Mind your step: learning to walk in complex environments.

In everyday contexts, children must respond to both self-related constraints (their own skills and abilities) and environmental constraints (external obstacles and goals). How do young children simultaneously accommodate these to support skilled and flexible behaviour? We used walking in a complex environment as a testbed for two hypotheses. Hypothesis 1: children will accommodate the self-related constraint of high foot placement variability via dynamic scaling. Hypothesis 2: children will plan ahead, even in complex environments. In our task, 3- to 5-year-olds and adults walked over obstacle sequences of varying complexity. We measured foot placement around the first obstacle in the sequence. Hypothesis 1 was partially supported. In simple, single obstacle environments, children engaged in dynamic scaling like adults. Those with more variable foot placement left greater margins of error between the feet and the obstacle. However, in complex, multiple obstacle settings, children employed large, un-tailored margins of error. This parallels other multisensory tasks in which children do not rely on the relative variability of sensory inputs. Hypothesis 2 was supported. Like adults, children planned ahead for environmental constraints. Children adjusted foot placement around the first obstacle depending on the upcoming obstacle sequence. In doing so, they demonstrate surprisingly sophisticated planning. We, therefore, show that in the motor domain, even very young children simultaneously control both self-related and environmental constraints. This allows flexible, safe and efficient behaviour.

The development of visually guided stepping.

Adults use vision during stepping and walking to fine-tune foot placement. However, the developmental profile of visually guided stepping is unclear. We asked (1) whether children use online vision to fine-tune precise steps and (2) whether precision stepping develops as part of broader visuomotor development, alongside other fundamental motor skills like reaching. With 6-(N = 11), 7-(N = 11), 8-(N = 11)-year-olds and adults (N = 15), we manipulated visual input during steps and reaches. Using motion capture, we measured step and reach error, and postural stability. We expected (1) both steps and reaches would be visually guided (2) with similar developmental profiles (3) foot placement biases that promote stability, and (4) correlations between postural stability and step error. Children used vision to fine-tune both steps and reaches. At all ages, foot placement was biased (albeit not in the predicted directions). Contrary to our predictions, step error was not correlated with postural stability. By 8 years, children's step and reach error were adult-like. Despite similar visual control mechanisms, stepping and reaching had different developmental profiles: step error reduced with age whilst reach error was lower and stable with age. We argue that the development of both visually guided and non-visually guided action is limb-specific.