An interview with Zachariah Reagh, a brain scientist at Washington University.
Do you remember what you did on this day six months ago? Do you remember who you were with? Or where you were?
Zachariah Reagh, Assistant Professor of Psychological & Brain Sciences at Washington University in St. Louis, has a new paper out in Nature Communications about how the brain remembers everyday-events by making mental sketches of people and contexts and copy-pasting them around different scenarios.
I’m a science journalist, and I had the pleasure of chatting with Raegh for Behind The Science, my ITM column where I unpack the behind-the-scenes of science changing the way we think about the world.
Tell me a little bit about what your lab does. I know you’re interested in how the human brain parse, store, and retrieve dynamic experiences, like those we encounter in our everyday lives, and how these processes change in healthy aging and with age-related pathology.
At the core, I study human memory, and how human memory changes as we get older. Specifically, I study memory for events.
The thing about memory is that it's not just one thing, there are a lot of different kinds of memory and a lot of different ways your memory can do its job. You've got procedural memory — so how you ride a bike, and other similar things that you don't necessarily need to explain, but just learn and know how to do them. Similarly, there are other implicit memories, like smelling something that you know tastes good might make you salivate. Those are things that your body has just kind of learned how to do.
Then there are more explicit forms of memory. For example, if you asked me what I did yesterday, and I tell you about it, that's an explicit type of memory. And within that domain, you've got a few other different kinds: semantic memory, which is like remembering the capital of Washington State, episodic memory, which is describing the things that happened to you yesterday, and so forth.
My research is more in the realm of episodic memory side.
This recent paper that you published on how the brain responds to everyday events, and what mechanisms are at play, how did this come about?
One of the things that we've known about episodic memory, is that your memory of events is made up of different pieces and different processes. For a long time, folks like myself have been obsessing over this idea of “item” versus “context”. What I mean by that is, if you go sit down and have lunch with your friend, the cafe that you're in is the context. And you know, the coffee cup in front of you is an item. Or even maybe your friend is an item within that context. If you think about it, that's a really intuitive way to sort out how you deal with your experiences.
On the one hand, people have been doing experiments that have shown it could work this way, but those experiments have generally been really, really simplistic: here's a picture of an apple overlaid on a picture of a scene. And I'm not knocking simplicity, simplicity is great, it helps you get down to the core pieces of what’s happening, but our real-world experiences are a lot messier and grosser than that. I walk into a coffee shop, there are people everywhere, and there are cups everywhere. There's a counter full of pastries, and no one's telling me okay, these are the items, and here's the context that you're in.
Over time, it started bugging me that we have no idea if that's how your brain is pulling these different pieces apart and storing them as memories.
On the other hand, people like myself, who study memory, have started using really complex things like Hollywood films or television shows, and audio stories to study how we remember things. But those are expertly constructed to capture your attention and be interesting, so that's also not quite like real life, right? There's no director who's making sure your day-to-day experiences are fun.
What this study really represented is somewhere in the middle. So, can we actually take something that looks like real life where your brain actually does what it's doing in everyday life and not what we're just telling them to do in the lab? And bake in these different elements of memories, like focal items and contexts? And if we set things up in that way, just like your real-world experiences, does your brain then pull those pieces apart the way we think theoretically that it should?
What was your approach to finding that middle ground?
I first tried to find episodes of sitcoms where the same characters show up in the same places. Naturally, my mind gravitated towards “Friends” because there are so many episodes, and the same people are constantly showing up in the same places. But the problem with that was that, again, it’s a fun show, not an experiment.
So, I had to do it myself. I took a GoPro and went to a few places in Davis, California, and took two of my colleagues and just filmed them: we went to a couple of different coffee shops, and a couple of different grocery stores, and I just gave them very vague instructions such as “sit down”, “open your laptop”, “start doing some work”, “someone's going to bring you a coffee”, and so forth. I was going for this slice of life, that you would actually see if you walked into a coffee shop.. The key part of the design were the two people — Lisa and Tommy — each person would show up in four different context (in two different coffee shops, and in two different grocery stores).
That allowed us to look at what the brain was doing while watching these different events unfolded. Are there parts of your memory system that care about when Lisa is there, no matter where she is? No matter what context it is? Are there parts of the brain that show a really nice, stable signal when you're in this particular coffee shop?
Then we went one layer up and asked if there were areas that just liked it being in a coffee shop — no matter which one. It sounds a little counterintuitive, but if you think about it, you can go to any coffee shop in the world, and know what's probably going to happen. You know that you need to order your drink, you need to find a table to sit down and sip it. That’s what we call a “schema”.
