Ther. 1990;70:763-775.
  25. Sporns O, Edelman G. Solving Bernstein’s problem. Child Dev. 1993;64(4):960-981.
  26. Fetters L. Perspective on variability in the development of human action. Phys Ther. 2010;90:1860-1867.
  27. Vereijken B. The complexity of childhood development: variability in perspective. Phys Ther. 2010;90:1850-1859.
  28. Dusing SC, Harbourne RT. Variability in postural control during infancy: implications for development, assessment, and intervention. Phys Ther. 2010;90:1838-1849.
  29. Thelen E. Grounded in the world: developmental origins of the embodied mind. Infancy. 2000;1(1):3-28.
  30. Harbourne RT, Willett S, Kyvelidou A, Deffeyes J, Stergiou N. A comparison of interventions for children with cerebral palsy to improve sitting postural control: a clinical trial. Phys Ther. 2010;90:1881-1898
Empty space, drag to resize
Empty space, drag to resize

Learn with IPAA!
We offer On-demand Introductory Webinars,
Live On-line Webinars, and Live In-person Courses.

Our courses are developed by experienced professionals!
Created with