Language Log

« previous post | next post »

Yesterday the NYT had a feature on Jean Twenge's work — Douglas Quenqua, "Seeing Narcissists Everywhere", 8/5/2013 — and also "A Back and Forth About Narcissism":

The social-science journal Emerging Adulthood recently invited Jean M. Twenge and one of her most prominent critics, Jeffrey Arnett, to debate “whether today’s emerging adults are excessively ‘narcissistic,’ ” as Dr. Twenge asserts. Both wrote papers outlining their positions, then each wrote a reply to the other.

The journal issue is Emerging Adulthood 1(1), and the NYT's "Back and Forth" consists of quotations from some of the papers. Here are the authors, titles, and full abstracts of the four papers in this interesting exchange.

Jeffrey Jensen Arnett, "The Evidence for Generation We and Against Generation Me":

This article addresses the question of whether today’s emerging adults are excessively “narcissistic” as claimed by Jean Twenge and others. The answer is a decisive “no.” There is no persuasive evidence that scores on the Narcissistic Personality Inventory (NPI) have risen among college students in recent decades. In any case, the NPI is a dubious measure of narcissism, and college students are a dubious sample of emerging adults. There is evidence that today’s emerging adults have high expectations and are confident in their abilities, but these qualities should be seen as psychological resources during a life stage that is often difficult. If it were true that their narcissism was rising and that narcissism leads to impulsive behavior, then impulsive behavior would also be rising, but in fact it is diminishing across multiple indicators. As a society, we can and should do more to support emerging adults, beginning with a halt to the negative stereotypes.

Jean M. Twenge, "The Evidence for Generation Me and Against Generation We":

According to the empirical evidence, today’s emerging adults (Millennials/GenY, born after 1980) are more Generation Me than Generation We when compared to previous generations. Five data sets show a generational increase in narcissism, including one that demonstrates significant increases when a confound is controlled. College and child samples increase in self-esteem over the generations. Some high school samples show no change, though high school students increasingly embrace other overly positive self-views. In nationally representative samples of high school and college students, values have shifted toward extrinsic (money, fame, and image) concerns and away from intrinsic (community, affiliation) concerns. These trends have mostly negative consequences, such as lower empathy, less concern for others, and less civic engagement (e.g., interest in social issues, government, and politics). Parents and teachers should focus on teaching children and adolescents the values of hard work and consideration for others instead of an inflated sense of self.

Jeffrey Jensen Arnett et al., "The Dangers of Generational Myth-Making":

We respond here to Twenge’s article “The Evidence for Generation Me and Against Generation We.” With regard to the question of whether “narcissism” is increasing among emerging adults, flaws are identified in the studies she used to make her case, and counterevidence is presented. We show that for the most part emerging adults’ values have not changed in recent decades, but to the extent that change has occurred, it has been in the direction of less selfishness and more engagement in global issues as well as greater desire to ameliorate problems in the community and the world. Finally, we emphasize the duty for scholars to avoid contributing to unjustified negative stereotypes about young people that lead others to have contempt for them and refuse to support their efforts to make their way into adulthood.

Jean M. Twenge, "Overwhelming Evidence for Generation Me":

Documenting trends in young people’s self-reported traits and attitudes is empirical research, not a complaint or a stereotype. Rising cultural individualism has both good consequences (more gender equality) and more negative ones (narcissism, mental health issues). Arnett seems to believe we should embrace studies of cultural change only if they find positive trends. A total of 11 studies show a generational increase in narcissism, 7 using noncollege samples. They include respondents from high school age to adults, four different ways of measuring narcissism, three different research methods, four different ways of recruiting respondents, three different countries, and eight sets of authors. Eleven additional samples show increases in positive self-views. Perspective taking, empathy, and concern for others have declined, not increased. Narcissism is not related to teen pregnancy or car accidents. An enormous body of research finds generational increases in anxiety, depression, and mental health issues, most in noncollege samples.

I don't have anything to say about the Narcissistic Personality Inventory, and I have no opinion about whether studies of cultural change ought to look for positive or negative trends. But last summer, I took a careful look at one of Dr. Twenge's papers that drew conclusions from historical trends in word counts (Jean M. Twenge et al., "Increases in Individualistic Words and Phrases in American Books, 1960–2008", PLoS One 7/10/2012):

"Textual narcissism", 7/13/2012
"Textual narcissism, replication 2", 7/14/2013
"It's all about who?", 7/31/2012

And I found no support for Twenge et al.'s main conclusion, which they stated this way:

This study demonstrates that language use in books reflects increasing individualism in the U.S. since 1960. Language use in books reflects the larger cultural ethos, and that ethos has been increasingly characterized by a focus on the self and uniqueness.

Instead, my replications suggested a very different story, with the words on Twenge et al.'s "individualistic" and "communal" lists showing a striking increase in "communal" references during the first half of the 20th century, and no meaningful overall change since then.

My impression of  that PLoS One paper was that the authors went into the study with their conclusion already determined, and engaged the data only long enough to find a (very thin) argument for the result that they wanted.

And my (perhaps naive and old-fashioned) reaction is that this is appropriate behavior for lawyers and politicians but not for scientists. I say this without any intent to insult lawyers — the basis of our legal system is the idea of adversarial argument, where the advocates for each side make the best case that they can, without actually lying or fabricating evidence. And political success, whether for people or for ideas, seems increasingly to require single-minded support for one side of an argument, exaggerating its positive features and ignoring, excusing, or even covering up its negatives.

I recognize that scientists have opinions, and that research is often motivated by the desire to support already-held beliefs. But the researchers that I respect are at least as skeptical about their own ideas as they are about the ideas of others.

Update — See Brent W. Roberts, "Science or law: Choose your career", 7/23/2013, for an extensive and thoughtful discussion of these issues.

August 6, 2013 @ 4:37 am · Filed by Mark Liberman under Computational linguistics

Permalink