Speaking of Psychology: Expressive writing can help your mental health, with James Pennebaker, PhD

Writing can be a powerful tool to help people work through challenges in their lives and improve their mental health. James Pennebaker, PhD, of the University of Texas at Austin, talks about why expressive writing can be good for mental health and how to try it. He also discusses his research on language use, and how analyzing the words that people use in their daily lives can offer insights into their emotions, motivations, and personality.

James Pennebaker, PhD, is professor emeritus of psychology at the University of Texas at Austin. He is continuing to explore natural language use, group dynamics, and personality in both laboratory and real-world settings. His earlier work on expressive writing found that physical health and work performance can improve by simple writing and/or talking exercises. His cross-disciplinary research is related to linguistics, clinical and cognitive psychology, communications, medicine, and computer science. Author or editor of 12 books and over 300 articles, Pennebaker has received numerous research and teaching awards and honors.

Kim Mills: Have you ever kept a diary? For many people, journaling is a lifetime habit. Others may turn to it in moments of stress or during major life transitions. But in fact, psychologists’ research suggests that we all might want to consider occasionally writing about our lives, even those of us who don’t consider ourselves writers. That’s because researchers have found that expressive writing can improve people’s mental and even their physical health and can help us work through challenges in our lives. Today we’re going to talk about why that is. We’re also going to talk about the power of words more generally and about how the words we use can offer insights into our emotions, motivations, and personality.

So why does writing about our lives improve our mental health? Do you have to write every day to see these effects? Does it matter how a person writes? Are there guidelines to follow if you want to try it? And how does language connect to emotions, motivations, and personality? What do the words we use tell us about ourselves and each other? Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association that examines the links between psychological science and everyday life. I’m Kim Mills.

My guest today is Dr. James Pennebaker, a professor emeritus of psychology at the University of Texas at Austin. He conducted pioneering research on expressive writing in the 1980s and ’90s. Since then, he’s focused on natural language use and how language, personality, and social behavior interact. Dr. Pennebaker is the author or editor of more than 300 articles and 12 books, including Expressive Writing: Words that Heal and The Secret Life of Pronouns: What Our Words Say About Us. He has received numerous awards for teaching and research, including the 2016 APA Award for Distinguished Applications of Psychology. 

Dr. Pennebaker, thanks for joining me.

James Pennebaker, PhD: It’s great to be here. Thank you.

Mills: Let’s start with your research on expressive writing. You’re a social psychologist by training, not a clinical psychologist. So how did you get started studying what eventually became recognized as this powerful therapeutic technique?

Pennebaker: Well, this is actually the story of my life. I’ve never really planned to go in any particular direction, just letting interesting things fall in my lap. Early in my career, I was interested in mind-body issues. Why do people get sick? How do you know if you’re sick? And I was working on what became an early book called The Psychology of Physical Symptoms. And I was thinking I should do a questionnaire just to get a sense of what kind of people, what kind of personalities, what kind of life experiences might be related to people reporting physical symptoms. And I came up with this giant questionnaire working with students, and we put on questions that—things that sounded interesting. They didn’t have any theoretical basis. We asked about people’s diet, how they got along with their parents.

And one of the students I was working with says, how about this one: Prior to the age of 17, did you ever have a traumatic sexual experience? And I thought, yeah, that’s an interesting question. That’s certainly nothing—nobody ever asked that question back in the early '80s. That one question ended up being one of the most powerful predictors of physical symptoms in health. And I later did this with a national sample. And again, people who endorsed that item were more likely to have been hospitalized for any cause in the previous years. It was related to colds, flus, cancer, high blood pressure, you name it.

And I became fascinated, what is it about a traumatic sexual experience? And as I started to do more studies, I discovered it wasn’t just a traumatic sexual experience. It was having any kind of traumatic experience that people kept secret. Now, we had known forever that traumas were bad for physical health. That wasn’t news. But having a trauma of any kind and keeping it secret really increased the toxicity of that trauma. In other words, secrets in and of themselves were toxic. And this made me wonder: If secrets are so bad, what if we brought people in the laboratory and had them talk about them or—that turned out to be way too complex. How about we just had them write about it? And that was kind of the birth of expressive writing.

