The Secret Life of Pronouns (9 page)

Statistically, these are very large effects. The patterns are identical for males and females, and younger and older applicants. Also, these are all smart and motivated students. Why are there social class differences in function word use? Clearly, something about the students’ upbringing and life experiences is influencing their word choices.

One promising idea is that there are differences in language use within families of differing social classes from the beginning. There is some preliminary research to support this idea. Perhaps the most cited and overinterpreted study was one done by Betty Hart and Todd Risley in the mid-1980s. In their study, thirteen professional families, twenty-three working-class families, and six families on welfare were tape-recorded in their midwestern homes for an hour once a month for over two years. When the study started, most of the children were under a year old. The tape recordings were carefully transcribed and all the words of the children and adults were counted.

The authors discovered that children in the poorest families were exposed to fewer than half as many words as those from the professional families. Those in the working-class families were in between. By the end of the study, children in the professional families had vocabularies (based on the recordings) that were twice as large as those of the welfare families’ children. On the surface, this was an impressive study. However, even the authors agree that the results should be interpreted with caution. For example, there were only six families in the lowest social class group. In addition, all the professional families had at least one parent who was a faculty member and almost all were white. In all likelihood, the professional families felt quite comfortable having students in their house every month recording what they said. The welfare families, on the other hand, were all African-American and likely had a very different view of the study than those in the professional families. For example, the lower-class families may well have been more suspicious of the experimenters and may have avoided talking on the recording days.

It is likely that the verbal experiences of children brought up in lower-class families are different from those brought up in wealthier, more highly educated families. Why this might result in different patterns of function word usage may be related to issues of power and status.

SPEAKING OF SEX, AGE, AND SOCIAL CLASS: THE SOUND OF POWER

In this chapter, two groups of words repeatedly emerge. The first, which we will call the
noun cluster
, includes articles, nouns, prepositions, and big words. The second will be referred to as the
pronoun-verb cluster
. It is composed of personal and impersonal pronouns (especially first-person singular), auxiliary verbs, and certain cognitive words frequently linked to hedge phrases. Men, older people, and higher social classes all use noun clusters at high rates; women, younger people, and lower social classes use pronoun-verb clusters at high rates. Most simple hypotheses to explain these differences fall apart at some point. For example, one might argue that people who are more verbal might be drawn to, say, pronoun-verb words. This might be true for women and older people but not lower social classes. Perhaps those high in noun clusters simply read more and are exposed to more words—again, this might work for older people and higher social classes but not men.

The simplest explanation is that people higher in power and status are drawn to noun clusters and people lower in power and status rely on pronouns and verbs. For the sake of argument, why would those with more authority use more noun-related words? And, equally important, why would lower levels of power be associated with the use of more verbs and pronouns?

One solution is to consider what power and status buy. Adam Galinsky, a researcher in the Kellogg School of Management at Northwestern University, has conducted a number of studies where people either think they have power in a group or think they don’t. If they believe they control their fate, they are much more likely to make decisions on their own and ignore others’ ideas. Those with less power are easily swayed by others. In short, if you don’t have power in a situation, it is in your best interests to pay attention to others. But if you are the boss, you should pay close attention to the task at hand.

As noted in the last chapter, when people are task focused, they don’t pay attention to themselves. Most tasks, in fact, require a clear awareness of the objects, events, and concrete features that are necessary to accomplish the task goals. Much like the two people with the artificially high testosterone levels, task-focused people are able to make decisions without human relationships getting in the way.

While signaling less status, the use of pronouns and verbs also suggests that speakers are more socially oriented. Most pronouns are, by definition, social. Words such as
we
,
you
,
she
, and
they
tell us that the speaker is aware of and thinking about other human beings. First-person singular pronouns are a bit different in that they signal self-attention. Indeed, in later chapters, we summarize evidence suggesting that humans (and nonhuman mammals) tend to pay more attention to themselves in settings where they are subordinate to more powerful others. Why verbs in general and auxiliary verbs in particular may signal lower power is not entirely clear. Auxiliary verbs (such as
is
,
have
,
do
) are generally part of a passive way of speaking: “He was hit by the ball” as opposed to the more active “The ball hit him.”

