A recent study from the lab of Professor James Russell, shortly to be published in the Journal of Cognition and Development, explores the ability of children to bind representations of themselves to past events that they witnessed.
Episodic memory is commonly interpreted as memory for “What, Where & When” information. This is because remembering what happened, where it happened and when it happened are seen as the essential components of an event, i.e., they are what make that event particular.
In addition to recalling what, where & when information psychologists also regard a special kind of self-conscious awareness as being an essential component of episodic memory. When you have an episodic memory, you not only bring into conscious awareness what, where & when information about the event, but you are also aware that this is an event that you experienced at a past time. Many developmentalists believe that it is this awareness of the self as an experiencer of past events that impedes very young children from having genuine (adult-like) episodic memories.
Our study examined this in more detail. We devised a memory task with 3.5-year-old to 6.5-year-olds, which required children not only to recall a previous event (i.e. the what, where and when information), but also required them to think of themselves as the experiencer of that event at a past time. Children visited a ‘jungle’ to ostensibly play a game in which they listened for animal sounds. The critical event was the occurrence of some monkey calls. When this happened two schematic “trees” lit up. One of the lights was on the child’s left and the other was on the child’s right. The child either saw the left light come on first and then the right one (left-right condition), or the right one and then the left one (right-left condition) or both lights come on together (simultaneous condition). This was the critical sequence that formed the basis of the memory test – a unique event that fits the what (flashing lights), where (on the left or on the right), when (sequential or simultaneous) criteria of episodic memory.
After a 24-hour delay, children watched three videos of the event in question, one showing a left then right light display, one showing a right then left light display and one showing a simultaneous light display. The videos showed the critical event either from the child’s own perspective (from a camera positioned just behind the child’s head at the time of the event – the behind condition) or from a camera positioned on the ceiling looking down on the event (above condition), or from a camera positioned on the other side of the room (the front condition).
The significance of the three different camera positions was that in the behind condition children could pass the task by simply relying on a sense of familiarity. However, in the above and front conditions, familiarity could not help as the videos show the scene from a perspective different from that of the child’s at the time of the event – therefore, all three videos should be equally unfamiliar. In these two conditions the child needs to recall the scene with the understanding something like ‘I was there and I saw the trees light up in such and such a way, therefore that must be me in that video’.
Our findings
Although this was a complex task, children from 4.5 years of age onwards performed above chance. Somewhat surprisingly though it did not matter which of the three camera positions children viewed on day two. They were as equally likely to pass the memory task in the behind, above and front condition, suggesting familiarity was not an aid in this task.
We administered a number of ancillary tasks to children, one of which was a second order Theory of Mind (ToM) task. Unlike standard ToM tasks in which children report what a character actually thinks, this 2nd order task required children to report what a character thinks he knows – this has an embedded representational structure and we speculated it would correlate closely with performance on the jungle memory task. Our data confirmed this, children who passed the 2nd order ToM task were much more likely to pass the memory task, even after differences in age and linguistic ability were accounted for.
However, performance on the jungle task did not correlate with three other ancillary cognitive tasks.
What we learned from this study?
Like other tasks that have been designed to measure episodic memory, the present task was not reliably passed by children until they were at least 4.5 years of age. Does this mean that children younger than this do not have episodic memory? We have previously argued that they have ‘minimal’ episodic memory (Burns, Russell & Russell, 2014). That is, they can recall past events but may not have the understanding that memory is a means of accessing past experiences one has had.
Why might this be the case? The results of the correlational part of our study may shed some light on this. Recall that performance on the jungle task correlated with performance on a task in which children had to embed one mental representation within another. This is also required when thinking of oneself in the past as having a certain experience. The true episodic recollector does not merely represent an experience as causing a current memory (first-order) but represents herself in the past as representing the world in a certain way (second order). We don’t just represent episodic memories: we represent ourselves representing them as our own.
One might ask whether this analysis captures something generally true of episodic memory, or whether our task was too idiosyncratic to yield general conclusions. We think our conclusions hold generally because (a) episodic memory is typically from a point of view and (b) includes the bodily location of the subject in relation to objects. The rememberer needs to include herself in the memory and needs to appreciate that the way she experiences an event determines how she remembers the event. This is why we have given our published paper the title, “Children’s binding of What-Where-When-Who in episodic memory: Identifying oneself in the past”. From our study, it seems children are capable of this after 4.5 years of age.
References
Burns, P., Russell, C., & Russell, J.R. (In press). Children’s binding What-Where-When-Who in episodic memory: Identifying oneself in the past. Journal of Cognition and Development.
Burns, P., Russell, C., & Russell J. (2014). Preschool children’s proto-episodic memory assessed by deferred imitation. Memory. http://dx.doi.org/10.1080/09658211.2014.963625