INFANCY, MEMORY IN

According to classic developmental theory, infants operate in the present, without thoughts of the past or anticipations of the future. Adults cannot remember events from infancy, a fact sometimes cited to corroborate the notion that memories are not formed during the preverbal period. However, experimental studies conducted in the late twentieth century demonstrate that young infants have more robust memories than heretofore believed. Indeed, modern theorists focus on the different types of memory infants might have. Infants seem to remember particular things under certain conditions and not others; they may also have privileged memory for biologically relevant signals such as faces and speech sounds.

Techniques for Investigating Infant Memory

Three experimental procedures have been developed to probe infant preverbal memory: visual preference tests; conditioning procedures; and deferred imitation. Each approach measures a different type of nonverbal memory. A fourth technique, object permanence, can also be used and it is only briefly mentioned in this entry because it is reviewed elsewhere in the encyclopedia.

Infant Visual Recognition Memory

The procedures used to evaluate infant visual recognition rely on infants' curiosity for exploring novel visual patterns (Bremner, 1994). Infants are shown a visual pattern for a certain length of time. A delay is imposed, and then they are presented with the old visual pattern and a new one. If infants devote more looking time to the new pattern than to the old one, this is taken as evidence that they have memory of the previously exposed target. Two specific techniques use this underlying principle: the habituationdishabituation technique and the paired-comparison technique. For habituation-dishabituation, infants are repeatedly exposed to the initial target until they become bored with it (habituated). The new pattern is then introduced, and if looking time increases significantly (dishabituation), this shows that the infants recognize the pattern as being different from the one in memory. For the paired-comparison technique, infants are initially shown two identical patterns side by side for a certain familiarization period. It is not required that the infants habituate, only that they visually examine the display (usually thirty seconds to two minutes). Then a delay is imposed, and two patterns (the old one and a new one) are again presented. The index of memory is their preference for the novel pattern.

Advances in technology allow researchers new ways of looking at the brain basis of memory in infants. Event-related potentials (ERP) can be used to examine the neural correlates of recognition memory and attention in infants (Nelson, 1995). Typically, one of two approaches is used. In one case, infants are habituated or familiarized with a stimulus. Brain activity is then measured by recording electroencephalogram (EEG) signals in response to a series of trials in which either the habituated stimulus or a novel stimulus is presented. Brain measures record whether infants discriminate between the new and the old stimulus. In a second type of study, infants are shown stimuli that are familiar: for example, their mother's face or a familiar toy, and brain activity to these remembered stimuli is measured and compared to novel stimuli (de Haan and Nelson, 1999). The first approach provides brain measures of memory for well-controlled, briefly exposed stimuli; the second approach measures brain reactions to naturally occurring stimuli that have been frequently seen.

Age-Related Findings.

Initially, studies of visual recognition memory reported no memory in infants younger than about ten to twelve weeks of age. However, researchers soon discovered that if the length of time infants studied the to-be-remembered stimuli was increased (up to five minutes) and the patterns were made very different from one another, even newborn infants could retain information in the visual recognition paradigm, at least for delays of a few seconds. Research then shifted to the effect of study time on memory; and the length of retention interval that can be tolerated by infants.

Study Time (Length of the "Encoding" Phase).

Infants require shorter study times to demonstrate the novelty preference (the measure of memory) when the choice stimuli are vastly different than when the stimuli are similar. This idea was illustrated in a study by Joseph Fagan (1990) using the paired-comparison technique. He tested five-month-old infants using pairs of patterns that were graded in the degree to which they were discriminable from one another. The results showed that when the easiest pair was used, infants needed only about five seconds of study time to demonstrate the novelty preference; when the pair of medium discriminability was used, they needed about twenty seconds; and when the least discriminable pair was used, they required about thirty seconds. Thus, infants, like students studying for an exam, seem to need relatively more time to study material if they are asked to remember subtle distinctions.

Length of Retention Interval.

Fagan showed visual patterns to five-month-olds and then imposed delays of three hours and one, two, seven, and fourteen days. The results revealed that infants could recognize which target they had previously seen even after the fourteen-day delay. What makes babies forget? Results from a variety of studies show that young infants will forget if they are shown highly related material during the retention interval. For example, if infants study photographs of faces and then are shown other face photographs during the retention interval, their subsequent memory performance will be poorer. The two factors that lead to maximum interference are stimuli that closely resemble the to-be-remembered material; and interfering presented soon after the initial exposure. This is reminiscent of adult memory, inasmuch as interference with remembering a telephone number is maximized by hearing other numbers soon after the initial information is delivered.

