Trial Transcript: Dr. Elizabeth Loftus, 1976

This is the transcript for the testimony of Dr. Elizabeth Loftus, an expert witness on human memory called by the defense. David Yocom represented the State of Utah, and John O’Connell was Ted Bundy’s defense attorney. This testimony was given on Wednesday, February 25, 1976, at the Third District Court in Salt Lake City.

DIRECT EXAMINATION BY MR. O’CONNELL:

Q: Will you state your name and address, please?

A: Yes. Elizabeth Loftus, and I live at 1105 Massachusetts Avenue, in Cambridge, Massachusetts, just for this year.

Q: And what is your occupation or profession?

A: I am an Associate Professor of Psychology in the Department of Psychology at the University of Washington, on leave.

Q: And where are you currently?

A: I am at Harvard University this year.

Q: And can you sketch your educational background for the Court, please?

A: Yes. Do you mind if I refer to my data, since it’s quite long?

THE COURT: No. That’s all right.

A: I have a Bachelor’s Degree majoring in mathematics and psychology from the University of California in Los Angeles, and I received that in 1966. I have a Master’s Degree in Psychology from Stanford University, which I received in 1967, and I have a Ph.D. in Psychology from Stanford University, which I received in 1970.

Q: And while you were in school, did you receive any honors or academic or professional honors?

A: Yes. I was elected to Phi Beta Kappa in 1965. I was also elected to Pi Mu Epsilon, which is the national mathematics honorary and to Mortar Board, which is the national senior women’s honorary. During my graduate training, I was supported all four years by fellowships from the Office of Education Traineeship, and the National Institute of Mental Health.

Q: All right. Have you published scholarly works in the area of psychology?

A: Yes. I have published approximately 45 scholarly articles, and I have completed two books.

Q: This is rather lengthy. Are those set out in your curriculum, or in your– how do you say that? Vita?

A: Vita.

Q: Could we just go through and pick out some that are maybe relevant to perception and memory and eyewitnesses?

A: Okay. Whereas most of these articles deal with the general problem in human memory and basic processes that are involved in the organization and retrieval of information, there are a few recent articles that deal specifically with the problem of eyewitness testimony, and I will point out some of those.

Q: What year are we talking about now, Dr. Loftus?

A: I was going to direct your attention to an article published in 1974. This was specifically on eyewitness testimony. The article I am referring to is Loftus and Palmer, “Reconstruction of Automobile Destruction: An Example of the Interaction Between Language and Memory.” It was published in the Journal of Verbal Learning and Verbal Behavior in 1974.

Q: From the vita. Have you published other articles in that area?

A: Yes. Well, later that year I published “Reconstructing Memory: The Incredible Eyewitness,” in Psychology Today. That was in the December, 1974, issue.

Q: Was that printed in the other journals an in the popular press?

A: Yes, that’s been reprinted probably six or seven times in newspapers and other journals and publications. In ’75, I published an article called “Leading Questions and Eyewitness Report,” in Cognitive Psychology. I published an article with Guito Zanni, called “Eyewitness Testimony: The Influence of the Wording of a Question,” and that appeared in the Bulletin of the Psychonomic Society.

I published an article with Diane Altman and Robert Gaballe entitled “Effects of Questioning upon a Witness’ Later Recollection,” and that appeared in the Journal of Police Science and Administration.

Q: Did that article deal with suggestions on how investigators should approach an eyewitness to avoid problems and things like that?

A: Well, that article described an experiment in which the subject viewed a very complex disruption, a classroom disruption, and afterwards they were interrogated in a manner which was aggressive, or in a manner which was very passive. And by “aggressive,” or in a manner which was very passive. And by “aggressive,” I mean the wording of the question suggested that the people who perpetrated the disruption were militant and were violent and so on. One week later, the subjects recall the incident as being more violent, more noisy and so on–

MR. YOCOM: I object to this as the contents of her article, I think it is going to the basis of the foundation at this time rather than the results of those tests.

THE COURT: Sustained at this point.

Q: Have you conducted experiments of your own in the area of Perception and Memory as they might relate to eyewitness memory?

A: Yes, I have.

Q: And could you state just generally some of the general conclusions that you have drawn from this experimentation and research in regard to the way Perception and Memory function in recalling events?

MR. YOCOM: I object at this time as to her conclusions, no foundation.

THE COURT: I need some more foundation on that. The objection is sustained.

MR. O’CONNELL: What I intended to do, Your Honor, so we’re clear, I am going from the general to the specific just because it makes more sense that way. Now, if the Court wishes, I can go from the specific experiments back up to the general, but I think it’s clearer if I work from the general back to the specific.

