Whistling Vivaldi, by Claude M. Steele

Amazon link

Subtitled “How stereotypes affect us and what we can do”, this is one of the most powerful books I’ve read in a while. Diversity issues are everywhere these days, and they seem hopeless to overcome – when looking at a problem like why so few women are in technology, people argue about whether it’s a pipeline issue (not enough girls and women being interested in tech), a retainment issue (doing a better job of supporting women once in the tech world), or pure prejudice (outright sexism by men in the tech world), but all of these require sweeping changes to improve. Whistling Vivaldi shares decades’ worth of experiments that pinpoint a primary issue that minorities face in such situations. And armed with that knowledge, Steele goes on to make specific recommendations how to address that issue.

Steele does not discount the existence of outright prejudice as being an issue that minorities face. He cites his first experience of “race” as a construct – when he was a kid, he was told that he could only go to the local public swimming pool in Chicago on Wednesday afternoons. He had not previously been aware of being black, but this restriction on what he could do illuminated it. As he puts it, “This is how I became aware I was black. I didn’t know what being black meant, but I was getting the idea that it was a big deal.” He became a social psychologist, and this book recounts his work and the work of others in trying to understand what impedes the performance of certain groups in stereotyped situations.

First, he explains the idea of stereotype threat via experiments. In one experiment, Princeton students were asked to play a round of miniature golf in the laboratory. Some students just played the course. Other students were told in advance that it was a test of “natural athletic ability”. White students who were told it was a test of “natural athletic ability” did 3 strokes worse over 10 holes than did white students who were told nothing. Black students had no variation in their scores whether they were told or not. The researchers hypothesized that students were putting more pressure on themselves because of the societal stereotype that whites have less athletic ability such that “their frustration on the task could be seen as confirming the stereotype, as a characterization both of themselves and of their group. And this, in turn, might be upsetting and distracting enough to add an average of three strokes to their scores.”

The researchers then tried to figure out how to reverse the situation, with the idea that “it should be possible to set up a stereotype threat that would interfere with black students’ golfing as well. All they’d have to do was represent the golfing task as measuring something related to a bad stereotype of blacks. Then, as black participants golfed, they’d have to fend off, like whites in the earlier experiment, the bad stereotype about their group. This added pressure might hurt their golfing.” So they told the next batch of students that the golfing was a test of “sports strategic intelligence”. And, voila. The results were the exact opposite of the earlier experiment – blacks who were told that scored 4 strokes worse compared to blacks who weren’t. Whites had no variation in their scores whether told or not. Same task. Similar students. Vastly different results depending on whether the instructions tapped into a negative stereotype that the students felt applied to them. This is what Steele means by a stereotype threat.

Steele’s insight is that we are all part of a larger society which has widely held stereotypes. When there exists a negative stereotype about some aspect of a person’s identity, that person feels pressured by it – they want to prove people wrong and show that it doesn’t apply. But that additional pressure is a drain on resources. When that person experiences a setback or frustration, they get distracted by thoughts of whether they are proving the stereotype in that moment. This additional cognitive burden is significant even in the absence of overt prejudice.

As Steele looked to test this hypothesis, he looked at the negative stereotype “that sex differences in math and science achievement were substantially rooted in sex differences in a genetically based capacity for math.” He and his collaborators looked for an experiment to test whether the performance difference was due to the stereotype threat that women felt about having to prove that stereotype wrong. Their ingenious idea was to give a difficult math test, but explain that “on this particular test, women always do as well as men.” And indeed, “among participants who were told the test did not show gender differences, where the women were free of confirming anything about being a woman, women performed at the same high level as equally skilled men. Their underperformance was gone.

This is not to say that stereotype threat explains all differences. These experiments were intentionally designed to remove most differences – so they were looking at similarly skilled people with similar backgrounds, where the only difference was the identity aspect under investigation. There are still many societal structures that contribute to differential performance more generally, including differential access to education and resources. However, stereotype threat is sort of the cherry on top – strong people that manage to fight through all of those disadvantages to get into top schools still have this remaining challenge to overcome.

