Interview Highlights

This is a condensed, lightly edited transcript of an audio interview. The full audio is available and highly recommended. The interviewee may post clarifications in the comments.

Adrian Bye: Today I’m here with Bill Ickes, who is a university researcher in Texas, and has written a book, Everyday Mind Reading: Understanding What Other People Think and Feel. I don’t tend to interview book authors here, but this book was just so incredibly phenomenal and can lead to such a profound understanding of empathy that I just had to get Bill here. Bill, thanks for joining us. Can you tell us about who you are and where you come from?

Bill Ickes: I didn’t set out with the goal of becoming a research psychologist. My first psych course suggested that maybe I needed to make a course correction, but I stayed an English major, and added psychology as a minor. Eventually, I realized that I did have strong research interests, and a lot of the ideas from the study of English Literature could be incorporated into psychology. In particular, I thought that it might be possible to treat the interactions of people in a laboratory setting as little mini dramas. Basically the concept is to put two people together, then eavesdrop on their conversations, and let their personalities write the script for them.

We spent about 10 years, after I got my PhD, studying effects of people’s personality traits on their initial interactions in the laboratory. Towards the end of that time, I began to realize we could extract a lot more information from these initial interactions. We had been looking at how much people look at each other, smile at each other, how close they sit on the couch, and so on. There was a whole other dimension of the interactions we weren’t tapping, and that was what was going on in the privacy of their respective minds.

That led us to videotape the interactions, making two copies instead of one. Then as soon as the interactions were over, we would tell them that there was more going on than they realized. We wanted to find out what they were thinking and feeling throughout that interaction. So, they are put in separate rooms, and they can stop the tape at all the points where they distinctly had a thought or feeling. They write down the content of each thought or feeling, using special forms that we provide them.

After a while, we were doing studies focused on the content of their thoughts and feelings. How the kinds of thoughts and feelings they reported were related to their personality characteristics, to their perceptions of each other as interaction partners, and so on. It wasn’t too long before I realized why don’t we take them through the tape of the interaction again and this time pause the tape at each of the points where their partner recorded thoughts and feelings, and tell them now it’s your job to make your best guess about what your partner was thinking and feeling at each of these points. In fact, we were asking them to read their interaction partner’s mind at each of the points where the partner had a thought or feeling.

Adrian Bye: What you were evaluating was the quality of the empathy?

Bill Ickes: The term we use, which is an inversion of an earlier term used by the clinical psychologist, Carl Rogers, is empathic accuracy.

After we have all their thoughts and feelings, plus their differences about the content of their partner’s thoughts and feelings, we can have our lab find accuracy points based on how closely the inferred thought or feeling matches the content of the actual thought or feeling. When you add up all those points, and divide by the number of inferences made, you get a measure of empathic accuracy or successful mind reading. The measure we use, varies on a scale from 0 to 100, where 0 is no idea what the other person was thinking or feeling, to a theoretical limit of 100, perfect empathic accuracy, which we never see in our research.

Adrian Bye: What was it you saw? 30-60%?

Bill Ickes: That’s right. If people get a general sense of the theme of the interaction and they apply other knowledge like general stereotypes that might apply to their partner, they can be somewhat accurate for those reasons, but not very. Accuracy at a base line only averages about 5%.

If you study several interactions between two strangers, their average empathic accuracy for that initial interaction is about 20%. If you put two close friends together, it goes up to 30%, which is a substantial improvement, but still far from that theoretical ideal of 100% accuracy. Married couples are comparable to best friends, and can be a little bit higher.

There seems to be an implicit upper limit on accuracy scores, around 60%. We just don’t see people scoring in the 80s, 90s, or even the high 70s.

Adrian Bye: Why do you think that is?

Bill Ickes: I’ve been giving it a lot of thought. There are a couple of reasons. One is because this kind of empathic accuracy depends on complex psychological inferences based on imperfect queues. People are not always telling you everything that is going on in their minds when you are talking to them, so it’s reasonable to expect less than 100% accuracy for that reason. But, I think that the more interesting and profound reason is that evolution doesn’t want us to be perfect mind readers.

Adrian Bye: You talk about that in your book, where you say we don’t want 100% accuracy because there needs to be things hidden to make relationships last.

Bill Ickes: I think the primary reason why evolution might have calibrated it much lower than 100% is that the prime directive in nature, according to Darwin’s theory of evolution, is making sure that your genes go on to the next generation. If you were 100% accurate in inferring everyone’s thoughts and feelings, and you were perfectly aware of their needs, you might try to divert your resources to them instead of to your children, and put your own genes at risk. Either your own direct genes, by giving away things that preserve your life to others, or your children’s genes, which, after all, are yours.

I don’t know how one would go about testing this evolutionary idea that nature has calibrated human empathy at this general level, but I think it’s a very plausible hypothesis. I encourage people to think about it and try to figure out ways to test it.

We obviously need empathy as a skill. People have to be good enough at reading other people’s thoughts and feelings that other people can’t lie to them repeatedly and take advantage of them. We have to be able to see bad things coming in our relationships with others. But, perfect empathy could be more of a curse than a blessing.

Adrian Bye: Did you do much research with people on the autism spectrum?

Bill Ickes: I have colleagues who have. I spent some time in Belgium, in 2005. I was at Ghent University, and there was a very active group lead by Herbert Roeyers there. He used our empathic accuracy methods to study the ability of high functioning Asperger’s patients to infer other people’s thoughts and feelings. In one of their studies, there were pairs of individuals, one with an Asperger’s diagnosis, and the other without, whose IQ scores were approximately the same, and were also the same age and gender. The research showed that the Asperger’s individuals did have more difficulty inferring thoughts and feelings when they viewed videotapes of interactions between strangers compared to the normally developing controls.