Mills: So what is it about expressive writing that helps people? And how is it different from other types of writing? For example, just writing letters, essays, or even lists that we write to keep ourselves organized?

Pennebaker: We’ll get into those questions. Those questions are actually much deeper than you might expect. The quick answer is there are—many of these approaches probably are just about as effective. What I did with that first study, I had no idea if should I have people write once or how many times should I have them write. And the first studies were all done with college students. I decided it would make sense to have people write maybe three or four times. I needed to do an experiment in an efficient way. So I would have people, they were randomly assigned by a flip of the coin to either write about a traumatic experience for 4 days or about superficial topics for 4 days. And I needed to have them write it in a laboratory at the university. And this study was done at Southern Methodist University in Dallas, and I needed enough rooms and I wanted to line everybody up, essentially, and essentially have somebody come in at 4 in the afternoon, 4:10, 4:20, et cetera, and just line them up and run all 50 or so people in an afternoon.

And to do that, I needed a lot of rooms and I was able to get enough rooms to run for 4 hours, and I had a lot of people helping me. And we could only get those rooms for a week, which was really 4 days, Monday through Thursday. And just doing the arithmetic that meant that everyone could write—in that first study, I believe it was 15 minutes, and they wrote for 4 days all we could get the times, and it had to be after all the classes had finished. And so all these practical considerations, and that first study was powerful because those people who were asked to write about upsetting experiences ended up going to the student health center at about half the rates of people in our controls. So there was no great theory of how much they should write. There was no great foresight in terms of what they should write about.

And that was the other issue. What should I have them write about? Well, I didn’t know. So I thought I’ll have them write about a really traumatic experience that they’d had. And actually, here were the first instructions that I gave: For the next 4 days, what I’d like to have you do is to really let go and explore your deepest thoughts and feelings about the most upsetting or traumatic experience of your entire life. I really want you to let go. And as you write, I want you to write continuously. Don’t worry about spelling or grammar or anything like that. And in your writing, you might write about how it relates to other issues in your life, maybe related to your childhood, maybe related to school, or how it could be related to your friendships or who you’d like to be in the future or who you’ve been in the past or who you are now. And you can write about the same traumatic experience all 4 days, or you could write something different each day. That’s entirely up to you. And in fact, many people have never had a trauma, but all of us have had major stressors or conflicts, and you can write about those as well. And ideally, I’d like to have you focus on something that you haven’t talked to other people about much. So for the next 4 days, I really want you to really let go and write.

So that was the instruction. And then people who were assigned to the control condition were asked to write about superficial topics. Initially, we had them write about things like the shoes they were wearing or their dorm room or something like that. But then later I became fascinated with time management. I’ve always been dismissive of time management. To me, I think that’s kind of training for obsessive compulsive disorder. In any case, what I did was after the first couple of studies, I would, in the control condition, I’d ask people how they use their time. What did you do during the last 24 hours? What do you plan to do during the next 24 hours? What did you do last week, et cetera. And having them focus on details. And I’m pleased to say that never was associated with any improvements in health or anything else, but everybody thought it was, they thought that, well, it’s time management. It must be good. But the point with this was in those first studies, we were focusing on traumas.

And then I started having people write about different topics. One of my next studies was having people not writing about the most traumatic experience of their lives. But I got college freshmen who were just starting college, and I would have them write about their deepest thoughts and feelings about coming to college—with the same general instructions, tie this into other experiences, and we got the same effects.

And then we started doing more and more studies with people who got laid off from their jobs, or you name it, it seemed in those first few years, we started doing more and more studies, having people write about anything. And those first studies, we had them write for the first one, I think it was 4 days, 15 minutes a day. And the second study I think was 20 minutes a day for 4 days. And then the next study was 3 days. And in that study after that was 5 days. And sometimes it would be as little as 10 minutes, sometimes as long as 30 minutes. And they all got similar effects.

