Infant memory shows the power of place.

Six-month-olds rely on surprisingly specific aspects of their incidental surroundings--such as the color or design of a crib liner -- to retrieve memories of a simple learned task, according to ongoing research by psychologists at Rutgers University in New Brunswick, N.J.

"Place information enjoys a privileged [mental] status much earlier in development than previously thought and seems to be the first level of retrieval for memories among infants and adults," asserts project director Carolyn Rovee-Collier.

The new findings, described in the March DEVELOPMENTAL PSYCHOLOGY, challenge current neuropsychological theories that consider basic language skills a prerequisite for memory development and assume that infants younger than around 9 months cannot store information about their surroundings in a systematic way.

Rovee-Collier's group conducted a series of experiements with a total of 85 infants. Each baby reclined in a seat placed in a playpen whose sides were draped by a yellow liner with green squares. A mobile featuring seven wooden figures and four jingling bells hung over the playpen. On two successive days, experiments tied a satin ribbon attached to the mobile around an infant's ankle for 6 minutes and an experimenter recorded the number of leg kicks as the infants learned to move the mobile. On the following day, an experimenter again charted kick rates for infants either in the original playpen or in one draped by a liner with a slightly different design.

Babies continued to kick at their previous rates if the liner displayed triangles instead of squares, but kicking dropped off drastically if circles or stripes adorned the liner. If the color of squares changed from green to red, or if the liner displayed no figures, infants continued to kick in response to the mobile. But if the color of the yellow background changed, or if colors on the liner were reversed (green background and yellow squares), infants showed no recognition of the mobile.

Kick rates also plummeted if experimeters removed the liner on the test day, leaving the familier context of the infant's playpen and bedroom. This suggests that altered test contexts produce kick-rate declines not because of their novelty but because infants cannot locate the modified context in their original training memory, Rovee-Collier argues.

Previous experiments by her team showed that 6-month-olds remember how to move the mobile up to two weeks after training sessions, but only if the design of the crib liner remains the same. Ever 3-month-olds learn that kicking sets the mobile in motion, and they retain this knowledge for three to five days, Rovee-Collier says. But a 3-month-old who trains in the bedroom and gets tested in the kitchen, or who goes from crib training to testing in a lower, portable crib, stares blankly at the previously encountered mobile.

"Young infants learn what happens in what place long before they are able to move from one place to another or learn the spatial relations between those places," Rovee-Collier contends. Although she declines to label an infant's reliance on specific features of a playpen liner as either conscious or unconscious, she adds that "the behavior of the babies we study is very deliberate."

She theorizes that context information serves as an "attention gate." When context during learning matches context at recall, recognition of basic perceptual cues -- such as colors and some forms -- permits attention to focus on memories for a learned task, such as moving a mobile with leg kicks. This theory holds that sensory receptors and the brain first break incoming information into "elementary perceptual units," which then get put back together to form a coherent perception.

Context's critical role in memory, which researchers have also observed among adults, suggests to Rovee-Collier that the brain harbors a single memory system rather than multiple types, as some researchers have proposed (SN: 11/17/90, p.312).