The study of the object processing abilities of infants is a growing area of research that extends our understanding of object cognition by complementing the numerous studies in adults and non-human primates. Not surprisingly, probing the mind of a 6-month old baby presents a series of problems that have led to some ingenious experimental procedures. One commonly used method involves measuring the amount of time that an infant will orientate to a given stimulus. This 'preferential looking' paradigm provides a measure of how interesting the infant finds a given stimulus and thereby offers a window into their previous experience with that object and their developing perceptual and cognitive abilities. Although this type of behavioural paradigm has been used with great success, the ability to make more direct measurements of brain processing in infants would provide a significant step forward for this area of research and would strengthen the link with the adult and non-human primate literatures. Csibra and colleagues provide an example of how non-invasive imaging can be combined with more traditional approaches to bring the development of object processing systems in the infant brain into the neuroscience arena.

Their experiment focused on one of the key questions in visual perception — if the visual features of an object such as form and colour are, to a large degree, processed separately, how is this information 'bound' together in the brain to allow the perception of a unified object? One theory suggests a prominent role for gamma-band (40 Hz) oscillatory activity in perceptual binding.

This form of binding can be studied through the use of illusory percepts such as the Kanizsa square, in which an illusory square can be induced between four appropriately orientated 'pacman' elements as shown in the figure. In adults the perception of illusory objects induces a burst of 40 Hz oscillations shortly after stimulus onset. Csibra et al. used electroencephalography to measure the oscillatory brain activity of 6 and 8-month old infants while they viewed either a Kanizsa square or a control stimulus consisting of misaligned elements that did not induce the illusory percept. There was an enhancement of gamma-band activity in response to the Kanizsa square in 8-month old infants but not in 6-month old infants. This activity was not observed in response to the control stimulus.

This work is consistent with, and extends, the behavioural studies that suggest that perception of the Kanizsa square begins at around 7 months of age. But the wider importance of this work may reside in the demonstration that direct measurements of neural processing can be undertaken in very young infants during the performance of behavioural tasks. By encouraging the dialogue between cognitive development and cognitive neuroscience, this study may therefore bind these two areas together and so open the doors of infant perception.