So that ended up being kind of the basis of this. And it’s ironic that there’ve now been over 2000 studies published on expressive writing. I’m astounded by this. I can’t even keep up with the literature. I was recently looking online with Google Scholar, how many papers have been published on expressive writing in the last year? And it’s about 200. It’s really quite surprising. And some of these studies work, some of them don’t work. I mean, it’s not a giant effect. But what’s been interesting is some people will have individuals write for 2 days, 3 days, 4 days, 5 days. Some will have them write once a week for 4 weeks. Sometimes they’ll have them write several times in 1 day. Sometimes they’ll have them write 20 minutes, 10 minutes. There’s even been a couple of studies where people write just 2 minutes each time. And again, they all seem to work.

Mills: Is there a greater effect for the deeper feelings? Is that what you’re seeing? So in addition to secret, it’s something that is really heartfelt and deep inside?

Pennebaker: I think the heartfelt deep inside is true with all of these. And even if you write for 2 minutes, what’s interesting is there’s a false sense that, oh, you’re only writing for 2 minutes, four times. That’s only 8 minutes. It can’t possibly work. Well, the reality is you’re writing for those 2 minutes, but you’re thinking about it. You leave the lab and you start pondering it and you might talk to somebody about it. You live with that experience over the course of the study. And so look, how most people are—and I’m one of these most peoples, I never was a dedicated writer, and I’m still not a dedicated writer. I write when I need to. And it used to be if I was upset about something, I had a fight with my wife or there’d been conflict at work or whatever, what I would do is I would not talk about it. I’d walk down the street and I’d start thinking about it, and then I said, oh, I should have said this. And then I was thinking about something else. And then I think, well, I should write about this. No, it’ll just make me think about it too much, which of course is stupid. I’m already thinking about it too much—

Mills: Right?

Pennebaker: And so getting people to actually sit down and confront it and to write it, you don’t have to write a lot, but you have to first of all just acknowledge it and put it into words. And that was really for me, the breakthrough.

Mills: Do people have to write on a regular basis to see a really significant effect? Do I have to write something every day to really experience the benefits?

Pennebaker: Oh my God, no. To give you an example, I write maybe two or three times a year when something miserable is going on. When I hear about journaling and writing every day, I just get nervous about it. I would hate that because I use writing when I’m dealing with something that is ugly, unpleasant, painful. And if life is going well, why in the world do I want to introspect that? I want to enjoy life when it’s here and when there’s bad things, writing for me helps me get through them. And so I view it almost as—I’ve often thought of it as an antibiotic. You have all these bad things going on and then you use this method to get past it, and then next time something bad happens, I’ll use writing again.

Mills: Is there a difference between writing something by hand or writing on a computer, or what about talking to a speech to text program? Any differences?

Pennebaker: There have been now a large number of studies comparing them. And I’d have to say that on average there’s no difference. If you stand back and you squint at the data, you can see a weak effect that writing by hand might work a little bit better, but my colleagues don’t really acknowledge the squint test, so it probably doesn’t matter. My recommendation is to do what you’re most comfortable with. And I’ve experimented with, and I’ve actually tested this with people in workshops, having people write, using finger writing—where you just with your finger write in the air about an upsetting experience.

I invented this years ago when my wife and I were on a vacation somewhere and we’d had a significant disagreement about something. I got up in the middle of the night, I was thinking, I should write, but if I turn on the lights, it’ll wake her up. And so I just sat there and started writing in the air and I thought, wow, this works. And it actually does. So I think the real art is translating an emotional experience into words. And by doing that, that changes the way that experience is organized in the brain, and it slows down your thinking process. And it also forces structure in creating a story. In other words, when you’re writing, you have to use a complete sentence. And if you write a complete sentence, your next sentence has to be related to that sentence you just finished. So it’s all part of putting together complex emotional topics.

Mills: So it sounds like a really great tool. Would you say it could substitute for therapy or should you do expressive writing in conjunction with therapy? Might that supercharge the effect that you’re getting?