Another factor that influences the length of infant retention is the temporal spacing of the initial studying time. In one study, two groups of infants were given the same length of time to study a face photograph (twenty seconds). However, for one group this study time was massed, meaning it consisted of four five-second intervals with only a few seconds separating each interval. For the other group this experience was distributed, meaning there were much longer pauses between the four five-second intervals. Both groups demonstrated immediate recognition memory; however, only the group that received the distributed exposure remembered over long delays. This effect of distributed study is also a well-documented aspect of adult memory.

Conditioning Techniques

The second approach to evaluating infant memory was developed by Carolyn Rovee-Collier (1997). It involves training infants to produce a foot-kick response to a mobile hanging in their cribs. The mobile is often attached by a ribbon to one of the infant's ankles, so that the frequency and intensity of the movement mimics that of the infant (known as conjugate reinforcement). Infants as young as two to three months rapidly learn the contingency, doubling or tripling their baseline rates during the nine-minute training period. Once the infant has learned the response, a delay period can be inserted between the initial training period and the reintroduction of the mobile into the crib. Memory for the learned response is indexed by an increase in kicking over baseline rates, even after this delay interval.

Age-Related Findings.

With this technique, Rovee-Collier and colleagues discovered a steady increase in the duration over which infants retain responses from two to eighteen months of age. Whereas the youngest infants tested in this paradigm remembered the response contingency for one day to one week, eighteen-month-olds remembered for thirteen weeks or longer.

Retrieving Infant Memories That Were Once Forgotten.

One remarkable discovery made using this conditioning technique is that infants can be reminded about a past event that they have forgotten (Rovee-Collier, Hayne, and Colombo, 2001). This reminder stirs (reactivates) a previously inaccessible memory. Rovee-Collier's classic demonstration involved three-month-old infants who had forgotten the learned response after two weeks. These infants were then exposed to a brief reminder (the mobile, which was being moved surreptitiously by the experimenter). Then the infant was given another twenty-four-hour delay; finally, the stationary mobile was reintroduced to assess memory. Infants administered the reminder had their memories "reactivated" and kicked vigorously when the mobile was reintroduced. Control infants who were not given the reminder did not show any memory under the same circumstance.

Memory Specificity and the Importance of Context.

In the first six months of life, infants are extremely sensitive to the context in which a behavior is acquired and show better memory if the test occurs in the same context as the learning episode. In one study, six-month-olds were given foot-kick training in a specially decorated crib (the context). As long as these infants were tested in the same context as the original training, they remembered to kick, even after a fourteen-day delay. However, if the crib decoration changed, the infants could not access their memories, even after a oneday delay (Rovee-Collier, 1997).

Deferred Imitation

The third procedure for testing infant memory is deferred imitation, that is, imitation after a delay (Barr and Hayne, 2000; Meltzoff and Moore, 1998). This technique capitalizes on the fact that preverbal infants enjoy imitating the actions of adults. To test memory, the infant is shown the to-be-imitated event, and then a delay is inserted before the infant is allowed to demonstrate the response. Memory is indexed by accurate reproduction of the target behavior after the delay. Control groups are tested to ensure that the production of the target behavior would not have occurred spontaneously in infants not exposed to the initial modeling.

Age-Related Findings.

Classic developmental theories, such as those of the Swiss psychologist Jean Piaget, had supposed that imitation from memory was a cognitive achievement that first emerged at about eighteen months of age. Empirical research revised this classic view by showing that infants can perform deferred imitation as early as nine months of age (Meltzoff, 1988).

Imitative Learning and Length of Retention Interval.

Scientists have been interested in infants learning novel material through observation. Andrew

[Image not available for copyright reasons]

Meltzoff (1990) showed infants a novel act that had not occurred in the baseline behavior of infants, and tested their memory for this act after a one-week delay. The act consisted of bending forward from the waist and tapping the top surface of a box with the top of one's forehead. Even after a one-week delay, 67 percent of fourteen-month-olds bent forward from the waist and touched their head to the panel (see Figure 1). Leslie Carver and Patricia Bauer (2001) found that ten-month-olds were able to retain at least some information about to-be-imitated events for delays of up to six months. Jean Mandler (1990) assessed memory for temporal order by presenting a sequence of behaviors that could be performed in one order or another. Both the sixteen-and twenty-month-olds showed immediate memory for the temporal order of these arbitrary sequences; the groups differed, however, on their long-term memory performance. The twenty-month-olds still showed memory for temporal order after a two-week delay, whereas the sixteen-month-olds were at chance levels. Long-term memory for arbitrarily sequenced events may develop toward the second half of the second year, and perhaps is aided by the emergence of language during this time.

