Esther Thelen

Introduction

Esther Thelen (May 20, 1941 – December 29, 2004) did more than anyone else to turn the study of how babies move into a testing ground for theory. She is best known for taking an unremarkable observation — that newborn stepping disappears for a few months before walking appears — and arguing that the cause was not in the brain at all, but in the changing ratio of fat to muscle in the infant's legs. That single move, made in the early 1980s, opened the door to a much larger claim: that development is not the unfolding of a plan written in the genes or the cortex, but the self-organization of many parts that cooperate, compete, and reorganize in real time.

Her colleagues described her view of development as closer to jazz improvisation than to a biological program. In her own words, dynamic systems approaches "seek to understand the overall behavior of a system not by dissecting it into parts, but by asking how and under what circumstances the parts cooperate to produce a whole pattern." She died of cancer at sixty-three, at the height of her career, while serving as president of the Society for Research in Child Development.

Early Life and Education

Thelen was born in Brooklyn, New York, to Alden and Sarah Stillman. She attended Antioch College and then took a bachelor's degree in zoology at the University of Wisconsin in 1964. In 1962 she married the historian David Thelen, and the next decade followed a path that would now be called unconventional and was then simply common: she was a faculty wife and the mother of two children, and did not enter graduate school full-time until the following decade.

She returned to the University of Missouri, taking a master's degree in zoology in 1973 and then a doctorate in 1977. The field was the biological sciences, not psychology — a fact that explains almost everything about how she later worked. Her training was in ethology, the study of species-typical behavior patterns in animals, with its habits of close observation, its suspicion of laboratory artifacts, and its assumption that behavior is something a whole moving organism does in a real environment.

Her dissertation examined rhythmical patterned movements in infants: the spontaneous, metronomic kicking that newborns produce. To most psychologists at the time this was noise. To someone trained to look at fixed action patterns it was a phenomenon, and it became the seed of everything that followed.

Academic and Professional Career

Thelen took her first faculty position in the psychology department at the University of Missouri in 1977. During those years she was treated for tongue cancer, and fought it into remission while relearning how to speak and how to eat. It is hard to read her later writing without noticing that she understood, at a level most theoreticians do not reach, that a body is a piece of machinery that has to be coaxed back into working.

She moved to Indiana University Bloomington in 1985 as a full professor of psychology, and founded and directed its Infant Motor Development Laboratory. Over nineteen years the lab became what colleagues called an intellectual hub for dynamic systems approaches to development, and a training ground for the motion-tracking technology that made microgenetic work possible. Her methods were deliberately strange and deliberately concrete: infants marched on motorized treadmills, kicked with elastic binding their legs together, and reached while wearing glitter sleeves so their limb movements could be recorded.

She held visiting positions at Ben-Gurion University in Israel and at the École des Hautes Études en Sciences Sociales in Paris. She was president of the International Society on Infant Studies from 1996 to 1998 and president of the Society for Research in Child Development from 2003 to 2005, and was a fellow of the American Association for the Advancement of Science and the American Psychological Association. Her research produced three books and more than a hundred and twenty scientific articles and chapters, funded continuously by the National Institute of Mental Health and repeatedly by the National Science Foundation. She died at Bloomington Hospital on December 29, 2004.

Major Contributions to Psychology

A place where nothing is in charge

The organizing claim of Thelen's work appears in an explicit form in her collaboration with Linda Smith: no single component of the system is privileged. Not the genes, not the cortex, not the environment — no single element is uniquely responsible for the appearance of a new behavior. New skills emerge when enough components happen to line up, and they can be destabilized again just as easily. The implication is that development should be studied at multiple time scales at once: the second-to-second unfolding of a reach, the week-to-week changes in a body, and the year-to-year trajectory, all of which turn out to be the same process viewed at different resolutions.

The disappearing reflex

The established explanation for the disappearance of newborn stepping was cortical maturation: subcortical circuits produce it, the cortex then takes over and suppresses it, and voluntary walking returns later. Thelen and her colleagues showed that the same alternating, coordinated pattern could be elicited in supine kicking, and proposed that the apparent disappearance was mechanical. Infants gain fat faster than muscle over the first months; lifting a leg against gravity becomes expensive; the behavior drops out. Adding small weights to an infant's ankles reduced or eliminated stepping. Immersing the legs in water, or supporting the body over a treadmill, brought it back. The apparent loss and mysterious reappearance became a problem in body mechanics rather than a milestone in neural development.

The A-not-B error

The second line of work took on Jean Piaget's most famous infant phenomenon. In the classic task, a toy is hidden at location A; after several successful retrievals it is hidden at B, in full view; the infant nonetheless reaches back to A. Piaget read the error as a limit on the infant's concept of the object. Working with Linda Smith and the theorist Gregor Schöner, Thelen built a dynamic field account in which the error is not a failure of knowledge but the output of a system with a memory: a reach to A leaves a residual activation in the movement field that decays only slowly, and given a delay of ten or twelve seconds that residue can carry the arm back to the wrong location even though the child saw where the toy went. The account predicts sensitivity to exactly the variables that turn out to matter — delay length, how many times the child reached to A beforehand, how distinctive the two locations are, whether reaching requires standing up or changing posture.

The long reach into practice

Her influence on pediatric physical and occupational therapy has been direct and practical. If new motor skills are assembled by each child from that child's own body and circumstances, then the unit of treatment is the individual, not the age band. Therapists began designing exercises around a particular child's strength, posture, and motivation rather than applying the standard sequence for a given month of life.

Major Publications

  • Thelen, E., & Fisher, D. M. (1982). Newborn stepping: An explanation for a "disappearing" reflex. Developmental Psychology, 18(3), 447–453.
  • Thelen, E., Fisher, D. M., & Ridley-Johnson, R. (1984). The relationship between physical growth and a newborn reflex. Infant Behavior and Development, 7(4), 479–493.
  • Thelen, E. (1995). Motor development: A new synthesis. American Psychologist, 50(2), 79–95.
  • Thelen, E., & Smith, L. B. (1994). A Dynamic Systems Approach to the Development of Cognition and Action. MIT Press.
  • Thelen, E., Schöner, G., Scheier, C., & Smith, L. B. (2001). The dynamics of embodiment: A field theory of infant perseverative reaching. Behavioral and Brain Sciences, 24(1), 1–34.

Impact and Legacy

Thelen's American Psychologist obituary credits her with having introduced developmental psychologists to dynamic systems theory and reinvigorated a field that had gone quiet. Two decades later her own former student John Spencer and colleagues reviewed the whole approach and judged it, en masse, extremely successful — in places revolutionary. Neither of those writers is hostile, and both identify the same unresolved problem, which is the right place to start with the criticisms.

The mechanism is stated in metaphor more often than in equations. The recurring objection is not that dynamic systems theory is wrong but that its central terms are loosely specified. What counts as a stability versus a loose one, which parameter is the control parameter, and where the threshold lies are questions that the verbal version of the theory does not answer in advance. In emigrating from mathematics and physics into developmental science the framework kept the vocabulary — attractors, phase transitions, soft assembly — and left behind most of the formal constraints that made those terms precise. The accounts it offers are often excellent descriptions of a pattern already observed, and correspondingly weak predictions of a pattern not yet measured.

Explaining away the regularities is harder than it looks. The theory emphasizes individual pathways: different bodies, different histories, different solutions. This is its strength. It is also a liability, because development across children is strikingly ordered, and a framework that stresses uniqueness must also explain why nearly every child reaches before walking and walks before running.

Its own practitioners disagree about what it is. Two decades after A Dynamic Systems Approach, even sympathetic reviewers noted that there is no one official version. Papers use different terms for the same idea, disagree about whether self-organization is a mechanism or a description, and split along metatheoretical lines over whether higher-level patterns have any explanatory standing at all. The contextualist reading descended from Thelen and Smith treats global order as something to be explained from local interactions; others treat it as genuinely causal. That is a philosophical disagreement wearing the clothes of a technical one. It was made explicit in a special issue of Human Development marking twenty-five years since A Dynamic Systems Approach (63(3-4):264-276, 2019). Witherington and Boom argue there that the field is moving back toward the organizational focus Thelen and Smith rejected: contemporary dynamical treatments are trying to reintegrate purposiveness, end-directedness, normativity and subjectivity as genuinely explanatory, where Thelen and Smith had treated levels of organization above the basic process dynamics as largely epiphenomenal outgrowths of a single dynamics common to any open system. They also name two problems the framework still has to solve - what it means for an ability or power, understood as a potential, to explain an action, and how real-time change and developmental-time change are supposed to interrelate.

The stepping account has competitors. The proposal that shifting fat-to-muscle ratio is what removes newborn stepping was a genuine advance over maturation-as-timetable, and the weights-in-water experiments are unusually elegant. But later work suggests the story is more crowded than "fat": strength, postural control, and the treadmill training itself, which is practice, and therefore learning. The identical controversy runs through the A-not-B literature, where the dynamic field model competes with graded-representation accounts, connectionist models of developing memory, and memory-based accounts that treat the error as forgetting. All of them predict delay effects. Distinguishing them requires quantitative fits, and a formal model with many parameters can be made to fit almost any pattern.

Set against that is the fact that the most useful criticisms now come from inside. Spencer, who was trained in her laboratory, frames the challenge the way she would have: not to abandon dynamic systems theory but to specify formally how multiple levels interact across multiple time scales. Her reply to the charge of vagueness would have been that the alternative was worse — a century of naming stages instead of explaining how change happens. The framework's largest growth has been outside the field she built it in. Complex dynamic systems theory in the form Thelen and Smith gave it is now a standard frame in second-language development research, where variability is treated as evidence of a system reorganising rather than as measurement error, and it is one of three families of account - alongside representational and radical-embodied ones - invoked in current work on how executive function develops. A 2025 review of embodied approaches to early executive function (Frontiers in Cognition) places Thelen and Smith's dynamical framework and its descendants as the systems alternative to representation-based models, while concluding that the embodied position remains "in its early stages" and that representational accounts still dominate the field.

Learn More

References

  • Thelen, E., & Smith, L. B. (1994). A Dynamic Systems Approach to the Development of Cognition and Action. MIT Press.

Last updated: 2026-10-20

Menu
Loading