THE COURT: I’m not so much concerned in which direction you head, so long as there is sufficient foundation for whichever direction.

Q: Dr. Loftus, is there a generally accepted theory of how Perception and Memory of complex events functions?

A: Yes. It is general accepted that human memory does not work like a tape recorder, especially like a video tape recorder. We don’t just take information and have it reside in our memories in the form in which it was originally acquired. Rather, what happens when we experience something, when we witness something, s that information from lots of different sources is being integrated, and so the way we comprehend the experience depends, first of all, on our prior biases, but more important, on things that happened subsequent to the experience. And so subsequent experiences can also become integrated into a memory for some event and actually cause an alteration of it. The important point is that we are passive receptors of information, but we are constantly operating on that information.

Q: What does the term “unconscious transference” mean?

A: That’s a term that has been used to refer to the mistaken recollection of confusion of a person seen in one situation as the person who has been seen in a different situation. And the classic example of unconscious transference is found in Wall’s book, “Eyewitness Identification in Criminal Cases.” There Wall describes the case in which a clerk in a train station was held up, I believe at gun point. Later on, the clerk identifies a sailor in a lineup. The clerk said that sailor was the person who had committed the armed robbery. It runs out that the sailor had an ironclad alibi, but had purchased tickets from the clerk on three prior occasions, so when the clerk went to the lineup, the face did, indeed, look familiar; it was familiar, but the clerk confused that familiarity and recalled the face as being the face of the robber rather than recalling it as being the face of the person who purchased the tickets. That is unconscious transference.

Q: Could it be applied to other than situations where the person had been actually seen? For instance, could it be applied to objects? I thought I saw an object here when I really saw it there?

A: It could. The term itself has been used usually to refer to faces, live faces or faces in photographs, but it could be extended, the concept could be extended to refer to objects as well.

Q: Have you done any experiments with this phenomenon with the use of photos?

A: Yes. If I could refer to some notes that I have made.

MR. YOCOM: If we are going to get into the areas of experiments, with regard to her experiments? I’m going to object to it if we are, because I think we’ve still not laid a proper foundation for this testimony.

MR. O’CONNELL: Well, she’s going to describe an experiment and the conclusions that come from the experiment, and I felt the description of the experiment would lay the foundation for the conclusions.

THE COURT: Well, I suppose we might get conclusions to a point where she’s describing an experiment.

MR. YOCOM: Then it would be immaterial as far as the facts in this case are concerned. She’s talking about something entirely different.

MR. O’CONNELL: I’m going to lay the foundation.

THE COURT: But a lot of foundation is immaterial sometimes. The objections is overruled.

DR. LOFTUS: I want to describe some of the experiments that have been done that lead one to know that the phenomenon of unconscious transference occurs in human beings. And first of all, I will tell you about an experiment by Robert Buckhout, then I will describe an experiment that was done in my laboratory.

MR. YOCOM: Are we going to proceed on a question and answer basis, or a lecture method?

Q: Would you describe the experiment of Mr. Buckhout that led up to your experiments?

A: All right. What Professor Buckhout did was to show approximately 141 people an event; in fact, they witnessed a staged assault in which a student assaulted a professor.

Q: Did the people witnessing it know it was staged?

A: I think eventually they were told, immediately afterwards, but not at the time. It was a spontaneous event. They were asked some questions about the event immediately afterwards, but for purposes of the identification part, seven weeks later, these people were returned to the experimenter, and they were shown six photographs. One of the photographs depicted the person who committed the assault. It turns out that in this experiment only 40% correctly, after nine weeks, could correctly identify the person who committed the assault. That means 60% incorrectly, either were incorrect or failed to make an identification. Now, what was of interest to the concept of unconscious transference is the point: one of these people was actually the criminal, that is, the person who committed the assault. Another person was an innocent bystander, somebody who had been near the scene of the crime, but had not committed the crime. And when people made a mistake, that is, the 60 percent of the people who either were incorrect or failed to make an identification, they tended to choose the innocent bystander. And this experiment has been interpreted as being a possible example of unconscious transference. This person looked familiar, and the familiarity was mistaken related back to the act of the criminal rather than to where it properly belonged, that is, an innocent bystander. In my experiment, the experiment in my laboratory– and this is an experiment that will be published later this year in the Law and Psychology Review– I showed people photographs, and at the same time they heard a story, and they saw six photographs, one at a time, while they heard a story of a crime being committed. There was 30 subjects in my experiment and three days later–

Q: Now, the people whose photographs were being shown, were each of their parts in this–

A: –story spelled out, yes. This was a story which described an incident. All of these people except for the fourth person were innocent people, and the fourth one was the person who committed an assault. So when this picture came on, my subjects learned that this was the person who was essentially the criminal. In this experiment, three days later, the subjects came back. They didn’t even know that they were going to be asked any more questions, they thought they were coming back to pick up their money for participation. And we presented them with some photographs and asked them to select the criminal. We presented five photographs and asked them to select the criminal.

