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

by James W. Pennebaker

  More broadly, could we get the same information about people by just looking at their regular language? This might include their e-mails, transcriptions of conversations, blogs, or professional writing. The answer is yes. And this is the very essence of this book. People who are analytic or categorical thinkers tend to use articles, prepositions, and negations when describing a boring bottle, discussing a backyard party, or talking with their neighbor about Mrs. Gilliwitty’s stomach problems. Of course, the ways we talk and think change depending on the situations we are in. In formal settings, we all talk more formally; when at wild parties, we are apt to talk, well, more wildly. Nevertheless, we take our personalities with us wherever we go, and no matter what the setting, we will leave behind a partial copy of our function-word fingerprint.

  A THUMBNAIL SKETCH: OSAMA BIN LADEN THROUGH HIS WORDS

  Consider the language of a public figure such as Osama bin Laden. Over much of his adult life, he left a record of his language in his interviews, speeches, letters, and written articles. Analysis of his words in Arabic or in English translation evidences his supreme self-confidence, even arrogance (very low rates of I-words, high use of we-words and you-words). Unlike most other leaders of extremist Arabic groups, including his sidekick, Ayman al-Zawahiri, bin Laden was a storyteller (high in narrative markers—past-tense verbs, social references) with a decidedly dour, hostile edge. Our overall analyses would peg him as high in need for power, moderate in need for achievement, and low in need for affiliation. Cindy Chung’s meaning extraction technique reveals that his real obsession in life switched from rage at his homeland, Saudi Arabia, to America’s incursion in Iraq and Afghanistan. Interestingly, he never showed much interest in Israel compared to his al-Qaeda colleagues. No data on whether he liked long walks on the beach at dusk.

  USING WORDS TO UNDERSTAND OTHERS AND OURSELVES

  Armed with the findings from this chapter, is it possible for you to “read” others better? Can the work on language help you to be a better or more effective person? Let’s go with a qualified “yes” to both questions.

  Recall from the second chapter the story about Senator John Kerry’s aides urging him to use the word we more and I less in his speeches? His very bright group of advisers falsely assumed that a person who uses we in a speech makes listeners feel closer to the speaker. This should serve as an object lesson to anyone who wants to read others’ personalities by analyzing their language. The primary rule of word counting: Don’t trust your instincts.

  If you want to get a sense of other people by examining their language, you will actually need to count their words. You can do this by hand—but it is a slow and painful process. Or you can use a computer program. (A brief overview of some programs can be found in the notes section at the end of the book.) Of course, simply counting their words is just the first step in decoding their personality.

  What do you want to know about other people? As outlined in this chapter, it is fairly easy to detect different thinking styles—whether formal, analytic, or narrative. Later chapters provide clues to detecting deception, dominance, ability to socially connect, and other dimensions. Capturing what grabs their attention through techniques such as the Meaning Extraction Method can be more difficult. Fortunately, our brains are much better at hearing content words than function words and so actual word counting may be less needed.

  If you want to find out what themes guide people’s attention, listen to what they talk about. One friend of mine is insecure about his intelligence. He’s undoubtedly smart but in virtually every conversation he drops information to prove how smart he really is. The last couple of times I saw him, he mentioned how he performed on an intelligence test in an airline magazine, a colleague’s offhand comment about something smart he said, how smart people watched documentaries (oh, and he watches documentaries). This has been a guiding theme in this person’s life and it is evident in virtually all his interactions.

  A second way to capture the themes most important to people is to watch how they guide the conversation. Years ago, I had dinner with two old friends—one who is an insightful clinical psychologist and the other an architect. In the middle of the conversation, the clinician casually noted to the architect, “It sounds like you are having some financial problems.” The architect was stunned by the clinician’s out-of-the-blue comment. But then he painfully admitted that he had recently lost his life savings due to a risky investment. I was surprised myself by the clinician’s comment because there was nothing I could discern in the conversation that hinted at the financial problems of my other friend.

  Afterward, I asked the clinician what made him think that our friend had financial difficulties. He laughed because in his mind, it was obvious. Several times during the meal, the architect would change the conversational topic—and it always had something to do with money, some kind of financial loss, or investments. Neither the architect nor I saw the regularity of his switching topics. Indeed, I’ve since learned that when someone changes the conversational direction, it serves as a powerful marker of what is on his or her mind.

  ANALYZING OUR WORDS TO KNOW OURSELVES BETTER

  If you had access to all the words you used in a day, what could you learn about yourself? Through my language research, I’ve been able to answer the question for myself. And, at least for me, it has been quite helpful.

  One way my laboratory team studies natural language is to record everyday speech in children, college students, married couples, and the elderly. One of my former graduate students, Matthias Mehl, was instrumental in developing a recording device called the Electronically Activated Recorder, or EAR. The EAR is a digital recorder that is programmed to come on for about thirty seconds once every twelve to fourteen minutes over the course of several days. Matthias, who is now an internationally respected researcher, spent thousands of hours perfecting the EAR so that it could withstand the punishment from its wearers.