Pennebaker: So, of course, if I’m talking on an APA podcast and say that this could take the place of therapy, I would be drummed out of the field. I do think that writing is a powerful method that anybody can use any time and it’s free. And if you write a few times and you don’t feel as though you’re getting any better, I think therapy is a really good alternative or jogging or talking to a friend. I also think writing works wonderfully in conjunction with therapy. And I would bet that a very high percentage of therapists actually use some form of writing in conjunction with therapy. And I’ve even recommended the therapists to, if they would like to speed up, that the effectiveness of a therapeutic session is when the person shows up to meet with you, and they typically are in the waiting room for 5 or 10 minutes, put them in a room and have them write for 5 or 10 minutes and then bring them into the therapeutic session. And when my friends have tried that, they’ve been impressed that the people are much more focused on what they want to talk about. They have a better sense of the issues, and it makes therapy go, I think more efficiently.

Mills: Let’s change gears now and talk for a minute about you more recent work, which is on language analysis and what we can learn about people from studying their word use. How did you move from expressive writing into this work? How are they connected? If they are?

Pennebaker: They are intimately connected. I was doing these first studies on expressive writing. This is we’re now into the mid-1990s, and I was trying to understand who benefited from writing. People didn’t always benefit from writing, and the stories that people were writing were unbelievably powerful. And I was curious, is there some way of writing that may be more effective than others?

And so I initially was able to get a large number of graduate students in the counseling program and have them read these essays and to evaluate them in terms of all these dimensions: How self-reflective is this person? How effective do you think writing wasn’t changing them? To what degree did they find some kind of meaning? How emotional were they? Et cetera, et cetera. And what I discovered was that having people, first of all just read these traumatic stories, and people, by the way did write really traumatic events that you and I and everybody listened to this, if you read them, you would agree that these were really upsetting experiences. What we discovered was that people who were reading these stories got depressed reading them because they were so upsetting.

And also the judges who were reading them didn’t agree on any of the ratings. In other words, here we were essentially having to do mini therapy with our raters, and their ratings were not internally consistent or valid. And so here was, oh, and it was really slow and expensive. This is not a way to build a career if you’re trying to understand writing. So I thought—I had taken a little computer science in college, and I had a sense of how you could write a computer program that could analyze it. And by the way, I’d looked to see if there were any computer programs that existed at the time and there weren’t. I talked to experts all over the country and I told them what I was looking for, and they all agreed it was an interesting project, but they didn’t know who could be helpful. So by the way, I should say about 4 years after I had finished writing this computer program with my students, I discovered that there in fact had been a really wonderful computer program written by a group out of Harvard in the 1960s, which that had proceeded me by quite a bit. It wouldn’t have worked for me at the time because it was built for a giant mainframe computer. But in any case, fortunately I had a graduate student in my lab at the time. This was about 1992, 1993, Martha Francis, whose undergraduate degree had been in computer science, and she had taken some years off and had come back to graduate school in social psychology. And I sat with her, I said, Martha, I’ve got a great idea for an experiment or for a computer program, and I think it should be pretty simple. I can’t believe it’ll take more than 2 or 3 weeks to write.

We worked on it. And that 2 to 3 weeks kind of expanded 2 or 3 years, but such is the nature of science. And it wasn’t the program itself that took so long. It was if you are going to tell a computer to look for anger words, you have to tell the computer what are anger words to look for? And so we had to make all of these dictionaries of words, and we started off with the way everybody starts something like this, looking for emotions, anger, happiness, guilt, sadness, et cetera. And then we also wanted some therapy related words. Words like “self-reflection,” “causal thinking”—words like “because,” “cause-effect” and other dimensions that psychologists were interested in. And then as long as we’re at it, we might as well do parts of speech. Those don’t require a lot of work to figure out. So articles and prepositions and pronouns and so forth.

So that’s how we put this computer program together. It was called—we called it Linguistic Inquiry and Word Count, LIWC. And I should say I had been given a lot of flak because earlier in my career, I’d come up with names of scales that were cute—and cute was not viewed as not too serious. Linguistic Inquiry and Word Count, LIWC. Now that’s serious. But then we called the computer program Luke. So if you ever hear the word Luke, it’s either a book of the Bible or it’s my computer program. In any case, the LIWC program ended up being a marvelous program. It’s now in its maybe sixth or seventh edition, and the newest version is a real machine.