In the experiment, we showed people these videos and had them recall those events while we scanned their brain activity. This allowed us to compare activity between when they were watching the event unfold, and when they were remembering what that event was like.
And what did you discover?
What we found is that there are indeed parts of the brain that lay down these patterns associated with a specific person or a specific context.
There was one network of cortical areas involved in episodic memory, that researchers have started calling the posterior medial network because it's in the posterior part of the brain and closer to the medial surface or the midline. These areas care more about context. So, when you were watching an event unfold, or when you were remembering an event that took place in a specific context, those patterns look similar, regardless of who was there.
And there were a couple of other things we found too. The medial part of the prefrontal cortex had similar patterns of activity that were present anytime something was happening in a coffee shop. This was probably my favorite result, just because it's so mind-blowing. I guess that there are parts of your brain that know when you're in a certain type of situation.
Finally, we found out that one area of the brain, the hippocampus, only liked a specific event. So this particular person in this particular context, and if you change either the person or the context, the pattern totally destabilised. And this isn't really surprising. If you know about the hippocampus, it's an area that makes very specific memories. Therefore, it makes sense that it would be a very hyper-specific combination of things.
Why does all of this matter and to what extent do these findings have implications for how we actually understand the world around us?
So first off, we were somewhat right about how we think memory works. And at least in my biased opinion, I think this makes an important push to say our memory probably works this way in the real world.
But I think there are some other cool things that can be taken from this. A main one is that your brain is lazy. And it's efficient. So, you can think about an alternative situation where every time something new happens to you, you've got to build from the ground up what that representation looks like. Like a kid with Legos, you're just having to piece it together from scratch every time. But what the brain seems to have is a template for what this person should look like when you're encoding or remembering an event. Or it has a template for what this coffee shop looks like, or what any coffee shop might be like. This laziness is really good. Actually, it lets you flexibly create an event representation on the fly, which we do all the time.
Going back to the concept of “schema”: Does it group coffee shops? Does it group all dining experiences? Does it group the act of having to be sat in an enclosed environment? Doing two things at once maybe like eating and talking?
That's a far better question than you realize. Schema is used very flexibly. In my case, for example, it's just very simplistically saying, okay, here are coffee shops, here are grocery stores. But a coffee shop has a lot in common with a standard restaurant, or it has a whole lot in common with a bar really, right? It's just instead of coffee, you might be getting a beer in this other situation. Everything is very similar otherwise, would the areas of the brain that we found here doing the schema stuff look the same? If they went to a bar? Maybe. I don't really know. It might depend on the situation.
But you also have schemas for situations. For example, you have a schema for what it's like to be on a date, which could be in a coffee shop, or it could be in a movie theater. We don't really know what the limits and boundaries of this are. But I think we as a field need to figure it out.
I guess I also have the same question for the other element of this — the item-focused networks. To what degree do we know that network would light up if it's somebody I know or somebody I care about? Or would it be different? Would it light up if it was my pet?
A lot of those studies that I was talking about, that showed a picture of an apple over picture, tend to use objects as the items. But what I'm seeing here is that a person can actually be treated by these brain systems as an item.
I mean, emotion and affect, in general, are one of those things that color a lot of processes in the brain. And I often joke that memory researchers are a little bit afraid of emotion, because it takes the simple, nice stories that we like to tell ourselves about how the brain works, and then throws a wrench into it.
Yes, how does emotion play a role in this whole differentiation of mechanisms?
What's kind of funny in this study was that we tried to capture these slice-of-life moments. But there's one thing that we're really missing here: meaning. Your actual day-to-day moments. some of them sure you just walk into a coffee shop, and there's nothing on your mind, and you're just observing what's happening, but a lot of your life is woven together by stories. So why did you get there? Where are you going next? Are you at the coffee shop with a friend? What are you meeting about? That's completely absent from this experiment.
I think what I've learned from this experiment, is that I am ready to dive into the more complex side of things such as stories and narratives.
How does this fit in with the bigger picture of the research that you're doing? And how do we go on from here?
One of the things that I'm interested in pursuing is: what happens to all this as we get older? And we've got data that’s been snowballing over the past couple of decades, showing that these memory systems might be impacted differently by age-related pathology. The context and item-related networks, for example, are affected differently by Alzheimer's disease. It’s maybe a golden opportunity that we're able to separate what these networks are doing by researching it. It’s an avenue that I'm pretty excited about, which I think is really needed.
This actually relates to some work I did in graduate school, where we found that your ability to make really high-fidelity memories of objects is more compromised in aging than contextual information. But that, again, that's one of those really simplistic studies.