Now, of interest for this particular discussion is what happened to 25 of these subjects. With the 25 subjects, we presented four people who had never been seen before, and one innocent bystander, that is, one of these other people. So the correct answer is, “The criminal is not here.” What my subjects did is as follows: 60 percent of the subjects chose the innocent bystander; 16 percent of the subjects chose another incorrect person — actually, all of these people were incorrect; and only 24 percent refused to make a choice, that is, said that the criminal was not in the set.

Q: Was it suggested to them in some way that the person should be in the set?

A: No. They were just given the instructions they should look at the photographs and pick out the person who committed the crime if he was there. So of these 60% who picked the bystander, this is a percentage that is significantly greater than you would expect by chance. What is happening in this case is the witness subject is looking at the face, it indeed looks familiar, and the familiarity is mistakenly related back to the act of the criminal instead of related back to the act of one of the innocent bystanders. So this experiment indicates that the phenomenon of unconscious transference can be demonstrated in the laboratory, and is a very real phenomenon.

Q: Now, did you review the statement of Carol DaRonch and the testimony of Carol DaRonch at the preliminary hearing that I submitted to you?

A: Yes.

Q: In those two documents, and adding to that something you saw this morning, can you describe for the court an instance of unconscious transference in a collateral manner?

A: Yes.

MR. YOCOM: Objections, Your Honor. Foundation.

THE COURT: Sustained.

DR. LOFTUS: If we extend–

MR. O’CONNELL: Just a minute. The objection was sustained. But foundation objections are just awfully broad, Your Honor.

THE COURT: Yes, I appreciate that, but you need to state enough by way of foundation so that Mr. Yocom knows what the subject matter is that you are going to, so that he can make a proper objection if one is appropriate.

MR. O’CONNELL: Well, I think my question went to whether or not she saw an example, let’s put it this way, a clear example of unconscious transference taking place in the testimony of Carol DaRonch other than the identification of Theodore Bundy. And I think that she said yes. I think that’s all the foundation I need. Now, is the objection relevance?

THE COURT: Is this based upon something she observed in the transcript of the preliminary hearing?

MR. O’CONNELL: Yes.

THE COURT: Then state the point so Mr. Yocom can follow.

MR. O’CONNELL: That was going to be my next question. The first question is, did she observe something and the answer is what?

Q: Yes. And do you have those documents before you?

A: Well, I know what is in them.

Q: All right. Well, can you describe what is in the documents that you are relying on before you get to your conclusion?

A: Yes. As far as I have learned from these documents, on November 8th when the incident occurred, the witness stated that she had been shown a badge that was gold. At the same time a police officer showed her a police badge that was silver, gold and blue, and said did it look like this and she said no, or something like that. Approximately a year later in the preliminary hearings he was asked to describe the badge, and she described it as being silver and blue or gold and blue. It appears to me as if what she may have done is integrate the image of the badge that she had seen that had been shown to her by the police officer into her memory, and now it is as if to her– it appears to her as if she had seen this police badge instead of the badge she originally described.

MR. YOCOM: May I voir dire on that point, Your Honor?

THE COURT: Yes. So long as it’s voir dire.

VOIR DIRE EXAMINATION BY MR. YOCOM:

Q: Doctor, would the fact she observed maybe hundreds of badges that same evening affect your opinion?

A: I would think that it would lead to a great amount of confusion as to– it could lead to a great amount of confusion as to which badge she had seen in the incident.

Q: Now, she describes it as silver and gold, and later says there’s some blue in it; maybe she had seen 10, 40, or 50 such badges in a book of badges that she observed that evening.

MR. O’CONNELL: This sounds like cross-examination to me, Your Honor.

THE COURT: I’m not so sure that it is no, Mr. Yocom.

MR. YOCOM: Well, she states she has seen an example of what she calls unconscious transference. I don’t think she has all the facts to state that opinion on.

THE COURT: That becomes cross-examination.

DIRECT EXAMINATION (RESUMED) BY MR. O’CONNELL:

Q: Has there been considerable literature and experimentation on the subject of the retention interval in memory?

A: Yes, there has.

Q: Could you describe, say, some of the history of what has been turned up in the literature on that subject? What do we mean when we express “retention interval?”

