The Secret Life of Pronouns: What Our Words Say About Us

by James W. Pennebaker

  Across the seventy thousand blog entries, the terrorist attacks resulted in an immediate surge in the expression of negative emotions that lasted about two days; then negative emotions returned to pre-attack levels in about eleven days. At the same time, the drop in positive emotion word use showed an even more striking pattern. After a precipitous drop on 9/11, the use of positive emotion words returned to pre-attack levels within four days. By ten days after the attacks, people’s positive emotions were higher than they were before the attacks.

  Note: Lines reflect the use of positive and negative emotion word usage of 1,084 bloggers. Baseline is the average of two months of blog posts prior to September 11. Data points through September 25 are by day, and thereafter by week through November 6.

  One final important finding emerged from the 9/11 project. In the five or six days after the attacks, the bloggers used cognitive words at much higher rates than before 9/11. Recall that cognitive words include words that reflect causal thinking (e.g., because, cause, effect) and self-reflection (e.g., understand, realize, meaning). The cognitive words typically indicate people’s trying to understand what is happening in their lives.

  An increase in the use of cognitive words immediately after an unexpected event makes sense—we all want to know what happened and why. However, starting about a week after the 9/11 attacks, the bloggers’ use of cognitive words dropped to unprecedented lows for the next two months. In fact, markers of thinking were far below the levels that existed before the 9/11 attacks.

  Consider the implications of all these effects. The analysis of the thousand bloggers paints a picture of how normal, everyday people think following a large emotional upheaval. Note that within a few days of the 9/11 attacks most of the bloggers returned to writing about their usual topics—shopping plans, boyfriends, pornography, pets, the usual. Across all topics, however, the same patterns were emerging. A summary of the bloggers’ language suggests:

  • Shared traumas bring people together. They pay more attention to others and refer to themselves as part of a shared identity—as can be seen in increased use of the warm you-and-I form of we.

  • Shared traumas deflect attention away from the self. Even though people may feel sad, they are not depressed. Recall that people who are truly depressed show higher I-word usage, not lower I-word usage.

  • Shared traumas, in many ways, are positive experiences. For at least two months after the 9/11 attacks, people expressed more positive emotions and were more socially connected than they had been in the months before the attacks.

  • Shared traumas make people stupider. OK, maybe not stupider but certainly less analytic. Within a week of the attack, people wrote in simpler ways, suggesting that they weren’t thinking deeply about their writing topics. In fact, they seemed more passive and accepting of new information.

  • People’s reactions to traumatic experiences change over time. The ways people think, feel, and pay attention to their worlds change drastically in the hours, days, and weeks after an emotional upheaval.

  BEYOND 9/11: MAN-MADE AND NATURAL DISASTERS

  Evolutionary psychologists look at the 9/11 findings and note how they make evolutionary sense. If we are in a small group on the savannah and are attacked by another group, it is critical that we band together, focus outwardly, and prepare for possible future attacks. These are adaptive reactions that can help to increase the survival of the individuals and the group itself.

  Do the same language patterns exist for shared emotional upheavals that do not directly threaten the group? The answer appears to be yes. Over the course of my career, I have studied the social and psychological effects of a number of large-scale upheavals, including the Mount St. Helens volcano eruption in Washington state in 1980, the Loma Prieta earthquake in the San Francisco bay area in 1989, the first Persian Gulf War in 1991, the death of Princess Diana in 1997, and the tragic death of twelve students at Texas A&M university during the building of their traditional bonfire in 1999.

  Across all of the studies, similar themes emerge that bolster the 9/11 findings. The most striking phenomenon is that all kinds of upheavals bring people together. People become more selfless, more concerned with others, and actively seek out relationships with others. Interviewing people in Yakima, Washington, several weeks after Mount St. Helens had dumped almost four inches of sandlike ash on their community, most of the residents reported that it had been a frightening experience but that, despite their losses, they were glad it happened in their lifetimes. Most reported meeting and talking with neighbors they never knew.

  The university community of Texas A&M was shaken by the deaths of twelve students caused by the collapse of their symbolic bonfire. Again, in interviews, online blogs, and newspaper accounts, the language of the community was selfless and warm. In fact, the social bonds of the university ended up being tighter than anyone had seen before. In tracking the physical health of the student body over the following year, we found that the students went to the student health center for illness 40 percent less compared to the year before the bonfire disaster. No such effects occurred at A&M’s neighbor, the University of Texas at Austin.

  It’s almost heretical to admit but terrible experiences can bring out the best in us. By their very nature, traumas can destroy some lives and enrich others.

