Infantile amnesia refers to the general inability of people to remember specific events from the early years of their lives. On the basis of both free recall studies (What is your earliest memory?) and studies for memories of notable and datable early events (the birth of a sibling, hospitalization, the Kennedy assasination) psychologists have concluded that there are very few memories from before the age 3 years. The average age of the earliest memory reported is 31/2, with a small but consistent gender difference indicating that females reported earlier memories. If people are asked to recall episodes from the entire life span the number reported before age 8 falls off sharply in comparison to other periods. This indicates that it is not the age of the memories per se that accounts for their relative paucity, rather it is the life period that they occur within (the earliest years) that is sparsely represented in long term autobiographical memory.
Sigmund Freud was one of the first psychologists to identify the phenomenon of infantile amnesia:
"What I have in mind is the peculiar amnesia which, in the case of most people, though by no means all, hides the earliest beginnings of their childhood up to their sixth or eight year." (Freud, 1905/1953)
His notion was that early childhood memories, particularly sexual ones, were too frightening and distasteful to the child to be preserved as such. Instead these emotional early memories required filtering from conscious awareness, and so took the form of more innocuous and seemingly inconsequential "screen memories". A major difficulty that arises in considering Freud's account as a complete analysis of the phenomenon is that not all reported early memories are emotionally neutral or concerned with trivialities. Memories of troubling experiences from early childhood appear to be no less common than negative memories from adulthood. In their excellent review articles Pillemer and White (1989, White and Pillemer, 1979) point out that in this, as in many things, Freud was multivoiced. In addition to his 'screen' or 'blockade' model he also proposed a 'selective reconstruction' model. The general premise of this 'selective reconstruction' model was that the inaccessibility of early childhood memories was due to a disjunction between the earliest and later modes of processing information. Modern theoretical accounts of infantile amnesia are more in line with the selective reconstruction than the blockade model. There is so diversity of opinion about the cause of the shift in processing that I will explore in the remainder of this lecture.
One of the prime candidates for the shift in processing is neurological development. Researchers have asked if the infant simply lacks the neurological equipment necessary for memory formation and storage. It is true that in the newborn, the hippocampus and the frontal lobes are immature. The period from 8 to 24 months is a period of intense synaptogenesis in frontal cortex. PET studies during this period of time show vastly increased neurological function in the frontal cortex. After 2 years the processes of myelination and increases in synaptic efficiency continue at a slower pace. While there is no abrupt neurological watershed corresponding to the offset of infantile amnesia, there is good reason to believe that neurological maturity must be one of the factors that limits early memory.
We can ask what it is possible for infants to remember given their relative neurological immaturity. A host of recent 'clever baby' studies have demonstrated that memory is more sophisticated in infancy than previously thought.
Meltzoff (1995) used the deferred imitation paradigm to demonstrate sophisticated memory abiltities in infants. Children of 14 and 16 months watched an adult perform a novel activity once. For example, the experimenter touched the top of a box with his forehead and it lit up. Four months later these infants were more likely to produce the same action as the experimenter than children who had not seen the previous action. Four months is a long time for a 14 month old to remember that a box will light up if you place your forehead upon it. It is not appropriate to label this ability simply motor or procedural for two reasons: the infants did not experience the extended period of practice that is characteristic of learning motor skills, and because the exposure was so much briefer than any habit based memory would allow. Meltzoff demonstrated that even young infants are capable of storing abstract representations of action sequences in long term memory, not simply learned motor responses.
Rovee Collier and colleagues used the mobile conjugate reinforcement paradigm to demonstrate memory in infants as young as 3 months. The infant's ankle was connected via a ribbon to an overhead mobile with the goal of having the infant learn the contingency between his or her kicking and movement in the mobile. In some of the experiments, an additional reactivation treatment is administered prior to tests of retention. Here, following an interval in which some forgetting has occurred, infants get a simple reminder of the set-up (e.g., the mobile, the room, the mobile moving) and are then tested at some later point (at least 24 hours after the reminder). Older infants (6 month olds) require more specific cues for retrieval than younger infants (3 month olds). Once reactivated a memory trace can be retained for at least as long as when it was originally encoded. If the infant's environment provides frequent opportunities for reactivation then theoretically "an individual's early experiences could be remembered over a lifetime". As a consequence of these studies Rovee-Collier et al stressed the importance of reactivation. As Howe & Courage (1993) express it:
Contrary to traditional views of forgetting in which traces are either present but inaccessible (retrieval failure) or simply absent from memory (storage failure in extremis), the memory modifications brought about by reactivation emphasize the dynamic properties of memory.
This is consistent with the view of adult memory drawn by Scachter, particularly his point that "memory is an emergent property of the cue and the engram" (p. 71). In other words, remembering changes what is "stored".
