— Interviewed by Bridget O'Brian for Columbia News, http://news.columbia.edu/research/3376
Photo Courtesy Columbia News

People often talk about different parts of the brain learning about the world in different ways: whether memorizing facts and events, or slowly learning habits. In her experiments of how the human brain learns, psychology professor Daphna Shohamy has found that there’s a lot more “cross-talk” between these two forms of learning than most people think, raising questions about what different parts of the brain are doing as people learn, and how what we learn affects the decisions we make.

"Many people think about decisions as being rational vs. irrational," she says. "For neuroscientists the question is: rational or not, what are the brain mechanisms?"

As one of the scientists at the Mortimer B. Zuckerman Mind Brain Behavior Institute, Shohamy is among dozens of Columbia professors unraveling the mechanisms of the mind and brain in interdisciplinary and collaborative ways. Shohamy herself exemplifies this approach, having graduated from Tel Aviv University with a double major in the humanities and neuroscience.

She came to her chosen field because of a long-held curiosity about how people think in different situations. She lived in seven different cities growing up as the family moved often so her parents could pursue interesting opportunities. "Moving around so much made me aware of how our experiences leave a lasting impression, changing our brains, changing who we are," she says.

“The idea that so much of what we do is biological fascinated me.”

Q How did you get interested in researching the brain?

A I was always interested in why people do what they do. When I was in my first semester as an undergraduate, I took what was then called a biopsychology course, which is equivalent to the Mind, Brain and Behavior survey course I teach here. The idea that so much of what we do is biological fascinated me. It still does. It inspired me to go to graduate school to study neuroscience, focusing on memory and decision making.

Q What are some of the differences between the field of cognitive psychology then and the kind of interdisciplinary neuroscience that’s prevalent now?

A Memory as a discipline is relatively old. In fact, most textbooks had a chapter on learning and memory and then a separate one on decision making. What I was interested in is how those things connect to each other. I was intrigued by the idea that we don’t just form memories arbitrarily, and decisions aren’t made in a vacuum, but that decisions we make are influenced by experiences we’ve had. How our brains encode those experiences changes the way we make decisions large and small.

Q In your research you ask the question, when are decisions guided by explicit knowledge and when by implicitly learned associations or biases? Can you talk about these two different brain systems?

A Our brain learns different kinds of information in different ways. The different parts of the brain that learn in different ways also have distinct neural architecture, structure and connectivity with the rest of the brain. One system is referred to as an episodic system and depends on the hippocampus. We need the hippocampus to answer a question such as “What did you have for breakfast yesterday?” If I ask you that, you see not just the toast but who you were with and where and when it was. Episodic memories are rich; they encode linked information about time and space. Then there’s a different kind of memory, like when you come home and automatically pull out the keys from your bag. We think of that more like learning a habit, and it depends on the striatum [part of the brain]. It describes behavior that is also learned from the past, but usually based on a gradual accumulation of experiences, not just one, and often happens without explicit memory. So one way of thinking of these systems is that one is for episodes and the other for the sort of actions or habits that are learned over many experiences. Of course, these systems are different, but there is also cross-talk between them.

Q You’ve said that from the very beginning, you had the intuition that this notion of two separate systems—habit versus episodic memory—didn’t make sense. Can you talk more about that?

A Initially people thought of these systems as being completely separate. They use different parts of the brain, they’re each encoding different kinds of information, and at any given moment you can use one and not the other. Now there’s some evidence that maybe they’re not separate. We know how highly networked the brain is. I felt there has to be a good way to understand the interaction between these systems, which we are still working on. It’s complicated because how they interact depends on what decisions you have to make, and what actions you are going to take.

Q Where does this leave brain science?

A The simple explanation can’t be right, but we haven’t yet developed as clear an alternative model as we would like. We have some scientific evidence supporting [the idea] that the different systems interact differently depending on what you’re trying to do. There are circumstances in which habits will take over, like when you’re distracted and don’t need to retrieve episodic memories. But in other cases you want to combine the two because those memories carry different kinds of information. For example, many of the decisions we make are new ones; a challenge we are especially interested in is how we generalize what we’ve learned but also maintain specific, distinct memories.

