More to Memory: Elizabeth Loftus

[MUSIC] Welcome back to the Cognition Project, a series exploring the stories of the researchers who created the field of cognitive science and helped to shape it, resulting in a new science of the mind. I'm Tom Griffiths, Professor of Psychology and Computer Science at Princeton University. We have been hearing the stories of some of the early cognitive psychologists, and in particular, people who began to use mathematics to model the mind. One influential group of mathematical psychologists formed at Stanford, with members like Richard Atkinson and Gordon Bauer, who were featured in previous episodes. In the last episode, we heard from Rich Schiffrin what it was like to be a graduate student in that group. Today, we hear a different perspective from Elizabeth Loftus, who grew disenchanted with mathematical psychology, but went on to create a new field studying the psychology of eyewitness testimony. [MUSIC] I went to UCLA as an undergraduate, and I was majoring in mathematics. I loved algebra, I loved geometry, but when it came to calculus, I really didn't love it that much, but I had invested so much time in math that I was determined I was going to finish in math. I started to take some psychology classes. I took intro psychology first, and I absolutely loved it. So I started taking more psychology, and by the time I was done, graduating at UCLA, I had enough credits for a double major in mathematics and psychology. So what was I going to do after that? Well, I heard about this field called mathematical psychology, and I thought, well, that must be perfect for me. And Stanford was the top place in mathematical psychology psychology. So I went to grad school at Stanford, and then I really learned what mathematical psychology was all about, and it just didn't captivate me. I ended up working with two very famous people in mathematical psychology. Pat Sopi's was my PhD advisor, Richard Atkinson, was my master's advisor on my dissertation committee. But what really turned me into an experimental psychologist was some work that I did separately at Stanford with Jonathan Friedman, a social psychologist who was studying semantic memory. So now I started doing these experiments with him on semantic memory, and I absolutely loved the work. The cognitive psychology, it wasn't the name of one of the sub areas of psychology at that point. I think I know there was social and development on maybe clinical psychology. They would be sub areas. I was in the sub area of learning, which later would kind of become cognitive psychology. But I do remember reading Ulrich Nyser's book, Cognitive Psychology in the late 60s, and getting my copy of it autographed by him at a conference, and I was very excited about that. But I then considered myself maybe an expert in or gaining some expertise, let's say, in the area of memory. It was semantic memory then, and I did not start to do the kinds of experiments on eye witness memory until after I was done with graduate school and had launched my career as an assistant professor. The math psych group kind of was housed in a different location than some of the other areas of psychology. And they had Friday seminars, which I continued to attend since I was officially part of this math psych group. The social psychologists, including Friedman and his operation were in a different building on the campus. So I was sort of had my fingers in both places. But my heart was really in the work on semantic memory, because that's when I learned how to design an experiment and actually run the subjects and analyze the data and write up the results for a publication, shepherd the work through the publication process, and become an experimental psychologist. One of the things that Jonathan Friedman and I did was to measure how long it took people to answer questions about the information that they had in semantic memory. And so for example, in one of our experiments, we showed that if you ask people for the name of like a bird that's yellow, people are faster to give you a response than if you ask them for a yellow bird. Or if you ask them for the name of an animal that begins with the letter Z, they're faster than if you give them a cue that says I want a Z word that's also an animal. And they're about 250 milliseconds faster about a quarter of a second. And so we speculated about why this might be true, that you need needed the conceptual category to enter it and then begin your search. And if I gave you the category first, you could begin the process a little earlier. Rather than if I gave you the other cue, you had to wait for the category. So it told us something about how knowledge is organized in the human mind. And so I did many experiments with Jonathan Friedman. I published a paper with Alan Collins, who was a superstar in the semantic memory field. Probably met Alan at a conference and we just started talking about what we were interested in. And he invited me to join him in a publication that ended up because of him in psych review. And so it became one of the more highly cited papers in the semantic memory, thanks to Alan. And I was doing that as an assistant professor, while I was developing strategies for studying eye witness memory, because I had decided I, well, really what happened is I had lunch with a cousin of mine. And I had done all this semantic memory work. My cousin, who was a lawyer, said, well, you're an experimental psychologist. Have you made any discoveries? And I said, yes, I have. I have found people are faster to give you the name of a, you know, a bird that's yellow, then the yellow bird. And she said something like, you know, on how much did taxpayers pay for that result? And I thought, you know, I really would love to do some work that has more obvious, practical applicability. And what would that be? Well, I knew something about memory. I've always been interested in legal issues. And maybe the memory of witnesses for crimes and accidents and other legally relevant events. And that's how I launched myself in that line of work. First, it seemed like it was practically important that the memory of people for crimes and accidents, and it was interesting to me because of a longstanding kind of interest in legal cases, court cases, maybe being kind of a true crime junkie. I also had a former professor who was now working for the US Department of Transportation. And I was telling him about some of the early