Part I: Studying Communities In Crisis, Dr. Pennebaker Speaks About The COVID-19 Crisis
Over his 50-year career, Dr. James Pennebaker’s research has focused on natural language use, health, social behavior, and the relationships between these. He currently holds the position of Regents Centennial Professor of Liberal Arts in the Psychology Department of the University of Texas at Austin. During the 1990s, Dr. Pennebaker developed the Linguistic Inquiry and Word Count (LIWC; pronounced “Luke”) program to facilitate further work expanding on evidence he found connecting the frequency of some types of words to psychological states.
This is Part I of this 4-part series in which Dr. Pennebaker dicusses the methods involved in studying communities in crisis, as well as some early results from research on the COVID-19 crisis.
Can you tell us a bit about your past work studying shared crises?
I've been studying people in crisis for a very long time. In fact, my very first study was one in 1980 looking at Mount St. Helens volcano. Two decades later I was looking at how people changed the way they interacted online through blogs on 9/11. And then I ended up studying many other things -- the death of Princess Diana, the Oklahoma City bombing, and many things since including the most recent, which is what's happening with the coronavirus.
In all of them, I've been able to track how large groups of people change in the way that they talk or use language, whether it's through instant messages, through blogs, through Reddit or Twitter or whatever. And I've tied how people change over the course of an event to how individuals change when they have their own traumatic experience. And what happens is, when we are faced with an upheaval in our lives, we start thinking differently. We connect with other people differently. We focus our attention very differently.
Can you tell us more about the methods you use to do your work?
Most of the work I've done is based on a computer program, LIWC, which we call "Luke". It stands for Linguistic Inquiry and Word Count. What it does is it goes through and it counts words and word categories, so in any given text we can calculate the percentage of words that are positive emotions and negative emotions, but also words that get at the heart of the way people think and the way they connect with others. They're often words that we don't pay attention to -- stupid things like pronouns or prepositions or articles, and by analyzing these we can get a sense of where people are paying attention, how they're thinking. If we're looking at a work environment, we can get a sense of whether people are paying attention to the topics that they should be attending to or whether their minds are wandering, thinking about other things, or thinking about nothing at all.
For September 11th, we got a thousand people who were big bloggers and we tracked them from two months before 9/11 to two months after. Because they were high frequency bloggers, we could track how all thousand of these people changed over time, so that once 9/11 occurred we found certain huge changes. For example, unsurprisingly, there was a big bump in negative emotions that lasted for several days and took about almost two weeks to get back to baseline. Positive emotion words showed a big drop, also unsurprisingly, on 9/11, but then by four days later their levels came back up to baseline -- and after that they went above baseline. They used more positive emotion words in the next two months we collected data, which was one of these fascinating phenomena. It showed that traumatic experiences are sad and upsetting, but by the same token they’re often associated with bringing people together. There's a greater sense of cohesiveness among people, and people's moods, ironically, are a little bit better. And again, this goes against common sense. Their mental health often is much better after an upheaval.
How does this current COVID crisis compare with those past crises and upheavals?
The current crisis is unlike any crisis or trauma I've studied. Think about most traumatic experiences -- it could be September 11th, it could be a horrible school shooting, it could be a natural disaster like an earthquake or hurricane, it could be almost anything. After all disasters, it brings people together. They often go outside. They talk with their neighbours. They go to memorial services. They are much more sociable. They are more connected. They might be nervous, they might be frightened, they might be sad, they might be angry -- but they are closer to their neighbours and they're closer to their friends. Oddly, it's a social experience.
This one is so different. So here we have this disease we can't see, we don't know who has it, we don't know how at risk we are at any given time. The death rate's pretty high -- it’s terrifying. And unlike every other disaster, instead of going and talking to others, we are being mandated to stay indoors and hang around with the same people we've been around for days, and then weeks, and now over a month. Who knows how long? So, all of a sudden, the rules are different.
We're not able to talk in the ways that we have -- nowadays we do have social media and we have other ways of connecting, but it's a profoundly different phenomenon than it's ever been. So now we’re doing a lot of research trying to get a sense of people's social lives. We're using large-scale surveys. We're using Reddit and Twitter and other sources to get a sense of how individuals and communities and countries and indeed the entire world is changing in the ways that we're connecting with each other, the ways we're thinking, the ways we're sleeping, the ways we're drinking, the ways we’re feeling depressed, the ways that we’re feeling elated, and how productive we're being. So, we're trying to bring together all of these different types of data. Some of them are big data using language and the LIWC program, others are questionnaires, and others are these big surveys we're doing where we have people just write and tell us what's going on in their lives.
Can you share some early results from your research on the COVID crisis?
The research we've been doing is almost overwhelming. I've been doing this with a group of graduate students, as well as some colleagues at University of Michigan and elsewhere around the world. So, we have the COVID surveys, which are approaching 20,000 people in the United States and around the world. And then we're getting other kinds of data as well. We're right in the middle of data collection. What I'm going to tell you are things that are just popping up almost randomly, so don’t hold me to anything. Although in some cases the effects are very robust.
Emotion: how are people reacting? Not surprisingly, the biggest, most overwhelming effects I have ever seen in my life are levels of anxiety and fear. I have never seen anything like this. And you can see it in Reddit. If you're not familiar with it, Reddit is a glorious online community -- glorious for a researcher like me. You have millions of people who are connecting, talking about all sorts of things. And we get people in city subreddits (Reddit sub-sections) -- for example, the New Orleans subreddit, or the New York City one, or the Seattle one, etc. We're able to just see what they're talking about. How much are they talking about COVID? What's their emotional state? Because we have access to the way these people were talking in the months before this ever started, we can see what the average level of an emotion like anxiety might be. Usually it's very, very low. And then on March 11th, when the World Health Organization called this a pandemic, for the first time people in America and Canada suddenly woke up and said, “my God, this thing's real”. There's this huge jump in anxiety at levels I have never seen in my life. They are unbelievable. And interestingly, Seattle was about five days ahead of the rest of the United States. It was the first city in North America where there were a lot of cases. And then the next one that jumped a few days before the rest was New York City. And what you see is very, very high anxiety levels, and now over the last month they have been dropping. And it bumps up and down depending on things like the number of deaths that day, but it's gradually returning back to normal.
And another big issue is how people are paying attention, how they're thinking. One of our variables is essentially verb tense. Are people looking forward? Are they looking into the past? Are they talking about now? And not surprisingly, starting around March 12, future-focus went up tremendously. In other words, all of a sudden people were talking about the future. And of course, that happens when you're under threat. And now that's been returning, getting closer and closer to baseline. And then the other one that is so interesting for me is how we are connecting with others, how we are socially connecting. A good example of how we're connecting is our use of the word we. With 9/11 there was a huge increase in we words (including us, our, ourselves, etc.), immediately after the September 11th attacks and it took two months to get back to baseline.
It's a lot weirder now. There has been an increase in we, but not anything like 9/11. And I think part of it is there's not that same sense of we. We're asking people to what degree do you feel closer to your family, your friends, your neighbours, your city, your country. And we're finding that they feel closer to their family, a little bit closer to their friends, even less close than they were before to their neighbours, less close to their city and less close to their country, which is bizarre. Almost any other crisis brings up a general sense of togetherness that we're not seeing here. And we're seeing this reflected in some of our survey studies as well.