A: Retention interval refers to a period of time between an incident and a test for one’s recollection of that incident. So in this case, applying it to this situation, we’d be talking about the period of time between a crime and subsequent identification of the perpetrator of the crime.

Q: All right. And could you describe some of the experiments done by other people that appear in the literature that are significant?

A: It turns out that we presently know more about the retention interval and its effect on memory than almost every other factor that you can imagine. It has been studied at least since 1885 when Ebbinkhaus first began studying this phenomenon with nonsense syllables. What he did was learn these syllables and relearn them at some later time, and he looked at how much more quickly he could learn them the second time, and this was an indication of how much was retained. So if you plot the percent that was retained, using this measure of how quickly he could learn it the second time, as a function of how long it had been since the last time he had studied the particular item in terms of days– and I believe he studies both sorts of retention intervals, in at least one study one to six days– and the function looks something like this: (drawing). This is called the forgetting curve. Basically, what this curve indicates is that when we learn some new material, most of the forgetting occurs immediately after the learning of it, and then the forgetting is more and more gradual. So this is the curve of forgetting, as first postulated by Ebbinkhaus in 1885. Since that time, the same sort of experiment had been done over and over and over with different subjects and different laboratories with different kinds o materials. And just to give you another example of the kinds of materials that have been used, another psychologist named Dallenbach used very complex pictures. He showed his subjects fairly complicated pictures, then asked them a series of 60 questions, and for different groups of subjects the questions were asked after a different interval of time. So some people were asked the question immediately, that is, after an interval of zero days; others were asked the questions five days later; still others were asked the questions 15 days later; and then, finally, one group of subjects had the questions asked 45 days later. I want to show you what happens to a number of wrong answers in Dallenbach’s study. On the average, after the questions are asked immediately, he found there were an average of 8 wrong answers. After 5 days, there were 10 wrong answers, then 12 wrong answers, and after 45 days, and average of 13 wrong answers. The subjects also had to indicate which of their answers they were absolutely positive of, that is, which ones they would be willing to answer to under oath. So let’s look at what happens to the number of wrong answers that would be sworn to. After zero days, it is 3, after 5 days, it’s 4, on the average. After 15 days, it’s 6. Then finally, after 45 days, his male subjects will swear to an average of 7 wrong answers. Much more recently, using pictures, a psychologist named Shepard showed clerical workers a large number of pictures and tested their recollection after 2 hours, 3 days, one week, or four months. What he did was to present two pictures — for his tests, that is — two pictures, and the subject had to choose which one he had seen before, whereas immediately afterwards, that is, after say a two-hour interval, people were almost perfect, that is, they were a hundred percent correct. By four months, they couldn’t recognize the pictures very well. They performed at a level of 57 percent correct, where 50 percent is just chance, just guessing.

Q: They were picking between just two pictures?

A: Picking between two pictures and identifying the one they had seen before.

Well, these are just some examples of psychologists and experimental research that has dealt with the phenomenon of the retention interval. And the conclusion from this research is that, well, really, the forgetting curve seems to apply; we forget things very quickly at first, then the forgetting is more gradual. But the same phenomenon is observed in a variety of situations.

**Several pages are missing from the transcript here. Transcript resumes during cross examination.**

CROSS-EXAMINATION BY MR. YOCOM:

Q: Now, you gave an example of feeding people false information in your testing with regard to questions given to an individual on their recall — and forgive me if I am getting a little mixed up– but this premise, or this test is based upon the fact that false information is given initially, is that right?

A: The first thing that happens is people witness some of that, namely, a filmed event. The second thing that happens is that they are asked a series of questions, and one of the questions contains a piece of false information. It’s a vehicle for introducing the information.

Q: But this test is again a laboratory type situation where these people are not the actual victims of crime?

A: That is true.

Q: Are they paid people, also, to participate?

A: We try to get our subjects from the introductory psychology courses where they are required to participate in five experiments, for example, in order to get credit for the course. sometimes we pay them.

Q: If, for instance, in a photo display to a witness in a crime, or even one of your students, there was a stack of a great number of pictures and they were trying to recall, say for instance in Buckhout’s test, the assailant, and the experimenter merely said, “Look through these and see if you see him,” would that differ a great deal from situations where the experimenter might say, the suspect or the criminal or the man is in these photographs, will you pick him out?

A: I would think in the latter case the witness would be much more likely to try to make an identification, and under more and more pressure to do so.

Q: Than in the first situation?

A: Than in the first situation.

Q: When you say merely, look through these and see if you see him?

A: Yes.

MR. YOCOM: No further questions.

MR. O’CONNELL: I believe I have no further questions of this witness, Your Honor, unless Your Honor has something he would like to go into.