  Part of the testing phase was to ask all the members of my research team to wear the EAR for a few days and then transcribe everything that was recorded on it. I have now done this myself several times. The first weekend I wore the EAR, my son was about twelve years old. In my own mind, that weekend was uneventful—chores, a family outing somewhere, the usual. A couple of days later, I transcribed my EAR recording and was distressed to see the way I spoke to my son. My tone was often cool and detached—matching his own aloofness (or was he matching my own aloofness)? In our interactions, I used big words, high rates of articles, relatively few pronouns—especially I-words. With my wife and daughter, my language was more warm and approachable.

  The experience of hearing my tone of voice and seeing my own words on paper had a profound effect on me. Thereafter, I made a conscious attempt to be warmer and more psychologically available to my son. Note that I did not make a conscious effort to change my pronouns and articles. Rather, I changed my behavior and attitudes with the assumption that my function words would follow.

  Over the years, I have also analyzed my language in e-mails, classroom lectures, professional articles, and letters of recommendation. Sometimes my language behaves in predictable ways; sometimes it doesn’t. And when it doesn’t, I learn something about myself. Indeed, it is always most striking to see those instances when my own view of myself doesn’t match the ways I’m objectively behaving. It also raises the question of what should I do—change the way I view myself, change my behaviors, or change my language.

  And this brings us back to a recurring theme of this book: To what degree does language influence psychological state or merely reflect it? Is it possible to change our psychological state by changing the ways we use words? For example, would John Kerry have won the 2004 presidential election if his advisers had told him to use I-words more and we-words less? Would his change in language have changed his formal speaking style and rigid body posture? I seriously doubt it. People can be trained to change their language but there is no compelling evidence that the language affe
cts their personality, behavior, or emotional state. Rather, had John Kerry attempted to loosen up, be more personal and genuine, his language would have followed. Language is a powerful reflection of a person but does not change the person on its own.

  CHAPTER 5

  Emotion Detection

  Who said there’s no crying in football? [New York] Jets coach Rex Ryan … in the wake of the previous day’s devastating 24–22 loss to the Jaguars, delivered an impassioned speech to his players that was so emotionally-charged it brought him to tears.… “He didn’t bash us at all; he was just very emotional … he was crying,” right tackle Damien Woody told The Post. “Rex believes in our team so much I can’t even put it into words and it would be a shame if we didn’t capitalize on our opportunity.”

  “I was a little upset to see him that way,” cornerback Darrelle Revis told The Post. “I’m upset for the same reasons he’s upset.”

  Asked if he’s ever been a part of a meeting with such high-powered emotions, Revis said: “No, I haven’t been a part of a meeting where a coach cried like that.… In the future, I hope there are more tears of joy than the one this morning.”

  —MIKE CANNIZZARO, New York Post, November 17, 2009

  All signs point towards trouble for [the New York Jets] this week, as Coach Rex Ryan cried as he addressed his team on Monday, feeling so overwhelmed with emotions. Staged or not, Ryan’s using tears to motivate and bring his team together is the official signal that the wheels have come off.

  —SAM HITCHCOCK, NewJerseyNewsroom.com, November 20, 2009

  WHEN PEOPLE BEHAVE emotionally, it gets our attention. An adult football coach crying is important information for his team, his opponents, and the football-watching public at large. The original New York Post article suggested that after Coach Ryan’s emotional display, his team was ready to take the next bus to Boston to better prepare for their upcoming game with the New England Patriots. The article a few days later from the respected NewJerseyNewsroom website viewed the same emotional display as evidence that the team was headed for disaster. And, indeed, it was. The following Sunday the Jets were crushed by the Patriots 31–14. No tears were reported the following week.

  Emotions change the ways people see and think about the world. They can motivate people to work harder or cause them to give up in despair. Emotions can broaden our perspectives or restrict them by causing us to ruminate about the same topics over and over. Emotions guide our thinking and affect the ways we talk and get along with others. Not only do we need to know our own emotions, we need to be able to read other people’s emotions to understand what they are thinking and planning to do.

  Reading other people’s emotions is usually easy if they are crying, screaming, or laughing hysterically. At other times emotions are conveyed more subtly through facial expressions, tone of voice, or nonverbal behaviors. Much of the time, however, people may be feeling one thing but not conveying it. All of us have had the experience of not knowing if our date, parent, teacher, boss, or client likes us or not. In our close relationships, someone may have failed to pick up on important emotional cues that may have damaged the relationship. In reading others’ e-mails, IMs, tweets, or letters, most of us have missed an emotional cue that the other person may have been intending to send.