Q One thing you’ve been researching is the role of the hippocampus, which does more than previously thought. What is its role?

A We’ve known for a long time that the hippocampus is important for creating explicit memories that you can verbalize. We’re also discovering that it seems to be important for a lot of memory-related behaviors, even perception and automatic decisions. We’re starting to learn more about how the hippocampus can guide decision making about things that we haven’t actually encountered before. A big challenge in education, for example, is how do we teach children skills that they can then transfer to the real world? It’s a question of learning something in one context and being able to use it in others. We have to have mechanisms in the brain allowing information learned in one setting to be used in other settings, and we think the hippocampus contributes to this.

Q How did you come to study Parkinson’s disease in connection with your research?

A An influential paper came out in my first year of graduate school that said dopamine allowed us to learn to predict rewards and then, based on experience, update our expectations about rewards. That same year a totally unrelated paper also came out about the fact that Parkinson’s patients have difficulty learning new habits. Now, in Parkinson’s disease there is a dramatic loss of dopamine neurons. By the time people have their first symptoms, they’ve already lost 70 to 80 percent of them. So I wondered if those two things were related, and whether what we think of as habit learning is something we could understand in terms of the emerging role of dopamine in learning. I started looking into the disease as a window into the role of dopamine in learning and decision making. Dopamine is important for a lot of behaviors, especially for the gradual formation of habits. There aren’t a lot of new habits you need to form at the later stages in life. But experiments have shown that is where we see people with Parkinson’s disease having a lot of trouble.

Q Are you also looking into other diseases to further your understanding of the brain?

A Disease can often be a useful window into how things work by looking at how they don’t work, and we learn something about the disease itself which hopefully ends up being useful. As we struggle to understand brain and behavior, one challenge that is relevant for many psychiatric disorders, such as addiction, eating disorders and others, is why—when people know what’s healthy and what isn’t—their behaviors don’t follow what we know. We all do this occasionally; we know what we’re supposed to do and we just don’t do it. But it’s not enough to tell an anorexic to just eat. Together with my colleagues Joanna Steinglass and Tim Walsh at Columbia University Medical Center, we’ve been exploring the idea with anorexia that starvation itself may essentially be a bad habit which, like a lot of habits, becomes reinforcing over time. This same idea is relevant for people suffering from addiction and many other conditions.

Q You teach a seminar called Cognitive Neuroscience and the Media. How did that start?

A I got the idea for the course the first year I started teaching Mind, Brain and Behavior to undergraduates. Some very ill-informed op-eds about brain science were published around the same time. For me this raised questions about our role as scientists in communicating our work publicly and also about our role as educators in preparing Columbia students to be discriminating consumers of science in the popular press. In the course I have students read media coverage of an issue or discovery, and then we read the scientific article on which it’s based, and we analyze the ways in which the media story deviates. There’s a lot of, “Oh my God, I can’t believe they wrote that.” What’s very humbling is at the very last class we get together with Marguerite Holloway’s students from the Journalism School’s science journalism program and we pair the students up to discuss the articles. Then my students realize, “Oh, you only have three hours to write this article?” It’s very illuminating for all of us.

Q Is there a part of neuroscience that fascinates you the most?

A For me, the most interesting decision-making problems are the everyday ones, the ones where we’re not stopping to think. And yet even these seemingly simple behaviors are more complex than we realize. When my daughter was 4, there was a period of about a year when she would keep asking me, “So wait, when I’m laughing, is that the brain?” And I’d say, “Yes.” “And when you tickle me, is that the brain?” The answer was always yes, and she just couldn’t get over the fact that everything we do stems from the brain. I share that enthusiasm—it’s such an exciting time to be doing research in mind and brain, and to do it here at Columbia. I have found that some of the most intriguing questions and greatest insights start from a conversation over coffee or a beer with a scientist who comes to things from a perspective other than my own. I think there is immense promise in the Zuckerman Institute, which is designed with this idea at its core. We can’t predict what it will lead to—hopefully, to important new answers to old questions, but also new questions that we can’t yet imagine.