EXAMINATION BY THE COURT:

Q: Dr. Loftus, do you have any data upon which to form a conclusion as to whether the fact a person is a victim in an actual situation would affect in any way the factors that you have heretofore set forth?

A: Yes.

Q: And they are eyewitness identifications?

A: I think that the role of stress becomes very important here, and again, there are some assumptions that I am making. If you assume that the victim of a crime is experiencing extreme stress and arousal and fear, or some combination of these, then your best guess– our best guesses as experimental psychologists, I am speaking now– is that the performance in this case we are talking about, memory performance is going to be lessened, be less good. So if you assume, then, in a crime situation the victim is experiencing something, some extreme arousal, that that memory arousal is not going to be as good as if the arousal were more moderate. And I can’t say, I have no reason to suspect, that the factors I describe would operate any differently, but the overall level of memory performance would be worsened with a person who was experiencing extreme stress.

Q: What you are really saying, then, is that the more stressful crime is, by virtue of its nature, as it affects the victim, the less likely the victim is of ever going to be able to identify the perpetrator of the crime?

A: That’s right. Now, people vary according to what causes them extreme stress.

Q: That would be a factor that you would have to take into account in an individual situation?

A: Yes. I’m now making the assumption that most people who are victims of fairly violent crime are experiencing extreme stress, and therefore, their performance is going to suffer. But the other factors that I have described should still operate in the same way.

Q: So if there were a period of time prior to the creation of a stressful situation where the victim was not acting under stress, would you think that any identification based on recollection during that period of time would be more reliable than where the entire event was stressful?

A: In fact, the relationship between stress and performance is called curvilinear or non-monotonic. If you look at– I will just draw it here– if you look at this, this is a law which–

Q: When you say “here,” you are referencing to Exhibit 61?

A: Yes, in this box. If I were to plot performance– and here I am talking about memory or perception or problems, all kinds of performances as a function of the person’s stress or arousal– the relationship looks something like this. That is under very, very mild stress, or mild arousal. This is also arousal. When you are waking up, your performance is not going to be very good. also, at extreme levels of stress, your performance is not going to be very good. I have assumed that victims of crime, particularly violent crime, are operating in this area, that is, they are experiencing extreme stress, and so for them, the performance is going to suffer. Now, if there is some other periods of time in which you are measuring performance as a function of stress, and the stress is less extreme, I need to know something about where in the range it was; was the person moderately aroused and stressed? In which case, you would expect fairly optimal performance. Or was there very low stress? In which case, once again, you would start to express a decremented performance.

Q: I see what you are talking about. Let’s assume, then, that you have a time continuum as represented by a straight line. At one point in time there is no stress whatsoever. As that continuum line proceeds, the stress begins to build, that is, assume that the victim begins to suspect that the victim might be a victim. and then at a point further down in the continuum line, the victim becomes assured of the fact that there is a real problem.

Now, at all times on that line of continuum, she– or he, or the victim– is able to observe the perpetrator of the offense. How would stress, in your judgment, affect eyewitness identification under those circumstances? That is, would the victim be more likely to be able to identify the perpetrator under those circumstances than in a situation, where there was a sudden violent event like a breaking-in through a door or a window?

A: Well, although that particular experiment hasn’t been done, my guess is that the sudden event would produce a less accurate identification, because the whole experience would be while the victim was in an extreme state of stress.

THE COURT: All right. Thank you. Any further questions based on the Court’s examination?

MR. YOCOM: Just one, Your Honor.

FURTHER CROSS-EXAMINATION BY MR. YOCOM:

Q: The thing that the Court covered, I believe they call “exposure time;” is it not?

A: Yes.

Q: So the longer the exposure time that the victim would have with the suspect would increase the person’s ability to later identify him?

A: In fact, exposure time is another factor in eyewitness testimony, and the longer you have to look at something, the more likely you are to be able to make a correct identification.

MR. YOCOM: No further questions.

REDIRECT EXAMINATION BY MR. O’CONNELL:

Q: A question I expected somebody to ask you, Dr. Loftus; people vary in their ability to remember things? Is there a rather wide–

MR. YOCOM: I stipulate to that, counsel.

THE COURT: I think she testified to that generally, but go ahead.

A: Yes, there are individual differences.

MR. O’CONNELL: That’s all.

MR. YOCOM: I have nothing further.

THE COURT: May this witness be excused?

MR. O’CONNELL: Yes. May she remain in the courtroom, Your Honor? She’s not going to testify.

MR. YOCOM: I have no objection.

THE COURT: Thank you, Dr. Loftus.

---

Special thanks to Erin Banks for her transcription assistance.