But what it allowed us to do was to go through and go into any expressive writing study and look at the essays people have written. And we started to see that there were certain features that predicted improvement. So if a person, ironically, if a person is writing about a trauma, the more they’re able to use positive emotion words, the more likely they are to benefit. And it’s kind of ironic. So if they say they’re not—nobody cares about them. They feel no love from anybody. Those words—“care,” “loves,” things like that—are positive emotion words. So even if they’re negating those positive emotion words, they’re better off than if they don’t use positive emotion words. In other words, if I say I’m not happy, I’m still thinking along that dimension of happiness. If I say I’m sad, I’m not thinking about happiness as the alternative. So that was one of the first surprising things.

The more powerful effect, however, were cognitive words. Early on we were focusing on a number of types, but one were causal words because cause, effect, reason, rationale, and another type were what we called insight words—“understand,” “realize”—meaning words such as that. And what we discovered was that the more that people use those words over the course of their writing, the more they improved. That is the more they were trying to understand what was going on, the more they were trying to construct a story, the more likely they were to benefit.

Some later studies, we discovered some other issues as well. Pronouns, for example, are really powerful. As a social psychologist, I should have picked up on this immediately, but when I’m using “I” words by definition, when I say I’m feeling so-and-so, my use of the word “I” is a very brief awareness of self. So “I” words are really interesting. We know that actually people who are depressed start using “I” words at very high rates. And when their depression lifts, “I” words drop significantly. And people who were anxious use more “I” words, people who were sad.

Ironically, we’ve used this to predict actually suicide. We’ve looked at suicidal poets versus non-suicidal poets. And the two poets don’t differ in terms of negative emotion words, the big differences is suicidal poets use “I” words at much higher rates. And again, they’re focusing so heavily on self compared to non-suicidal. That’s really fascinating. Isn’t that interesting? Yeah, yeah. What we found was that people who benefited the most from writing would change in their use of pronouns over the course of the three or four writing sessions. So one time they might use “I” words at high rates, and other times they’re using other pronouns: she, they we. And then back to I and then others. And what they’re doing is they’re changing their perspective.

And I think this—therapists do this themselves. When a client comes in and you ask them, what’s going on? Well, my marriage is a problem. And then they’ll say, well, I feel this. I feel this. I feel, I feel this, I feel this. And the therapist at some point will say, yeah, yeah, yeah. What about the other person? What’s going on with the other person? Or if the person comes in problems with my marriage, yeah, my spouse does this, they do this, they do this. And the therapist will say, yeah, yeah, yeah, but how do you feel? Tell me about yourself. In other words, the therapist is naturally doing what people who benefit from writing do. They’re able to go back and forth between self and others.

Mills: So changing perspective is helpful.

Pennebaker: That’s right. That’s right. And people who write the same way all 3 or 4 days hardly ever benefit. And in fact, writing about the same story in the same way is really the same as rumination. It’s almost a marker that writing is not beneficial. And if you find that you tell the same story over and over again, you’re not getting any traction.

Mills: Yeah. I was just going to ask another word based question. You have talked about the difference between content words and style words. How are they different? What do we learn about analyzing those words?

Pennebaker: It’s funny. That was what I think the biggest breakthrough I had with the language work. When I started doing this, it was just at the creation of the internet. This was about 1995-96, and there was this new entity called America Online. And I suspect, Kim, you know about American Online, and I’m guessing I do many people out there. Many people out there in podcast land won’t know it. AOL was this incredible breakthrough. You got a disc you could put on your computer that if you had a modem, it allowed you to connect to the internet. And was what was fascinating was AOL had these chat groups and you could be in these chat groups with people from all over the world. And for somebody like me, it was unbelievable because you could download chats. And what I would do is at night, we’d put the kids to bed, and then I’d go into my study and I would just go to chat rooms and I'd just download.

And sometimes I’d go to chat groups that were predominantly women and others that were predominantly men and so forth. And I started looking at the differences at the beginning between men and women and how they use language. And I had all of these theories that many people probably would share that I had, that I just assumed women would use “we” words at higher rate than men. No, there was no difference. Well, at least men will use “I” words a lot more than women because they’re much more narcissistic? Actually, women consistently use “I” words more. I was thinking, what the hell? It must be a fluke. So I tried it with other data sets. I kept finding it over and over again. And what about cognitive words? Well, men, I would assume they use them more. No, women use cognitive words at much higher rates than do men articles and prepositions.