  USING LANGUAGE ANALYSES TO GUIDE MENTAL HEALTH TREATMENT IN THE IMMEDIATE AFTERMATH OF AN EMOTIONAL UPHEAVAL

  One of the best-kept secrets in the mental health world is that most people actually cope quite well when faced with truly horrible life events. Most of us think just the opposite. When researchers ask us how we might react if something horrible happened to us, they find that we tend to believe that personal traumas would psychologically cripple us. In fact, the majority of people who have experienced torture or rape or who have survived terrible car or plane crashes or other unimaginable events don’t evidence symptoms of post-traumatic stress disorder (PTSD) or major depression.

  Humans, it seems, are remarkably adaptable. They also naturally know what to do when a trauma strikes. Keep in mind that the overwhelming number of bloggers that have been studied are in good psychological health—meaning that whatever they do is probably not a bad idea. One finding is clear and a little counterintuitive: It can be healthy to distract yourself from your pain. Paying bills, helping others, playing video games, or cleaning your house are all time-honored methods of coping with profoundly disturbing news. If you want to talk with people about the event, talk to them. If you want to be by yourself, be by yourself. There is no convincing evidence that any single coping method works well for all people.

  If you are a mental health professional, this advice is very unsatisfying. If a disaster hits and you see people in great distress on television, your natural inclination is to run out and help. Again, all the evidence indicates that the best way you can help is to make sure people have their basic needs fulfilled, that someone can answer their questions, and that someone will listen to them. People do not need in-depth psychotherapy right after a disaster hits. Far from it.

  In the 1990s, a well-meaning psychological intervention became popular that encouraged traumatized people to disclose their deepest emotions about their trauma within the first seventy-two hours after it occurred. The movement, then called Critical Incident Stress Debriefing (CISD), sounded quite reasonable and was adopted by emergency workers, large corporations, and wings of government around the globe. Despite the good intentions behind the program, there were some serious problems with it. After a large number of studies, it is now generally agreed that CISD probably causes more harm than good. In looking at the ways people use language online in the first hours after a trauma, deep emotional processing immediately after an emotional event is not what healthy people do.

  Most people need social support after an upheaval. One of the most striking phenomena my students and I have seen in studying blogs is the degree to which people seek and receive support from others. When people
are distressed after an upheaval, they merely mention their emotional state and a flood of well-wishing comments from friends and strangers follows.

  As part of her doctoral dissertation dealing with the language of diet blogs, Cindy Chung tracked what factors led to successful weight loss across several hundred dieters in a diet blog community over several months. Many of the bloggers focused on the details of losing weight but the majority also wrote about their relationships, life experiences, and emotional issues. Many dieting experts would recommend that the best way to lose weight is to obsessively list the foods you eat, the number of calories you have taken in, and the amount of exercise you have done on a daily basis. Not true according to Chung’s research. Instead, she found the very best predictor of successful weight loss was being involved in the online social network. That is, the more comments or posts a person sent out and received, the more successful they were at losing weight. In addition, bloggers writing about personal and emotional issues were far more successful in losing weight than those who wrote only about their foods and diets.

  THE LONG-TERM EFFECTS OF TRAUMAS: HOW WORDS CAN HEAL

  As the blog data suggest, most people initially cope with a traumatic experience by seeking out friends and family. The ways people talk or write about the upheavals vary considerably and may not involve any deep emotional discussions.

  If people could always respond to emotional events the ways they wanted or needed, there would probably be less of a demand for therapists and physicians. Traumas, even highly personal traumas, are deeply social. If I have been raped or mugged, I may be hesitant to tell others because of the humiliation or the fear that others will treat me differently. Consequently, my natural reactions to talk openly about the experience will be stifled. Similarly, I may need to appear strong and happy to my family to avoid their becoming anxious or depressed.

  THE DANGER OF SECRETS

  Events that are shameful, embarrassing, or could damage one’s reputation are often kept secret for years. I discovered this early in my career when I included a questionnaire item that asked people if they had experienced a traumatic sexual experience prior to the age of seventeen. In surveys of thousands of adults, approximately 22 percent of women and 11 percent of men said they had. Particularly striking was that this same group had terrible health compared to people who had not had traumatic experiences. Later studies showed that the problem was that the sexual traumas were almost always secret traumas. Any type of major upheaval that people kept secret from others tended to compromise their physical and mental health.

  Big emotional secrets are toxic for several reasons. One of the first interviews I conducted that illustrated this was with a thirty-five-year old woman I’ll call Laura, who had been about twelve when her mother remarried. Starting a few months after the marriage, her new stepfather, Jock, would occasionally sneak into Laura’s bedroom in the middle of the night and fondle her. Terrified, she tried to get him to leave even though he would make light of the situation. This continued off and on until Laura was fifteen, when she left home to live with an aunt. As she described it:

  I had always been close to my mother. The divorce had nearly killed her and she was so happy with Jock. If she had known what Jock was doing to me, it would have broken her heart. I wanted to tell her so much. Do you know what it is like to be in a family like that? I’d get up in the morning and Jock and my mother would come down together. He would smile and be friendly, like nothing had happened. I hated his guts but could never tell anyone why. Every morning, every evening, every time I saw that bastard, I felt sick to my stomach.