  The central question of this chapter is how can we detect people’s emotions through their words. On the surface, this sounds like a simple task. If people are happy, they should use happy words. If sad, they should use sad words. If only it were this easy. Counting emotion words is a fine start to measuring feelings but these approaches miss the central point: Emotions affect the ways people think. If we could just come up with a way to measure the ways people think, we could come up with a richer way to study and understand emotions. There is just such a way. Function words do a fine job of tracking people’s thinking styles. It should come as no surprise that these same words can provide insight into people’s emotional states as well.

  DIFFERENT EMOTIONS, DIFFERENT WAYS OF THINKING

  A good starting point is to consider three distinct emotions: happiness, sadness, and anger. We all agree that these are common emotional states that have their own distinct physical and psychological feelings. These different emotions also cause us to look at the world differently. When thinking about emotions and words, it can be instructive to see how poets write. After all, poets spend much of their time writing about their emotional reactions. Do poets use function words differently when writing about happiness, sadness, and anger? Look at some of the work by Edna St. Vincent Millay, the first woman to win the Pulitzer Prize for poetry. Writing in the first half of the twentieth century, Millay was celebrated as a free spirit who wrote powerfully about love, loss, and relationships. Compare the ways she uses words in lines from the poems on the following page.

  The examples from Millay illustrate what researchers have found. When writing about positive experiences people tend to use we-words at particularly high rates. People who are happy are also more specific, relying on concrete nouns and references to particular times and places. Other studies find that positive moods change people’s perspectives so that they look at the world in a more open way—sometimes referred to as the broaden-and-build way of thinking. Sadness generally causes people to focus inwardly. Pronouns tend to track people’s focus of attention, and when in great emotional or physical pain, they tend to use I-words at high rates. Sadness, unlike most other emotions, is associated with looking back in the past and into the future. In other words, people tend to use past- and future-tense verbs more when they are sad or depressed compared to other strong emotions.

  Although classified as a negative emotion, anger has a completely different profile than sadness. When angry, people focus on others and rarely themselves. In addition to using high rates of second-person (e.g., you) and third-person (he, she, they) pronouns, angry people talk and think in the present tense.

  When events happen to us that cause us to feel sad or angry, we tend to try to understand why they occurred. We use cognitive words that reflect causal thinking and self-reflection. Not true for positive emotions such as pride and love. When happy and content, most of us are satisfied to let the joy wash over us without introspection. In other words, negative feelings make us thoughtful; positive emotions make us blissfully stupid.

  Most of the studies that have examined transient emotions in the laboratory have relied on asking college students to write about powerful events that had elicited feelings of happiness, sadness, or anger. The lab studies serve as a helpful road map by which to understand how we all feel and express emotions in the real world.

  PRONOUNS AND MISERY: THE LANGUAGE OF SUICIDAL POETS

  In any given year, over 5 percent of all adults experience a major depression. Depressive episodes are associated with physical health problems, breakdowns in people’s social and work lives, and greater risk of suicide. A number of factors are known to influence the likelihood of depression, including major life upheavals, genetic predispositions, and social isolation. One particularly prominent theory of depression argues that when people become depressed, they tend to focus on their own emotions at a pathological level. They ruminate on their feelings of anxiety, sadness, and worthlessness while paying less and less attention to the world around them.

  Recall that pronouns reflect people’s focus of attention. Given that depression causes people to look inward, it follows that a depressive episode would be associated with higher rates of self-referencing pronouns, especially first-person singular pronouns such as I, me, and my. Several studies have found this. The more depressed a person is, the more likely he or she will use I-words in writing or speaking. Most striking is that use of I-words is a better predictor of depression among college students than is the use of negative emotion words.

  Depression rates are particularly high among writers, most notably for successful poets. Recent studies indicate that published poets die younger than other writers and artists and as
many as 20 percent commit suicide. Although the job of writing poetry may be stressful, a more compelling explanation is that depression-prone individuals are drawn to writing poetry, in part, to try to understand their mood swings. This is especially true for a form of depression called bipolar depression, sometimes referred to as manic depression. Bipolar disorder is especially toxic because it has a clear genetic basis and often catapults people through extreme mood swings without any apparent cause. Unlike other forms of depression, people diagnosed with bipolar disorder are much more likely to commit suicide.

  Kay Redfield Jamison, a respected scientist at the Johns Hopkins School of Medicine, has written extensively on the close link between the artistic temperament and bipolar disorder. In her research, she finds that a disproportionate number of poets have symptoms consistent with bipolar disorder, which she discovered through their memoirs; reports from family members, friends, biographers; or the authors’ poetry. Would it be possible to identify bipolar disorder and suicide proneness through the computer analysis of the poets’ published works? Working with Shannon Stirman, who is now a clinical psychologist, we examined the published poetry of eighteen poets, nine of whom committed suicide. We discovered that suicidal poets used far more I-words in their poetry than nonsuicidal poets. Particularly striking was that the two groups of poets did not differ in their use of negative emotion words. Although this was a small sample of poets, the effects were statistically impressive.