Who knew? I had no idea. But men consistently use those more than women. The only thing that I, oh, what about emotion words? Well, obviously women use emotion words more. No, there’s virtually no difference. The only difference were social words, references to other people, friends, men, woman, child, et cetera. Women use social words at much higher rate. And that’s the only one I got right. And I should tell you, when I’ve given tests to people all over the world and asked them, who uses these different types of words, more men or women, everybody makes the same mistakes that I did. Although nowadays, people are getting more sophisticated. But what’s happening is the words that I had discovered are called function words. And these are words that don’t have a fixed meaning. So “she” means one thing in one context, but in another context, it’s referring to another woman versus “a, the” refers to something that you and I both know what it is. When I talk about the table, somehow it’s been referenced in our conversation before. If not, we say a table, my table, that table, et cetera.

And these, all these words are interesting because they’re processed the brain differently, and they’re heavily, heavily psychological. They’re related to, as I suggested, depression. They’re related to telling the truth versus lying. They’re related to relative status. They’re related to, you can get a sense of how long the two people have known one another. It tells us about their personality and much, much more. So this was where I started to become so excited about how language could be used to understand individuals, but also couples, but also groups. And then I started to become interested in entire societies. And nowadays with social media, we can start to track how a culture is shifting.

So for example, with COVID, this is not going to come as any shock to anybody, but COVID turned our culture upside down, and in some ways, in possibly permanent ways, certainly put us on a different cultural track. And we can see this through the analysis of Twitter or Reddit or other social media data sets that when an entire culture is anxious, the entire culture starts using “I” words at higher rates. They start using anxiety words at higher rates. One other thing that I think people will be interested in, working with something like Reddit, we’re able to download hundreds of thousands of Reddit posts, or we could easily do it up until about a year ago, but we could still do it in other ways.

Now, working with one of my students, Sarah Seraj, what we started to do was to look at how do people behave when they are going through a breakup? And the ways that we study relationships in psychology historically has been find people who’ve had a breakup and give them questionnaires. Tell me what it was like. Come on, this is a questionnaire. People are constructing stories. What’s happening when they’re going through a breakup? Well, if you’re lucky doing a big study, a few people in your group will have a breakup and you can study them a little bit. You give them questionnaires.

But I don’t find that very compelling research. I want to capture a large group of people going through breakup, and I want to track them every day of their lives from before to after. So with social media, I can’t do every day of their lives, but I can go to—there’s one of the subreddits is breakups. And people go there and the average person who goes to breakups will post once about their breakup, sometimes asking for advice, sometimes just venting. So what we did in our study was we went to the breakup subreddit and downloaded everybody who posted at least once and then asked them to be in our survey or our study, we never actually met any of these people, it’s anonymous.

But what we would do is if a person had posted at least a month before somewhere on Reddit, and they had posted at least once or twice in a month afterwards, somewhere, they were eligible to be in our study, and our admissions were a little bit more stringent than that, we ended up getting about 6-8,000 people. And imagine now we go and we find when they make their very first post, the average person makes their very first post in the breakup subreddit typically within 2 weeks of their breakup. And then what we do is we go and look at all of their posts up to a year before the breakup and to 1 year after the breakup. And they’re posting in all sorts of things. The average person doesn’t post very many times in the breakup subreddit. And in fact, we pull out all of their posts dealing with relationships. So these, they’re doing ask Reddit, they’re into photography or jokes or what, it doesn’t matter to us. We get all of their posts and comments. And the average person had over a hundred posts and comments in that year before and after, meaning we are capturing their lives and their interactions with people that they’re not even telling that they’re having a breakup. Does their writing or thinking change? And it does. And in fact, we see signs of this breakup that their language starts changing at least 3 months before the breakup.

Mills: So you can predict that they’re going to break up, right?