  Looking back on it all, the very worst thing was that I couldn’t talk to my mother anymore. I had to keep a wall between us. If I wasn’t careful, the wall might crumble and I’d tell her everything. The same was true of my friends. I’d go out with my girlfriends and we would all giggle about boys and dating. Their giggles were real, mine weren’t. If they had known what was happening in my bedroom they would have died.

  I have heard or read variations on this theme hundreds of times over the last twenty-five years. In Laura’s case, the fondling was horrible but the collateral effects were worse. All of her close relationships with family and friends were damaged, her physical health deteriorated, and she was not able to talk about the experience to anyone until several years later.

  Not talking about a major emotional upheaval violates our natural ways of behaving. As we have seen, emotional events provoke conversations. If we witness a terrible accident, discover that our favorite sports team has won an important game, learn that our closest friend has just left her husband, we feel the need to talk about it. This urge to talk about unexpected and upsetting events is fundamentally human. It is found in all cultures that have been studied. People talk about emotional experiences to learn more about them. Talking, it seems, is one of the primary ways that we are able to understand complex experiences. Conversely, when people are not able to talk about emotional events, they tend to think—even obsess or ruminate—about them.

  TRANSLATING SECRETS INTO WORDS: EXPRESSIVE WRITING

  As described in the first chapter, the links between secrets and illness motivated me to turn this idea on its head. If we encouraged people to talk or write about upsetting experiences, would their health improve? The answer turned out to be yes. In our first experiment, college students took part in a writing study where they were initially told that they would be writing about assigned topics for fifteen minutes a day for four consecutive days. After agreeing to participate, half of the people were told that they would write about deeply traumatic or stressful experiences for the four days. The other half of the students were instructed to write about superficial topics, such as describing objects or events.

  Overall, people who were asked to write about emotional topics exhibited better physical health than those who wrote about superficial topics. Those in the experimental condition who wrote about traumas went to the doctor at half the rate of people in the control condition in the six months after the experiment. Later studies found similar patterns. Writing about traumatic experiences improved people’s physical and mental health.

  Other researchers soon found that expressive writing affected immune function and other biological processes associated with health and illness. Josh Smyth from Syracuse University and his colleagues published a powerful study with arthritis and asthma patients showing that writing influenced the course of the diseases. Other projects with people dealing with AIDS, cancer, heart disease, depression, cystic fibrosis, and a range of other physical and mental health problems benefited from expressive writing. Now, almost twenty-five years later, over two hundred scientific articles have been published pointing to the power of writing.

  The early writing studies were the impetus to develop a text analysis method that could help us figure out why writing worked. Awakening people’s emotions about earlier upheavals forces them to think differently about them. Exploring emotional topics demands that people look inward (as we see by their increasing use of first-person singular I-words). It helps them to organize their thoughts and construct more meaningful explanations or stories of their lives (as seen in the increasing use of cognitive words). By tracking the ways people write about their traumas, we are witnessing how they are changing in their thinking.

  Writing is not a panacea and its effects are limited. For example, there is no evidence that writing about an emotional upheaval immediately after it occurs is helpful. Although writing for a relatively brief time over a few days has generally worked, it is unclear that long-term diary writing is necessarily helpful. In fact, writing too much about a particular problem may be a form of rumination. My recommendation is if you are interested in expressive writing, try it out for a few days. If it is beneficial, great. If not, try something else.

  TYING IT TOGETHER: EMOTIONS AND THINKING STYLES AS TWO SIDES OF THE SAME COIN

  Remember Rex Ryan, the coach of the New York Jets football team who cr
ied in front of his players after losing an important game? His emotional display reflected a change in the ways he was thinking about his team and its potential. Equally important, his tears were a powerful social signal to the players. You will recall that one of his players was quoted as being impressed by the coach’s commitment to the team and how his crying brought the team together. In a separate article, the same crying scenario signaled to a sportswriter that the New York Jets were falling apart.

  Emotions are not just reactions to events. Different emotions can change the ways we think and influence how we respond to others. Emotions are intensely social in that they can draw us closer or push us farther apart. Emotions are also meaningful signals about other people’s motivations, goals, and intentions. The intimate connection between function words and emotional state naturally follows. Emotions make us think about the world differently and function words reflect this change in thinking.

  The relationship between thoughts and feelings has been the subject of heated debate in philosophy and psychology for centuries. Both Aristotle and Plato argued that logic and emotions were fundamentally different processes. Descartes, writing in the seventeenth century, went farther by claiming that emotions undermined people’s abilities to think rationally. The early American psychologist William James also emphasized how emotions and passions frequently blinded people’s judgments. Sigmund Freud argued that fundamental emotional issues were the driving force of personality and behavior.