The similarity in word use across age, sex, and social class is only a small part of the puzzle of function words and social psychology. Being a woman or a man, young or old, rich or poor is only a small part of our identities. We have just glimpsed the language of who we are. In the next chapter, we dig more deeply into the individual psychology of each of us.

CHAPTER 4

Personality: Finding the Person Within

W
HO ARE YOU? You, yes, you—the one reading these words right now—are the sum of your traits, values, stories, skills, job, friends, relationships, material possessions, and, of course, the words you use. Stop for a second and think how you would describe yourself to a stranger. Be careful, however. In this chapter, you will see how your choice of function words reveals more than you might guess.

Most online dating services encourage their members to write something revealing about themselves on their websites. As it happens, my students and I have collected a few thousand such self-descriptions in our research. Some examples:

I’m a nice person. Very down to earth. Drama-free. Hard worker. Caring. Honest. Sympathetic. Supportive. Intuitive. Inquisitive. Curious. I’m not superficial. I don’t keep superficial relationships. Got great friends and a good life.

—Juan, 27-year-old male

I have a great build. I am a good cook, have a fantastic job, home owner, exceptional morals, good decision maker, non smoker, no baggage and marriage minded. I love biking, kayaking, swimming, jet skis, hiking, snowboarding, rollerblading, horseback riding. I also have a great interest in music, movies, plays, comedy clubs etc.

—Marcus, 39-year-old male

Father of two girls. I have a house, truck, dogs, kids. like to be outside and love football and my girls.

—Tony, 31-year-old male

*            *            *

I am a romantic, enthusiastic, passionate and fun-loving woman. Immensely important to me are family, friends, kindness, integrity and especially laughter. I balance my dedication to my work with my spiritual life through daily meditations and a healthy lifestyle. I celebrate every day.

—Margaret, 53-year-old female

I consider myself a very eclectic individual. I can have fun almost anywhere and will try anything once … (except maybe country line dancing). I’ve had some unique experiences in my life and will continue to explore what life has to offer. I dig antiques, knitting and crocheting, and my 13 year old car that I refuse to let go despite the fact that my window is temperamental and only rolls down when it feels like it and I have to tape some wires together for my brake lights to work. I’m a makeup artist so naturally there’s a girly girl side to me.

—Gigi, 31-year-old female

The people who know me best consider me to be a warm-hearted, generous friend, independent, resourceful, but a bit stubborn at times.

—Mirah, 34-year-old female

Reading personal ads is a little like watching parts of movies. Each description tells a story about the writer both in
what
is said and in
how
it is said. For example, compare the first three men. Juan, the twenty-seven-year-old, describes himself in traditional personality dimensions that are inherently social—honest, sympathetic, supportive. And then he ends with a factual statement that he has great friends. The thirty-nine-year-old Marcus describes himself by his actions and possessions. There is no sense of his emotions or connections with others. But he does have a great build. The description by the thirty-one-year-old Tony is oddly poignant. In a handful of words, he conveys a socially isolated life with his few possessions, his dogs, but, most important, his daughters.

The ads from the three women are equally revealing. Margaret, the fifty-three-year-old woman, says less about her personality traits than about the central values that guide her. Gigi is a storyteller. You know on your first date that she will talk about some of her unique experiences. Even the eccentric story of her malfunctioning car window is part of who she is. And the last woman, Mirah, may not even trust her own views of herself and so tells the readers what other people think of her personality. Even though all of the men and women on these dating sites are consciously marketing themselves, we still snatch a glimpse of who they are.

In addition to revealing something of each person’s identity, the different ads point to the strikingly different ways that people think about personality. The very essence of some people can be their possessions, their occupations, their skills, their traits or dominant characteristics, their values, or their stories.