Pennebaker: That’s right. Not very well, I’ll be honest with you. But what I love about it is about 75-77% of the people got dumped by their girlfriend or boyfriend, and when they write, they say, this came out of nowhere. It didn’t come out of nowhere. There were signals. And that person was broadcasting these signals and it got closer and closer to breakup. We saw changes in their thinking, changes in their emotional state, et cetera. And then we could see how long the breakup affects people. And the average person you see released significant effects in terms of their thinking and emotions for 6 months afterwards. The breakup—we came away with a much deeper appreciation of how a breakup really messes people up. And you can read it. They think much less logically. They are disoriented. They are really under a huge amount of stress. And most employees or employers are not going to say, oh you’ve had a breakup. Yeah, take a week off. They never think about that. But the reality is they probably should. Anyway. So we’ve been using this in all sorts of ways to start to get a sense of how can we measure the effects of a breakup or taking a psychedelic drug or COVID or other events, everything from Black Lives Matter, et cetera. How can we now use this to get a sense of how individuals and groups change?

Mills: Well, I just want to ask you one last question. You’ve studied what we can learn about politicians by analyzing their word use. And given that we’re in the middle of a political season right now, I’m just wondering what have you found? Are you looking at current candidates, and if so, what are you learning from their choice of words? And can you predict what the outcome of the election might be based on those words?

Pennebaker: I have never been able to predict who’s going to win based on words, but I can tell you how that candidate thinks. And the reality is we have two likely candidates, and we already know how they think, and they think probably in ways that you would imagine. It’s quite interesting. One of the two candidates uses “I” all the time, and some people think that’s narcissism. Well, it can be narcissism, but it’s also insecurity. And the one who uses “I” words at very, very high rates more than any previous president, of course, is Donald Trump. And he’s been using it at particularly high rates over the last couple of months. But he always has.

Another dimension that we always look at is something that we call analytic thinking. Analytic thinking is really kind of interesting. It’s this goofball algorithm that we came up with. And it’s essentially, if you come right down to it, people who use a high rate of nouns and a low rate of verbs or another way to think about it, high rates of articles and prepositions and low rates of pronouns.

And people who are high in this analytic thinking, looking at their admissions essays, they do much better at university than people who are lower in analytic thinking. It’s correlated with IQ, et cetera. And it’s essentially a marker of formal, logical, analytic thinking. Donald Trump, not surprisingly, is the lowest of any president ever in terms of analytic thinking. And Biden’s kind of there in the middle. He’s not high, but he’s not extremely low in terms of clout. This is another measure that we have, which is this sense of certainty and conveying that they are certain of what they’re saying. Both Biden and Trump are high. Trump is again highest, and he’s one of the highest ever. Obama was very high in clout also.

In fact, if there’s anything—a really interesting project that one of my former students, Kayla Jordan did, and Ryan Boyd’s been involved in all these studies as well, was looking at the history of U.S. presidents, but also prime ministers in the U.K. and Canada as well. And ever since about 1910 or so, we have been electing presidents who are higher and higher in clout. That is, more and more certainty with a corresponding drop in analytic logical thinking. And I think what’s happening is we, the people, want people who convey things in a very simple way with absolute authority. Which—hey, people, that’s not a very smart way to pick a leader, because what we’re picking are people who often are not that bright, but they’re certain, and that’s what we like. And it’s the Cheetos theory of elections, which is if it tastes good and it’s crunchy, of course we like it. And that’s how I think we often vote. And by the way, Democrats, Democratic candidates have been shifting this way, just have Republican candidates. It’s because the populace—and the Brits the same, even though they select the prime minister differently. And we see the same thing in Canada as well. It’s increase in certainty and decrease in logical analytic thinking.

Mills: Well, Dr. Pennebaker, I want to thank you for joining me today. This has been really interesting. We could have gone on and on, but I think we’re out of time here. So thank you again.

Pennebaker: I’ve enjoyed it.

Mills: You can find previous episodes of Speaking of Psychology on our website at www.speakingofpsychology.org or on Apple, Spotify, YouTube, or you get your podcasts. And if you like what you’ve heard, please subscribe and leave us a review. If you have comments or ideas for future podcasts, you can email us at speakingofpsychology@apa.org. Speaking of Psychology is produced by Lea Winerman. Our sound editor is Chris Condayan. 

Thank you for listening. For the American Psychological Association, I’m Kim Mills.