Contextual Constraints on Memory Retrieval at Six Months

Abstract

In 3 experiments, 6-month-old infants learned to move a mobile by kicking and were tested 1 to 21 days later for retention of the newly acquired memory as a function of the training and testing contexts. In Experiment 1, decreasing the relative distinctiveness of the training and testing context did not impair retrieval of the newly acquired memory. In Experiment 2, however, testing in a different context completely eliminated retention after delays of 1 and 3 days, when retention was otherwise perfect; after progressively longer delays, retention improved paradoxically. The familiarity or novelty of the test context was not a factor in the failure of infants to recognize the mobile in the altered context after 1 day. In Experiment 3, the effect of an altered context was assessed in a reactivation paradigm. After the training memory was forgotten, infants were presented with the original mobile as a reminder and were tested for retention of the training memory 1 day later. When either the reminding context or the testing context was different, they exhibited no retention. These findings reveal that memory retrieval at 6 months is highly specific to the setting in which the memory is acquired. We propose that infants learn what specific events are associated with what specific places prior to the age when they can locomote independently and acquire a spatiotemporal map of the relations between those places.

Visual‐preference and operant measures of infant memory

“…All infants were trained and tested as before except that blue-and-red striped cloth panels were draped over the sides of their cribs or playpens. This change was necessitated by the fact that a memory cannot be reactivated at 6 months of age unless infants are trained and reactivated in a distinctive context (Borovsky & Rovee-Collier, 1990). Three-month-olds received a 3-min reactivation treatment 13 days after the end of training and were tested 1 day later, on Day 14; 6-month-olds received a 2-min reactivation treatment 20 days after the end of training and were also tested 1 day later, on Day 21.…”

“…In addition, when we had previously tested 3-month-olds with either a novel mobile or the training (familiar) mobile 3 to 4 days after training, they had treated them equivalently, responding as robustly to the novel mobile as to the familiar training one (Rovee-Collier, Adler, & Borza, 1994;Rovee-Collier & Sullivan, 1980). In Experiment 3, therefore, we also gave a group of 3-montholds a paired-comparison test 4 days after the end of training and predicted that they also would kick robustly, evidencing operant retention, but would treat the novel and familiar mobiles equivalently and look at them nondifferentially (A corresponding group of 6-month-olds was not tested because infants of this age operantly discriminate a novel mobile from the training one for 2 weeks, which is the longest delay that they also exhibit operant retention; Borovsky & Rovee-Collier, 1990;Hill et al, 1988).…”

Developmental changes in the specificity of memory over the first year of life

“…For 3-month-old human infants, changing the context at the time of testing has no impact on retention after 1 day (Butler & RoveeCollier, 1989;Hayne, Rovee-Collier, & Borza, 1991) but impairs it after longer delays (Butler & RoveeCollier, 1989;Rovee-Collier, Griesler, & Earley, 1985). At 6 months, a context change impairs retention after both 1 day and 3 days but has no effect on retention after longer delays (Borovsky & Rovee-Collier, 1990;Shields & Rovee-Collier, 1992). At both ages, training in one room and testing in another one has the same effect as training in the presence of a distinctive cloth panel and testing in the presence of another one (Borovsky & Rovee-Collier, 1990;Butler & RoveeCollier, 1989;Hartshorn & Rovee-Collier, 1997;Hayne et al, 1991).…”

“…In addition, memory reactivation is impaired by a context change at both 3 months (Butler & RoveeCollier, 1989;Hayne et al, 1991) and 6 months (Borovsky & Rovee-Collier, 1990;Hartshorn & RoveeCollier, 1997), and at both ages, context-dependent retrieval can be overridden by training infants in multiple contexts (Amabile & Rovee-Collier, 1991;Rovee-Collier & DuFault, 1991). We have not tested 2-month-olds in an altered context, but we have attempted to enhance their long-term retention by training and testing them in a highly distinctive context.…”

“…The answer to this question is particularly important because recent studies of deferred imitation during the 2nd year of life have presented evidence that infants 12 months of age and older are generally insensitive to cue and context changes (Barnat, Klein, & Meltzoff, 1996;Hanna & Meltzoff, 1993;Hayne, MacDonald, & Barr, 1997), whereas evidence from mobile studies indicates that, after the same absolute delays, infants 6 months of age and younger are highly sensitive to such changes (Boller, Grabelle, & RoveeCollier, 1995;Borovsky & Rovee-Collier, 1990;Butler & Rovee-Collier, 1989;Gulya, 1996;Hill, Borovsky, & Rovee-Collier, 1988;Merriman, RoveeCollier, & Wilk, 1997). The present experiments, therefore, were designed to complete this picture for the last half of infants' 1st postnatal year and then to provide a systematic analysis of the effect of cue and context changes on infants' memory performance over the entire 1st year of life.…”