If you look back at the six personal ads, you will see large disparities in the function words that each person uses. The usual sex differences emerge—women use far more pronouns (especially I-words), auxiliary verbs, and cognitive words. Men use more articles and nouns. But there are large differences even within the men’s and women’s descriptions. For almost every function and emotion word category, each person has a unique pattern of word use. It is these different patterns that provide insights into their worlds.

In all likelihood, those who posted the ads went on dates. They probably shared more about who they were and, in return, expected their dates to divulge information about themselves as well. During the dates, they undoubtedly talked about their daily lives, the recent weather, food, clothes, or other topics that grabbed their attention. Similar discussions can happen on job interviews, when meeting an office mate for the first time, or simply talking to someone at a party. We directly and indirectly share and seek information about personality all the time. One might think, then, that modern-day psychology would have some fast and efficient ways to categorize people’s stories about themselves. One would be wrong.

Studying people’s personalities by listening to what they say is enormously difficult. Everyone’s stories are different in substance and style. In online dating, some write for pages, others include only a sentence or two. How can you compare one person who just lists his occupation and possessions with another who lays out his life goals? How does one even go about organizing and categorizing people’s self-descriptions? Historically, researchers have hired an army of raters to read each self-descriptive essay and then make judgments about the essay writers’ personality, writing style, goals, and interests. The process is slow, unreliable, and very expensive.

Recently, however, my colleagues and others have begun to apply some promising computer solutions to analyzing open-ended personality descriptions. As you might guess, different patterns of function words reveal important parts of people’s personalities and the ways they think.

FINDING PERSONALITY IN FUNCTION WORDS

In the early 1980s, a young graduate student in the linguistics department at the University of Southern California decided to investigate how different literary genres varied in their respective use of language. For example, do literary novels, nonfiction books, plays, mysteries, and even romance novels differ in the ways their authors employ words, grammar, and syntax? The student, Douglas Biber, happened on a statistical method not common in linguistics at the time, called factor analysis (more on that in a moment), and ultimately produced a seminal book titled
Variation Across Speech and Writing
that answered this question. English scholars, of course, had long been able to tell the difference between a mystery and romance novel—the plots were different. But Biber’s computer-based method suggested that romance novels were more personal, as measured by pronoun use, and relied more on present-tense verbs than mysteries and other genres. In fact, he found that virtually every genre had its own unique linguistic profile.

I came across Biber’s book in the mid-1990s and was fascinated by this thinking. Although he focused on ways to distinguish great and not-so-great literature by analyzing parts of speech, there was no reason that someone couldn’t distinguish people in the same way by looking at their writing. About this time, I teamed up with a former colleague of mine, Laura King, from the University of Missouri, who is among the most creative personality psychologists in the world. Whereas Biber analyzed words in books to distinguish genres of literature, Laura and I wondered if we could use people’s everyday writing to distinguish genres of people. That is, do different writing styles reflect different personalities?

The first step was to get hundreds, even thousands of writing samples from people who were all writing on the same general topic. By good fortune, I routinely teach a large introductory psychology class every year. At the beginning of each semester, my students are introduced to the nature of consciousness and thinking. One of the pioneers of modern-day psychology, William James, wrote extensively in 1890 about the nature of thought—likening it to a stream. One thought or feeling naturally leads to another, which then leads to another. Each thought or sensation stimulates the next, which may be completely unrelated to a thought that appeared a minute earlier or might occur a minute later. Further, as James noted, as our minds float down the stream of consciousness we can be aware of only a single thought at a time.

Since James, many researchers have experimented with stream of consciousness by having people either talk or write continuously about their thoughts as they occur. For many years now, I have required my students to track their thoughts in writing for twenty minutes to give them a better understanding of William James and the nature of consciousness. Although the exercise is required and students submit their writing online, their writing is not graded. We now have almost eight thousand essays that students have given us permission to use for research.

