Interactions between statistical prediction and episodic memory
Every day is a new day, but much of our experience is consistent across days. That is: While we are constantly encountering new information, we are also constantly encountering information that is richly structured and predictable. Human memory systems reflect this distinction. It is thought that there are separate memory systems that process unique details of individual experiences (episodic memory; e.g., your last birthday), versus shared properties across experiences (statistical learning; e.g., how you tend to spend your birthday). But how do these distinct memory systems function in the mind and brain? Do they operate independently, or in parallel, or in opposition? After all, any given experience contains both new, unpredictable information as well as structured, predictable information. Traditional memory systems theories posit that these two memory systems operate fully independently, in separate regions of the brain. In this talk, I will explore three lines of work demonstrating that, instead, episodic memory and statistical learning can adaptively constrain one another, perhaps due to their shared reliance on a single brain region (the hippocampus).