Allocentric coding is a spatial orientation strategy whereby the location of a hidden goal is coded in relation to landmarks. Research showing that the ability to use this coding strategy emerges at around 9 months has led to the proposal that its development is linked to the onset of crawling; this proposal is referred to as the crawling hypothesis.
Spatial orientation abilities were measured by task performance in a circular enclosure whereby, during training trials, infants were presented with an event (the appearance of an adult playing "peek-a-boo") from various viewing positions. During subsequent test trials from a new viewing position, no event was presented, and infants' looking responses were recorded. The experimental condition environment contained visual features landmarking the event's location; the control condition, by contrast, contained no visual features.
A series of cross-sectional studies examined whether infants younger than 9 months were able to use allocentric coding after a simpler reorientation (Study 1) and with a facilitating training regime and more salient landmarking (Study 2). A further study (Study 3) replicated the methodology used by Tyler and McKenzie (1990), who reported high levels of performance in 6- and 8-month-olds. Overall the findings from these studies left open the possibility of the crawling hypothesis as a viable proposal.
A longitudinal study (Study 4) examined the link between the onset of crawling and allocentric coding-by monitoring the development of spatial and motor abilities from 5 to 11 months-and provided little evidence to support the crawling hypothesis.
Consequently, alternatives to the crawling hypothesis are considered (such as visual attention and exploration linked to brain maturation).