Pillemer & White (1989) propose that infantile amnesia is overcome through the linguistic sharing of memories with others. They view memory as composed of two functionally independent systems:
| Dual Memory System | |
|---|---|
| Nonverbal, Image-based System | Socially Accessible System |
|
|
Various pieces of empirical evidence support the idea of a later onset of narrative memory. In a study of vivid ("flashbulb") memories Winograd & Killinger (1983) found that the number of informational categories of narrative (who, where, when, what,...) increased with age: those under 5 produced only 2 or 3, whereas those 6 or older produced 4 or 5. Eisenberg studied two girls between the ages of 2 and 3 and found that the percentage of conversations in which they discussed unique rather than routine aspects of past events increased sharply over the year. Simultaneously, the percentage of memories that were elicited by an adult's question decreased. Tessler (1986) studied 3 year olds and their mothers on a visit to a natural history museum. She found that none of the children remembered any of the objects that they viewed in the museum if they had not talked about them with their mothers. Mother's talk alone was not facilitative, nor was a child's mention alone effective in leading to subsequent remembering. Creating a narrative together cemented particular objects in memory.
Pillemer & White endorse a dialectical or Vygotskian model rather than one of simple cultural transmission or socialization. Through sharing memories with others language becomes available as a means of reinstating memory. After overt recounting becomes established, covert recounting to oneself may take place, and take on the function of reinstatement. Language has a dual function: communication and mental representation.
This is Howe & Courage's (1993) theory: they see the developmental changes in sense of self over the first 21/2 years of life as pivotal in the development of autobiographical memory. Upon reviewing a large body of literature, their conclusion is that before the age of 2-3 years children lack the concept of self necessary to build enduring personal memories.
The development of a sense of self is chronicalled mainly through visual recognition, or mirror tests. By about 3 months of age infants are extremely attentive and positive toward mirrors; as a consequence much early "play" consists of hilarious games involving approaching and retreating from the faces in the mirror. By 9 months infants show awareness of the meaning of the mirror image through their use of the tandem movement of the image and themselves for deliberate play and imitation. Further, their knowledge of the reflective property of mirrors is demonstrated by their turning to see someone who approaches in the mirror. At the age of approximately 18 months (although this can vary widely) full visual self recognition is demonstrated in the rouge test: the infant will try to wipe a red blob off her own nose rather than reach out for the mirror. By about 22 months infants can correctly label their image.
Linguistic markers also evidence the growing sense of self. The personal pronouns I and me are acquired first at approximately 22 months, "you" follows about 2 months later. Normal 2 year olds rarely mix up I/me with you. At about 2 toddlers begin to use grammatical morphemes that indicate tense. The first to appear is "-ing" to refer to ongoing events, followed by the introduction of the past tense.
While Howe and Courage make much of the difference between their "self" hypothesis and Pillemer & White's "socially accessible memory system" hypothesis, they can be viewed as two sides of the same coin. The concept of self and the linguistic ability necessary to share narrative memories are related developments.
Is it due to the lack of a theory of mind?
Between 3 and 4 years of age there is a dramatic watershed in the child's ability to take another's perspective. 3 year olds find it difficult to attribute mental states to others that differ from their own present state of mind. The critical experimental paradigm used to demonstrate this phenomenon is the false belief task. If a child is shown a box of candy (the British Smarties) which contains pencils and is asked what they think it contains they naturally reply "Smarties". They are then shown that they were tricked and its true contents are revealed and asked what they think their friend who is waiting to play the game will think. If the child is less than 4 he or she will usually respond that their friend will think that the box contains pencils. They are unable to attribute a false belief that differs from their own knowledge to another. The ability to create another person's perspective and to explain the motivations of others is a key aspect of story telling, so this account dovetails well with Pillemer's "socially accessible memory system" hypothesis.
Is it due to a lack of understanding?
Other aspects of children's understanding in addition to the theory-of-mind leap play a role in the formation of a narrative memory. Pillemer, Pacariello & Pruett's 'fire alarm study' provides a good demonstration of this. 31/2 and 41/2 year olds were interviewed 2 weeks and a year after an emergency evacuation of their preschool because burning popcorn set off the fire alarm. All preschoolers could answer some memory questions at Time 1 (2 weeks) but the 41/2 year olds produced more coherent narratives and showed a better understanding of the temporal and causal sequence of events. At Time 2 (7 years) 57% of the older group (41/2 year olds) provided at least a fragmentary narrative memory and 86% selected their correct classroom location. Only 18% of the younger group (31/2 year olds) produced even a fragmentary verbal response, and their identification of classroom location was no better than chance.
Sources
Goleman, Daniel (1993). Studying the Secrets of Childhood Memory, New York Times, Tuesday, April 6, pages C1 & C11.Back to text
Howe, M. & Courage, M. (1993). On resolving the enigma of infantile amnesia. Psychological Bulletin, 113, 305-326.Back to text
Pillemer, D.B. (1998). Momentous Events, Vivid Memories. Cambridge, MA: Harvard University Press.
Pillemer, D.B., Picariello, M.L. & Pruett, J.C. (1994) Very long-term memories of a salient preschool event. Applied Cognitive Psychology, 8, 95-106.
Pillemer, D. B., & White, S. H. (1989). Childhood events recalled by children and adults. Advances in Child Development and Behavior, 21, 297-340.Back to text
White, S.H. & Pillemer, D.B. (1979) Childhood amnesia and the development of a socially accessible memory system. In J.F. Kihlstrom and F.J. Evans (Eds.) Functional disorders of memory (pp. 29-73). Hillsdale, NJ: Erlbaum.