To appreciate stream-of-consciousness writing, imagine how you might write if you were tracking your own thoughts and feelings. The instructions are simple—simply write continuously about your thoughts and feelings as they occur to you. Don’t stop, just write. Here are two fairly standard examples from college students:

The refrigerator is making this funny noise. People are walking down the hall talking loudly. Someone is playing their music very loudly. My stomach is growling. I think I want a snack. My head is itching. My mouth is dry. I want something to drink. I need some more bottled water in my room. I need to repaint my finger nails. The color is chipping off. My ring on my finger needs to be sized, it is too big for me.

Or another:

I am not feeling well. I wonder if Nick got home safely. Maybe I will call him and see if he’s okay. I’m nervous about the psychology test. I should have read the chapters tonight but I am sick. I miss home. I miss my family and my dogs. Buzz is so cute and Red always barks but I still love her. I hope I’m not sick when I have to take this test on Thursday.

In both of these examples, people are writing in the here and now. That is, they report their experiences in the immediate present without much structure or analysis. We often refer to this type of thinking as an example of
immediacy
. We all think like this occasionally. And there are times we even talk like this—perhaps when tired or simply chatting mindlessly with a close friend—using small words, the present tense, and high rates of personal pronouns, especially I-words.

Across hundreds of essays, we separated out the various classes of function words, including personal pronouns, articles, prepositions, conjunctions, and auxiliary verbs. We then used a statistical method called factor analysis to see what clumps of function words emerged. With the stream-of-consciousness essays, three factors appeared that pointed to very different writing styles:
formal
,
analytic
, and
narrative
. In fact, no matter which texts we analyze, we generally find the same dimensions within almost any genre of writing, including similar types of literature, song lyrics, college admissions essays, or suicide notes.

FORMAL THINKING

The most consistent function word factor that always emerges,
formality
, often appears stiff, sometimes humorless, with a touch of arrogance. Formal thinking can be thought of as the opposite of immediacy. Highly formal (or low-immediacy) thinking and writing typically include big words and high rates of articles, nouns, numbers, and prepositions. At the same time, formal writing has very few I-words, verbs (especially present-tense verbs), discrepancy words (e.g.,
would
,
should
,
could
), and common adverbs (
really
,
very
,
so
). Here is an example of someone who the computer identified as very high in formal thinking:

I hear the sound of sandals scuffing the ground. Vague sounds of the television and radio blend into an almost incoherent hum and beat. A monolog comes onto the television in the living room. Hum -the sound of the TruAir air purifier catches my attention amidst the bath of sounds in the apartment room.… Cosmic riddles come to my mind, why does the universe have entropy?

Compared to the examples of immediacy, the high-formality writer is much more intellectual and a bit distant. There is a sense that he is putting on a serious performance. Interestingly, Biber’s first factor in his genre analysis was also formality. Academic writing and general nonfiction tend to be highly formal compared with romance novels, which are high in immediacy.

Formality in writing and speaking is related to a number of important issues. Those highest in formal thinking tend to be more concerned with status and power and are less self-reflective. Compared to the less formal writers, they drink and smoke less, are more mentally healthy, but also tend to be less honest with themselves and others. As people age, their writing and speaking styles shift from more immediate to more formal. In other words, the first dimension of stealth words has tremendous social and psychological implications.

ANALYTIC THINKING

The factor of analytic thinking identifies people who work to understand their world. The hallmark of analyzing is making distinctions. These distinctions could be between what people did and what they didn’t do, which part of the test they passed and which part they failed. Words that contribute to analytic thinking include exclusives (
but
,
without
,
except
), negations (
no
,
not
,
never
), causal words (
because
,
reason
,
effect
), insight words (
realize
,
know
,
meaning
), tentative words (
maybe
,
perhaps
), certainty (
absolutely
,
always
), and quantifiers (
some
,
many
,
greater
). An example of high analytic thinking: