Future Directions Multiple Systems across the Lifespan

Memory Professor System

10x Your Memory Power

Get Instant Access

Up to this point autobiographical memories have been considered as single entities. An alternative view is that they consist of information stored in component processes, each process occurring in a separate behaviorally and neurally defined system (Rubin, 1995, 1998; Rubin & Greenberg, 1998, in press, 2000; Rubin, Greenberg & Schrauf, 2000). According to available evidence, these component processes are an integrative memory system (which is classically attributed to the hippocampus and more recently expanded to include the frontal lobes), imagery in each individual modality and a multimodal spatial-imagery system, language, narrative reasoning, and emotions. On this view, having a full-blown autobiographical memory requires the use of the integrative memory system, at least one modality-specific imagery system (usually visual imagery), spatial imagery, to varying degrees imagery in the other senses, language, narrative reasoning, and emotions. For different memories within an individual and for different clinical populations, the degree to which each system contributes varies. Thus, for instance, flashbulb memories are more likely to have considerable visual and spatial imagery, and people who are depressed or who have posttraumatic-stress disorder are likely to have voluntary memories that lack sensory details and narrative coherence. In contrast, the intrusive memories of people with posttraumatic-stress disorder are likely to have considerable sensory detail and, compared to other memories, sensory details in the olfactory and gustatory system. There is considerable support for this view from behavior (Rubin, Greenberg & Schrauf, 2000) and neuropsychology (Rubin & Greenberg, 1998, 2000), but what does it mean for understanding the bump?

How do the individual processes needed to produce a full-blown autobiographical memory develop over the lifespan? It is hard to know for sure because the processes are usually not studied developmentally in a way that would provide a direct answer. Figure 10 is my best guess. It is based on a what has been measured and some poorly supported assumptions, which I will make explicit. Figure 10 serves to clarify our level of ignorance and to frame questions for future investigation. I assume that the integrative memory system develops and declines something like fluid intelligence, that the language system develops something like vocabulary, that the narrative reasoning system and especially the narrative skills needed to tell a life story develop more slowly than language (Habermas & Bluck, 2000), and that visual imagery declines with age. I have no idea of how to plot the development of the system that processes emotion. From figure 10, the bump would result from the interaction of the component systems and how they interact with the environment present at the time. For example, under such a view, flashbulb memories, which require a strong visual component, would be more likely to form earlier in the lifespan (see Cohen, Conway & Maylor, 1994, for some support), and childhood memories should have less narrative coherence. Similarly, the same traumatic event would be more

Working Memory Lifespan

Figure 10

Hypothetical developmental trajectories of abilities needed to form autobiographical memories.

Figure 10

Hypothetical developmental trajectories of abilities needed to form autobiographical memories.

likely to produce a vivid intrusive memory not well integrated into the person's life story if it occurred early in the lifespan, when imagery was fully functioning and narrative coherence was not, as opposed to if it occurred later in the lifespan, when the imagery system was weaker and narrative reasoning was stronger.

Earlier the section "Differences in the Properties of Autobiographical Memories'' provided evidence that the ratings of memories do not differ over the lifespan; there are just more memories from the bump period. There is an apparent discrepancy between this finding and the theory just put forth. If, for instance, visual-imagery ability decreases and narrative-reasoning ability increases over the lifespan, then one might expect older adults to rate memories from youth higher in imagery and lower in narrative coherence. Two possible resolutions exist. First, the ratings we have are measures of the ability of older adults now to produce a visual image and a coherent narrative from what was encoded a long time ago. Thus, the effects of differences in encoding might be minimal in comparison with the effects of current retrieval abilities. As reviewed earlier, we know that cuing with high-imagery words does produce older memories, and so it still might be the case that cues that differentially tap systems may be able to access memories from periods when a particular system was functioning best.

The second possible resolution is that if all memories could be accessed, there might be differences in the properties of the memories across the lifespan, whereas for the memories that come most easily to mind, this is not observed, because what makes the memories easy to recall is that they are high on several of the properties rated. An experiment to approximate an exhaustive search would have older adults spend an hour or two recalling and rating memories from a single year from when they were 10, 25, and 50. One might expect the memories from age 10 to be rated higher on imagery and lower on narrative coherence relative to the memories from age 50, whereas the bump memories might rated highly on both scales. I know of no data on these issues.

Was this article helpful?

0 0

Post a comment