Combined Brain and Imitation Measures.

Carver, Bauer, and Nelson (2000) tested nine-month-olds for deferred imitation of ordered sequences of actions. About half of the infants recalled the events after a delay of one month. Event-related potentials (ERP) were measured during the delay interval to pictures of toys. Brain electrical responses mapped onto individual differences in behavior: Infants who successfully performed deferred imitation showed recognition memory as measured by ERP patterns during the delay interval; infants who did not imitate showed no recognition memory as measured by ERP. These results, and others, suggested that there are changes in the organization of brain systems involved in memory near the end of the first year of life, and that these changes can be observed in measures of individual differences in both brain and behavior.

Real-World Implications.

Infants remember what they see on television. In one study infants were shown how to manipulate a new toy by an experimenter who appeared on television. The infants were not allowed to handle the real toy, but the next day the real toy was presented on the table. The results showed that infants accurately imitated the television-presented actions they had seen one day earlier. Another study examined whether fourteen-month-olds could remember actions performed by other infants. A "tutor infant" was taught how to perform a series of particular acts. This tutor infant was then brought to day-care centers where he demonstrated the acts to "naive infants." After a forty-eight-hour delay, the naive infants were visited at their homes by a researcher who laid out the toys on the floor. The results showed that the infants imitated what they had seen their peer do two days earlier. A general implication of this work is that imitation and memory are robust enough to play a significant role in the social and personality development of the preverbal infant (Meltzoff and Moore, 1998).

Conclusion and a Look to the Future

Multiple Memory Systems in Infants

The three techniques used to explore infant memory complement each other but do not address precisely the same aspects of infant memory. The distinctions are important for theory.

Recognition Versus Recall.

In deferred imitation, infants go beyond the regulation of attention; they do more than react to the "newness" of a pattern. They must produce an absent act without seeing it and without having previously imitated it. Deferred imitation taps something more than visual recognition memory and provides a measure of recall memory prior to the acquisition of language.

Imitative Learning Versus Conditioning.

Like deferred imitation, the conditioned foot-kick technique goes beyond visual recognition memory because the infants do more than recognize the familiar mobile; they also retrieve from memory what to do (kick). However, the conditioning procedure and the deferred imitation task differ in the type of information retained. Deferred imitation is based not on an incrementally learned procedure (as in the case of conditioned foot kicks) but on the performance of an act that was simply perceived during a brief previous episode. The deferred imitation test does not involve any motor practice during acquisition of the to-be-remembered event (no immediate imitation is allowed). The two tests also differ because the link between the stimulus and the infant's response is not forged through conditioning in deferred imitation. These distinctions are relevant for theories of memory and its development. Cognitive scientists and neuroscientists have made a distinction in different types of memory systems, particularly between procedural and declarative memory systems. One hypothesis is that deferred imitation provides a technique for exploring a primitive form of declarative memory, which researchers believe is mediated by the medial temporal lobe.

Relations Between Infant Memory and Childhood IQ

Do scores on infant memory tests predict later cognitive performance? Empirical work demonstrates that infants who perform better on tests of memory between two to nine months of age score higher on IQ (intelligence quotient) tests given later in childhood (McCall and Carriger, 1993; Rose and Feldman, 1997). There is a (heated) social policy debate as to whether tests of early memory should be advertised as "infant intelligence tests." This debate should not mask the scientific discovery that there is continuity between the mental performance of infants, as measured by their performance on tests of memory, and childhood IQ scores.

Early and Rapid Memory Formation for Biologically Relevant Signals

Newborn infants are predisposed to encode and remember biologically important signals such as facial and speech signals. Newborn infants, with only a few hours of exposure to the mother, look longer at their own mother's face than at a stranger's face. Newborns also choose to listen to the voice of their own mother than to that of a strange female talker, suggesting that perhaps the sound of the mothers voice is learned prenatally. Finally, recent discoveries show that young infants respond differently to nativeas opposed to foreign-language speech sounds. For example, Patricia Kuhl and colleagues (2001) showed that six-month-old Swedish infants have committed Swedish but not English speech sounds to memory, whereas American infants have done the opposite, demonstrating that infants are listening to the ambient sounds in the environment and remembering them even before they can talk.

Infantile Amnesia Revisited

Research indicates that infants have far more robust and complex memories than classic theories predicted. The puzzling phenomenon of infantile amnesia becomes more of a mystery when considered in light of this modern infancy research, because it can no longer be thought that infants do not form memories or that they are confined to sensorimotor skill routines during the preverbal period (Meltzoff, 1995). It is possible that the amnesia adults experience about their own infancy is due to the extreme mismatch between the cognitive, emotional, and physical context of the initial learning and that of the adult. It is sometimes reported that adults can gain access to "lost" childhood memories by immersing themselves in unique situations they have not encountered since childhood. Alternatively, the fact that adults use a linguistic code may shroud their memories of the preverbal period.

See also:AMNESIA, INFANTILE; CHILDREN, DEVELOPMENT OF MEMORY IN; DISTRIBUTED PRACTICE EFFECTS; INTERFERENCE AND FORGETTING; LANGUAGE LEARNING: HUMANS; OBJECT CONCEPT, DEVELOPMENT OF; REINFORCEMENT

Bibliography

Barr, R., and Hayne, H. (2000). Age-related changes in imitation: Implications for memory development. In C. Rovee-Collier, L. P. Lipsitt, and H. Hayne, eds., Progress in infancy research. Mahwah, NJ: Ablex.

Bremner, J. G. (1994). Infancy, 2nd edition. Cambridge, MA: Blackwell.

Carver, L. J., and Bauer, P. J. (2001). The dawning of a past: The emergence of long-term explicit memory in infancy. Journal of Experimental Psychology: General 130, 726-745.

Carver, L. J., Bauer, P. J., and Nelson, C. A. (2000). Associations between infant brain activity and recall memory. Developmental Science 3, 234-246.

de Haan, M., and Nelson, C. A. (1999). Brain activity differentiates face and object processing in 6-month-old infants. Developmental Psychology 35, 1,113-1,121.

Fagan, J. F., III. (1990). The paired-comparison paradigm and infant intelligence. In A. Diamond, ed., Annals of the New York Academy of Sciences, Vol. 608: The development and neural bases of higher cognitive functions. New York: The New York Academy of Sciences.

Kuhl, P. K., Tsao, F. M., Liu, H. M., Zhang, Y., and de Boer, B. (2001). Language/culture/mind/brain: Progress at the margins between disciplines. In A. R. Damasio et al., eds., Unity of knowledge: The convergence of natural and human science. New York: The New York Academy of Sciences.

Mandler, J. M. (1990). Recall of events of preverbal children. In A. Diamond, ed., Annals of the New York Academy of Sciences, Vol. 608: The development and neural bases of higher cognitive functions. New York: The New York Academy of Sciences.

McCall, R. B., and Carriger, M. S. (1993). A meta-analysis of infant habituation and recognition memory performance as predictors of later IQ. Child Development 64, 57-79.

Meltzoff, A. N. (1988). Infant imitation and memory: Nine-montholds in immediate and deferred tests. Child Development 59, 217-225.

—— (1990). Towards a developmental cognitive science: The implications of cross-modal matching and imitation for the development of representation and memory in infancy. In A. Diamond, ed., Annals of the New York Academy of Sciences, Vol. 608: The development and neural bases of higher cognitive functions. New York: The New York Academy of Sciences.

—— (1995). What infant memory tells us about infantile amnesia: Long-term recall and deferred imitation. Journal of Experimental Child Psychology 59, 497-515.

Meltzoff, A. N., and Moore, M. K. (1998). Object representation, identity, and the paradox of early permanence: Steps toward a new framework. Infant Behavior and Development 21, 201-235.

Nelson, C. A. (1995). The ontogeny of human memory: A cognitive neuroscience perspective. Developmental Psychology 31, 723-738.

Rose, S. A., and Feldman, J. F. (1997). Memory and speed: Their role in the relation of infant information processing to later IQ. Child Development 68, 630-641.

Rovee-Collier, C. (1997). Dissociations in infant memory: Rethinking the development of implicit and explicit memory. Psychological Review 104, 467-498.

Rovee-Collier, C., Hayne, H., and Colombo, M. (2001). The development of implicit and explicit memory. Philadelphia: John Benjamins Publishing Co.

Andrew N.Meltzoff

Revised byAndrew N.Meltzoff

andLeslie J.Carver