Karl Newell (1986) *Constraints on the Development of Co-ordination”* in M. G. Wade, & H. T. A. Whiting’s book; Motor Development in Children: Aspects of Coordination and Control.

Aims:

This impactful publication was a chapter that was originally targeted at a motor development readership, but it quickly became clear that it was relevant for understanding coordination in different contexts including motor learning for sport performance. It had clear aims:

1. To show that by considering the role of constraints on co-ordination, researchers’ could test the utility of contrasting theoretical models that were prevalent at that time.

2. To propose that the constraints approach could give researchers interested in movement coordination an opportunity to bring together motor development and motor learning theories. This integration of theoretical ideas on motor development and learning was an important challenge that Karl Newell and his group of PhD students took on in the late 1980s and early 1990s. This was because there were separate theories to explain motor development in children and how they learned to move.

Newell and colleagues were convinced of the importance of providing one common theoretical model to explain acquisition in all categories of co-ordination, captured in this book chapter:

Newell, K. M., & van Emmerik, R. E. A. (1990). Are Gesell's developmental principles general principles for the acquisition of coordination. In J. E. Clark & J. H. Humphrey (Eds.), Advances in motor development research (Vol. 3), (pp. 143-164). New York: AMS Press.

Background to the paper

Newell’s decision to publish his treatise on constraints in a book on motor development, rather than motor learning or motor control, is significant since at that time, the only research that had truly focused on co-ordination of behaviour had been within the field of motor development. That was because of the over-emphasis on motor control and learning in the dominant theories at that time, despite the fact the Nikolai Bernstein (1967) had focused attention on understanding coordination of the many different parts of the body decades before.

Also, for many practitioners working with children and youth it is confusing as to what actions and skill should be classed as belonging to motor development or motor learning. What are the differences proposed between motor development and motor learning?

Here are some standard definitions from previous texts:

Motor Development: The sequential, continuous age-related process whereby movement behaviour changes (Haywood & Getchell, 2005). Continuous change in motor behaviour throughout the life cycle, brought about by interaction among the requirements of the movement task, the biology of the individual and the condition of the environment (Gallahue & Ozman, 2006).

Motor Learning: Refers to the relatively permanent gains in motor skill capability associated with practice or experience (Schmidt & Lee, 1999).

Motor Control: The study of the neural, physical, and behavioral aspects of movement (Schmidt & Lee, 1999).

Like Newell, for skill acquisition specialists interested in understanding the processes underpinning co-ordination of human movement, studying those at the beginning of their movement journey, that is infants, would seem to be an obvious place to start. However, up until recently, the basic movement patterns, categorised as phylogenetic skills, that we tend to associate with infant motor behaviors, such as, reaching, grasping, rolling, crawling, sitting, standing and walking have been conceptualised as different to more specialist skills needed for sport performance (i.e., ontogenetic skills) such as kicking a ball, hitting or throwing a ball or implement, gymnastics or athletics skills, or individual pursuit skills such as canoeing or climbing. An interesting question is: Why the distinction?

The distinction between co-ordination and skill:

Nikolai Bernstein (1967) emphasised the importance understanding how we learn to continually re-organised body parts (he termed ‘degrees of freedom’: muscles, joints, limbs, and more) to achieve our intended movement goals, such as maintain balance on a mountain surface using all four limbs or intercept a moving ball. Later Peter Kugler defined Co-ordination as: the function that constrains the potentially free variables into a behavioural unit.

Skill is the optional parameterization of this [coordination] function.

(Kugler et al. 1980)

Essentially, it was long believed that movement skills of infants were down to maturation and occur naturally, and simply down to (motor) development. In contrast, the other category included movements, classed as ontogenetic actions, had to be specifically learned-hence categorised under motor learning. What evidence is there to suggest that the two classes of activities emerge through different processes? One potential reason is that given that ‘everyone’ with normal development is capable of all of the basic patterns of co-ordination, but not everyone can perform specific ontogenetic actions, then these basic patterns must be learned by repetitive programming resulting in a set of internalised instructions. In contrast, phylogenetic movements were thought to appear through prescriptions embedded in the genetic make-up of each individual, appearing in a standard development sequence beginning within the uterus and onwards into infancy and childhood. Consequently, these phylogenetic movements are likely to appear in sequence, irrespective of the experiences of the individual, with delays or lack of appearance being designated as a developmental disorder. Evidence from studies (prior to the 1980s) seemed to support this view, even though most of our understanding came from work undertaken over 40-50 years earlier by the likes of Gesell (1929), McGraw (1943) and Shirley (1931). These studies were meticulous in their observations and analyses but were mainly descriptive in nature revealing that fundamental movement and posture patterns occur via a progressive ordered and regular sequence of stages of motor development. For example, Shirley (1931) observed 25 infants and identified five sequential stages in the development of “upright” locomotion. Similarly, McGraw (1943) categorised seven stages of progression from birth to normal gait. Arnold Gesell reported observing 23 different stages of infant behaviours playing with ‘rattles’. Thus, the finding that basic movement patterns follow an invariant and universal sequence appears to show strong support for the maturational view of motor development. Over time this assumption has come under question due to research that demonstrated the nonlinearity and variability of maturation and typical motor development, exemplified by work of Linda Smith and Esther Thelen in the 1990s.

See: Smith, L. B. & Thelen, E. (Eds.) (1993). A Dynamic Systems Approach to Development - Applications. Cambridge, MA.: MIT Press.

Thelen, E. & Smith, L.B. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA.: MIT Press.

In his work spread over 30 years, Gesell proposed that maturational perspective which is generally taken to refer to the specific patterns of co-ordination exhibited in early childhood is characterised by (i) the appearance of new behaviours without practice (ii) consistency in these new patterns of behaviour across subjects within the same species and (iii) an orderly and invariant sequence in the development of these behaviours.

One point worth noting before we move on is that biological factors were acknowledged by Gesell as being important in the emergence of ontogenetic skills; however, in reality they have been down played in maturational theory.

So, if maturational development processes could explain infant co-ordination, how could actions that need to be specifically acquired be explained? The answer is that in 1986 was: not very well. This was mainly due to motor learning research being limited in its study of co-ordination by its insistence on utilizing tasks in which participants used single degrees of freedom (one joint, digit or limb), or by requirements that participants needed to produce a co-ordination pattern on the first attempt. In effect, this approach meant that motor learning scientists were hampered by the experimental paradigms of dominant motor control theories and were not studying true co-ordination problems. Consequently, and perhaps understandably, for those interested in understanding how co-ordination emerged through learning, the ideas from motor development were very attractive. For example, Roberton (1982) mapped the invariant steps in throwing actions of children.

The emergence of cognitive science models of motor control and learning, such as Schmidt’s (1975) Schema theory, might have been thought to have superseded the maturational perspective. However, at the most fundamental level, the schema approach was still largely based on internal focus on symbolic knowledge structures in the form of representations in the brain prescribing actions.

The advent of cognitive models to explain motor skill development along with emerging evidence that in some instances individuals regressed to earlier less ‘progressive’ stages and in others there were omissions between stages, resulted in maturational perspective losing efficacy due to its failure to explain the processes underlying co-ordination. However, as Newell highlights in the paper, maturation theory is implicit in current (circa 1986) accounts of the development of co-ordination.

A final point before me move on, is that nowadays, due to the dominance of cognitive theories, emphasising top-down movement control models, we often see the terms motor development, motor learning and motor control used inter-changeably. This is a misunderstanding that spills over into coaching and teaching where performance and learning are sometimes confused.

Co-ordinative structure theory

Scientific theory evolves through dissatisfaction with current explanations and falsification of testable hypotheses. This was where movement scientists were in 1980. Influenced by the ideas of Bernstein (1967) and dynamical systems in science, the time was ripe for attempting new explanations of coordination, and this gap was filled by advocates of co-ordinative structure theory: Kugler, Turvey and Kelso (1980, 1982). These authors proposed a new model that was framed in dynamics where they posed the question regarding how n information in the environment could be contextualized so that it is continuously co-ordinated with changes in an individual’s as skeleton-muscular dynamics, across different timescales of performance, learning and development, for example, those brought about by changes in skeleto-muscular dimensions (1982, p.5). Put simply, the question of interest for Kugler and colleagues was: how do individuals successfully co-ordinate their movements during periods of growth, when the scaling of a person’s dimensions change?

Brief Aside: The fusion of key ideas of dynamical systems theory and ecological psychology was first mooted by Peter Kugler, Michael Turvey and Scott Kelso as a basis for understanding movement control and co-ordination positing that information and dynamics are complementary in complex systems. The term ecological dynamics emerged later (Araújo et al., 2006) and around the same time, William Warren proposed how to integrate perception-action coupling and dynamics of movement coordination.

See:

Araújo, D., Davids, K. & Hristovski, R. (2006). The ecological dynamics of decision making in sport. Psychology of Sport and Exercise 7, 653-676.

Warren, W. (2006). The dynamics of perception and action. Psychological Review 113 (2), 358–389.

Kugler and colleagues framed their answer in the context of Bernstein’s (1967) degrees of freedom problem. That is, how does one re-organise (co-ordinate and control) the 792 muscles and over 100 mobile joints to ‘construct’ movement patterns which function to achieve intended task goals? In this context we will class a functional movement as one that solves a movement problem. They proposed that the solution to any co-ordination problem is through a systematic linking of muscles in such a way that the ‘set’ of individual muscles used in a specific context (e.g. throwing a ball) is reduced to a much smaller set of muscle collectives, that are constrained to act as a single functional unit. This obviously reduces the problem of coordinating so many muscles, limbs, segments and joints, which traditional cognitive models such motor programming and schema theory wrestled with. They termed these collectives co-ordinative structures that are temporally organized as a single, coherent unit intended to achieve a task goal such as reaching, stepping, intercepting, kicking and so on.

In the co-ordinative structures approach, the emergence of an optimal pattern of co-ordination is predicated on the capability of a biological system to self-organise components in response to the constraints imposed upon it (adapt and adjust as conditions change). To adapt and adjust precisely an organism needs information from within and from its surrounds. This information for self-organising system degrees of freedom comes from interacting constraints as Newell argued. The role of interacting constraints was, of course, the focus of Newell’s impactful treatise and we will specifically discuss how this model works in detail in the next blog.

One final point that is of value for contrasting the traditional approach and the ‘new’ co-ordinative structures approach is the notion that co-ordination patterns are only assembled temporarily. The value of temporally, or ‘softly’, assembling a co-ordinative structure, rather than hard-wiring it in into the brain as proposed in a motor programming model to enable them to be ‘run-off’ or reproduced later, makes sense as humans are complex highly adaptive systems which are dynamic in nature, with changes being the norm. This means that a co-ordinative structure has a shelf life and is only of value if conditions in the individual-environment system remains stable. However, when some aspect of the system changes (either immediately or more slowly) beyond a critical value (e.g., a change in leg strength for an infant, or the medium through which an child is invited to walk (hard (rocks), uneven (pebbles) or soft (sand) surfaces) the co-ordinative structure will become less functional and a new one needs to emerge to help the mover adapt to the new context. This is an important consideration for practitioners as it demonstrates that there is no need for highly repetitive practice attempts to attempt to hard wire specific co-ordination patterns. Soft-assembly requirements in all movement contexts require an emphasis on seeking stability of course, but mostly adaptations.

Summary

When Newell published his paper, the traditional approach to motor development had not moved on too much from the decades-old maturational and normative-descriptive approaches and theories. Newell’s ideas emerged in the context of contrasting thinking that was prevalent at that time that centred on the relative contributions of nature and nurture in terms of movement skills. Essentially this approach suggested that some skills were simply due to maturation, whilst others had to be learned. Newell proposed that constraints could provide researchers with a vehicle to test the utility of the contrasting theoretical models as well as providing one common theoretical model to explain all types of co-ordination: for motor learning and development over the life course.