Putting Together the Puzzle of Adaptive Constructive Memory

Q&A with Daniel Schacter

The image most often used to describe how memory works is that of a video recorder retaining impressions in real time of each event, and your brain then plays back those impressions when calling up a memory. But that is but a memory myth. The image Harvard cognitive neuroscientist Daniel Schacter draws on to describe memory is instead an act of construction, “something like assembling a jigsaw puzzle from fragmentary bits and pieces.”

Indeed, he calls memory an “adaptive constructive process” – a process he has illuminated throughout his career and for which he is receiving the Distinguished Career Contributions Award from CNS this March at its annual meeting in San Francisco.  “Although reliance on construction can be a source of memory errors, the ‘adaptive’ part of my phrase ‘adaptive constructive processes’ refers to the idea that the same constructive processes that can get us into trouble with memory can also support adaptive functions, such as solving problems or generating simulations of future experiences.”

CNS spoke with Schacter about this process, recent studies that are changing how we think of memory, and what he is most looking forward to it at the CNS annual meeting.

CNS: The title of your award talk includes both memory and imagination. What is the relationship between memory and imagination? And future thinking/planning?

Schacter: The relationship between memory on the one hand, and imagination and future thinking on the other, nicely illustrates the notion of adaptive constructive processes. About a decade ago, Donna Rose Addis and I put forth an idea we called the “constructive episodic simulation hypothesis.” The hypothesis was stimulated by findings from several laboratories, including our own, of striking cognitive and neural similarities when people remember the past and imagine the future.

For example, in an early fMRI study, we found that when people were presented with word cues and either recalled a past experience or imagined an experience that might occur in the future, a common core network of regions that overlaps extensively with the well-known default mode network showed increased activity compared with control conditions. Previous observations from Endel Tulving and others had shown that amnesic patients who have difficulties remembering their personal pasts also have difficulties imagining their personal futures. Similarly, we found that older adults recalled past experiences with less episodic detail than young adults and showed a similar reduction in episodic detail when they imagined future experiences.

CNS: Why is this important?

Schacter: These and related findings suggested that the observed similarities might reflect a common reliance of remembering and imagining on episodic memory retrieval. Putting these observations together with the literature on constructive memory and memory distortion, we suggested that an important function of episodic retrieval is to support the construction of imagined possible future experiences by allowing people to flexibly recombine elements of prior experiences into simulations of situations that might occur in the future. This kind of flexible retrieval and recombination should be adaptive for purposes of thinking about the future because the future is rarely an exact repetition of the past – we want to be able to simulate novel upcoming events in light of our past experiences. However, we argued that this flexible, constructive use of episodic retrieval also has a downside: It can produce memory errors that result from miscombining elements of past experiences or confusing imagined and actual events.

I think that we are seeing a focus on memory not just as a means of retrieving past experiences and stored knowledge, but also an increasing interest in how various forms of memory impact a range of cognitive and affective functions, including future thinking, scene construction, problem solving, decision making, emotion regulation, and creativity.

CNS: What are some recent studies in this area that you are excited by and that you might talk about in your award lecture?

Schacter: As I mentioned earlier, a key part of the constructive episodic simulation hypothesis is the idea that flexible retrieval processes, while serving adaptive functions, can also lead to memory errors. We did not have any direct evidence on that point when we first discussed the idea, but several recent studies in my lab have provided experimental evidence for a link between flexible recombination and memory errors. Other recent studies in the lab have provided novel fMRI evidence concerning the brain regions and network interactions that underpin the involvement of episodic retrieval in future imagining and related functions such as creativity. These studies are exciting to me because they are starting to paint a picture of the mechanisms that support adaptive constructive processes.

CNS: What do you think are some of the biggest changes happening right now in the study of memory? And where do you think we’re heading?

Schacter: In my own lab and in the field more generally, I think that we are seeing a focus on memory not just as a means of retrieving past experiences and stored knowledge, but also an increasing interest in how various forms of memory impact a range of cognitive and affective functions, including future thinking, scene construction, problem solving, decision making, emotion regulation, and creativity. I think that this broadens considerably the domain of memory research, and is a trend that I expect to continue in the future.

CNS: What’s next for your own research?

Schacter: I think it is important to both deepen and broaden our understanding of adaptive constructive processes. Deepen in the sense of understanding better at both cognitive and neural levels the mechanisms that both support adaptive functions and contribute to memory errors. And broaden in the sense of identifying new domains in which these processes operate.

CNS: What are you most looking forward to at the CNS annual meeting in San Francisco this March?

Schacter: The chance to catch up with former lab members and other colleagues attending the meeting, and to give the award lecture to an audience that will include my wife and two daughters. When they were young, there weren’t too many situations where my daughters would sit and listen to me go on for the better part of an hour, but now that they are adults – one of them a developmental psychologist, the other an associate at a global health consulting company – I’m hoping that they will enjoy the lecture.

-Lisa M.P. Munoz