Other research showed that this difference in performance is reduced in a situation where the interactions are highly structured. If the interactions become more structured and predictable, people with Asperger’s have less difficulty making accurate guesses about what the people on the videotape are thinking and feeling. Where they have the most difficulty is where things are completely unstructured.

Adrian Bye: What do you mean by structured or unstructured?

Bill Ickes: There are different ways to structure the interaction. One way would be to turn it into an interview with prepared questions. Another way to structure an interaction is to organize it in terms of a game with rules. An unstructured conversation is when people just freely talk about whatever comes to mind.

Adrian Bye: You said you found average empathic accuracy is 20%. Do you have a percentage of what they found in terms of Asperger’s?

Bill Ickes: They do in their research, and the percentage won’t vary greatly. People with Asperger’s, as I recall, didn’t fall much short, but it was enough that the difference was statistically significant. They are doing quite well in comparison to small children who aren’t very good at mind reading, but they aren’t doing as good as the average person.

If you look at people who have more severe cases of Asperger’s, or are in the range where a diagnosis of autism is made, then the deficit is more profound. Autism researchers suggest that the victims of extreme autism are effectively mind blind. They are incapable of inferring the specific content of other people’s thoughts and feelings. In fact, in some cases, clinicians have argued that their patients are unaware even that people have thoughts and feelings.

Adrian Bye: Another thing I found fascinating was your research into the empathy of abusive men.

Bill Ickes: The view we discovered was prevalent among maritally abusive men was that all women are constantly having very negative, critical, rejecting thoughts and feelings about men. Not only does this bias interfere with the men’s ability to accurately infer what a woman is thinking or feeling, but it seems to justify, in the minds of abusive men, that it’s ok to show contempt for the woman, to physically abuse her, because she deserves it.

We got a wide sample of married men who ranged from kind, loving husbands all the way to men who were both physically and emotionally abusive. We wanted to have non-abusive men in the sample to use as a control condition to compare men who don’t abuse their wives with men who do.

When both kinds of men view videotapes in which a woman is describing to a male therapist her problems in her marriage, at times when the woman gets emotional and distraught, the non-abusive men report that they feel sympathy for those points. To the degree that they report feeling sympathetic, their empathic accuracy is higher.

The abusive men act very differently. Those same points where the woman is highly emotional and distraught, the abusive men report they are feeling contempt for her and they are not nearly as accurate in inferring her thoughts and feelings.

Adrian Bye: You found that you can motivate people to have higher levels of empathic accuracy.

Bill Ickes: There is probably a limit to how much they will improve. In general, the research shows that women are easily motivated and try to do their best in inferring the specific content of other people’s thoughts and feelings. This is very consistent with the social stereotypes that women are supposed to be empathic. If you remind women of this stereotype, they try to put more attention and effort into it, and they do better.

Men, to motivate them, it took a different motivator initially. That was to simply pay them. They were given some practice trials, told to be as empathically accurate as possible, and to the extent that they were, they got paid actual money. If you really do your best, we’ll pay you more. It was sufficient to motivate these guys so that they could close any motivation gap between themselves and the women, and they wound up doing just as well.

You can extrapolate this to all the marriages and cohabitation relationships that the woman is thinking he is not very good at being sensitive to my thoughts and feelings and she’s wondering is this something he’s unable to do? No. The answer seems to be he’s perfectly capable of doing it. It just takes more to motivate him.

Adrian Bye: Have you tried testing psychopaths or sociopaths in terms of their empathy accuracy?

Bill Ickes: I’ve been at conferences where people have discussed this and agree that it’s a good idea to do. Interestingly, people disagree what the prediction should be. Some people argue that psychopaths should have very good empathic accuracy, because they are often exceptionally good at manipulating people. They seem to have real insight into what other people are thinking and feeling, so they can push all the right buttons to get people to do what they want. By that perspective, the problem with psychopaths is not a lack of empathic accuracy, it’s a lack of any sympathy for other people’s thoughts and feelings. Other people argue that maybe psychopaths do have a deficit in empathic accuracy. We’ll need to see the data to know what the answer is.

Adrian Bye: Could this come down to a gap in mirror neurons?

Bill Ickes: This is a special case by which the distinction between cognitive empathy and emotional empathy may turn out to be quite important. It may be different mirror neurons that are implicated in cognitive versus emotional empathy. Perhaps psychopaths have only a deficiency in regards to the ones that mediate emotional empathy. That is, they can infer other people’s mental states quite well, but they don’t have any sort of an emotional resonance or sympathy with the other person.

Adrian Bye: Is there anything that you’d like to talk about that we haven’t covered.

Bill Ickes: The potential application. One agenda that remains to be addressed is can you use the techniques that we’ve developed to identify people who are more empathically accurate than average, and then use it as a selection tool for occupations where that is an important skill?

Adrian Bye: If someone wanted to increase the empathic accuracy of their team, is this something they can reach out to you for?

Bill Ickes: They could contact me. There is information in the book, Everyday Mind Reading, about our research on feedback and its effect on empathic accuracy. We found that, in the single session in which feedback was provided, empathic accuracy for a particular target could be boosted on average of about 10%. 10% may not seem like a lot, but it can be a real advantage if you learn to do that consistently.