thoughts I was having about studying eyewitness memory. And he at the Department of Transportation said, well, you know, if you study memory for accidents, there's money at the Department of Transportation, you can get a research grant or something. So that's why my first experiments involved accidents and not crimes because I was able through the advice of this former professor to tap into some funding from the US Department of Transportation. And that was, you know, one of the first grants, research grants, national research grants that I received to do this work. So, what was my going to study? I mean, there's so many factors that can affect the accuracy or completeness of somebody's eyewitness report. But I decided that I would, I'm seized upon the idea of studying the questioning process. How questioning witnesses about their experience might affect what they remember. And that's what some of my early studies looked at. So, for example, in one study that many people know about, we question people about the speed of the vehicles involved in the accident. So we might ask how fast were the cars going when they hit each other? Or for other subject witnesses, how fast were the cars going when they smashed into each other? And we found that just using the word smashed rather than hit, increased the speed that people reported. So just changing a word or two in your question could make a significant difference in how fast people remembered the vehicles were going. And then what we showed is those early questions can not only affect the answer somebody gives right away, but can have a long range effect, can affect the answer to some other question that you put to them often much, much later. So we did a follow-up study after people asked, were asked the smash question or they hit question, come back to them a week later, remember that accident we showed you? And did you see any broken glass? There was no broken glass and yet people to whom we had used the word smashed in questioning them, were twice as likely to say they saw this non-existent glass. And so that just told me that these questions are, they're not just affecting the immediate answer, they're causing maybe a reconstruction of the event and when you come back to it later, you are accessing this altered representation and answering the new questions based on that. And then we did many other studies where we showed that leading questions could affect the answer somebody gave. Did you see some children on the playground? Versus did you see any children on the playground? Did you see the broken headlight? Versus did you see a broken headlight? Just changing a word or two can affect the answer somebody gives. And that's what I was busy doing in the 1970s when I was launching this program of work, thanks to the Department of Transportation. Back at that time, memory scientists who would say they were studying memory were not showing people films of accidents and crimes. So this was, they were using stimuli that were selected to be deliberately more palette and simple, wordless, occasionally isolated pictures but not films of crimes and accidents. They were very busy. They were at consent with both doing his mathematical psychology work and also at the same time doing the computer assistant instruction projects that he was doing along with Snoopies. And so it was a little hard to get FaceTime with these busy professors. Snoopies has sadly passed away. I mean, I do have some very fond memories. I'm married a fellow grad student. That's how come my name is Loftus. Now Jeff Loftus, we got married in 1968 and Atkinson came to the wedding in Southern California. That was very sweet. Snoopies didn't come, but that's okay. Loftus was one of the few women in the mathematical psychology program at Stanford in the 1960s. There is that photo that circulates a lot. That would have been taken in 1966 or early '67. And yeah, I guess then possibly the only woman. It was a photo of faculty and graduate students so you can see some other young shift friend, for example, and some other graduate students in that photo. But then, soon after, more women came. Bobby Klatsky came within a year or two and she became a friend. And then others, these Friday seminars where people would present their mathematical, their formulas and trying to use mathematics to predict behavior and so on. And I just was not all that interested in that work and I would sometimes sit in the back and write letters to my uncle who lived across the country or I'd hem scurts, did salt, you know, hemming. And I don't think that the fellow students thought that I was gonna have much chance of succeeding in this profession. I had heard that they voted me most likely to drop out kind of thing, but that didn't happen. I will say one thing, as long as students who may listen to this or hear this or think about this, many times graduate students. And I have said this when I've met with groups of grad students here at UC Irvine, where I still teaching. You may wonder what you're gonna do and what kind of work you wanna do and who you wanna be and how you wanna identify yourself and so on, but you may find that that thing doesn't come until even after you get your PhD because my whole career now is started with work that was after I was already out of graduate school. So don't worry, I say to them, if you're not sure what it's gonna be, just do something and let it lead you wherever it leads. I love what I do. So there isn't anything else. I mean, before I decided I wanted to go on to grad school, I thought I might be a private eye or something, or for a while I thought I might teach, be a teach mathematics, like high school math or something. But I think that being a college or university professor is just a fantastic life, especially if you like the work you're doing because there's just a great, perfect amount of structure, yet freedom for entrepreneurial exploration. - The Cognition Project was created by me, Tom Griffiths and it's produced by Yael Evan Orr with support from the Solomon Argoogonheim Foundation, the University of California Berkeley and Princeton University. Our engineer is Taric Fudah. We thank Brian, Christian, Tanya, Labroso and our interviewees for their help in making this project possible. If you're interested in more of a story of the origins of cognitive science and how it led to many of the ideas behind modern AI, you can find it in my new book, The Laws of Thought.