The rest of the book goes into the detailed experiments that Steele and other social psychologists have done to explore how stereotype threat works. One set of experiments show that stereotype threat manifests itself as anxiety, including high blood pressure, a racing mind, difficulty focusing, and reduced working memory. The person may not even consciously be aware of being anxious, but the effects were clearly measurable in the lab. These direct physical consequences of stereotype threat have a meaningful effect, and can even lead to a shorter life due to the high blood pressure and anxiety.

So now that Steele identified stereotype threat as a major factor in the differential performance of identity-based groups, what can we do about it? Steele continues: “the next idea [was] that what determines how much identity threat a person feels in a setting are the cues in the setting that might signal these contingencies.” The rest of the book is about how we can set about removing or reducing those cues “because cues and contingencies are things that, at least some of the time, you can change. You can get your hands on them, and you can shape how people think about them. If identity threat were rooted in an internal psychological trait, a vulnerability of some sort, then it would be harder to remedy.”

The most powerful cue, of course, is whether there are others “like me” in an organization. Steele calls this “critical mass”. It is not a specific number, as it is highly situation dependent. Steele gives the example of Sandra Day O’Connor on the Supreme Court. As long as she was the only woman on the court, every decision she made was examined carefully and assumed to be guided by gender stereotypes. Once Ruth Ginsburg was added to the court, all of the additional pressure of being a female on the Court went away. Steele theorizes that this observed effect is why O’Connor voted in favor of affirmative action – she understood the additional pressure of being an exception in a situation, and that “a critical mass of minority students [was] essential to these students’ ability to function and learn in a university environment”.

Another cue is how critical feedback is given. If not delivered carefully, critical feedback can be seen as evidence of unconscious bias, of believing that the student can’t do the work. At the same time, any overt acknowledgement of identity can be seen as pandering. An experiment showed that the most effective form of feedback was the following. “The feedback giver explained that he “used high standards” in evaluating the essays … Still, he said, having read the student’s essay, he believed the student could meet those standards. His criticism, this form of feedback implies, was offered to help the student meet the publication’s high standards. … Why was it so effective? … It told them they weren’t being seen in terms of the bad stereotype about their group’s intellectual abilities, since the feedback giver used high intellectual standards and believed they could meet them.”

Another effective intervention was to facilitate intergroup conversations among students from different backgrounds. He found that minority students at top universities were so afraid of confirming stereotypes that they didn’t deserve to be there that they refused to ask for help from professors or other students – they locked themselves into their rooms and worked harder and harder. But by cutting themselves off from the support system, they were working uphill. By fostering discussions, and discovering that students of all backgrounds were struggling to make the transition to college, it reduced their anxiety that they were confirming the negative stereotype.

And, unsurprisingly, Steele references the work of Carol Dweck on the growth mindset. In particular, he cites the challenge of facilitating those intergroup conversations, as both sides are wary – one side of being judged prejudicially, the other of being judged as prejudiced (e.g. racist). “When it is identity threat that keeps people apart and uncomfortable with each other”, we can’t have those conversations and start to work past prejudice. Steele found, though, that when those conversations were framed with a growth mindset, students were more willing to participate, as “with a learning goal, mistakes become just mistakes, not signs of immutable racism”.

I highly recommend this book to get a better sense of some of the challenges that minorities face, and some ideas of how they can be addressed. There are so many more great anecdotes and experiments in the book that I have not shared, so please go read it, especially if you are in a position of power where you can effect some changes about the cues shown by your organization.

P.S. One aside that I quite liked was Steele’s observation that identity is often claimed precisely as a result of negative stereotypes. “… the kind of contingency most likely to press an identity on you is a threatening one, the threat of something bad happening to you because you have the identity. You don’t have to be sure it will happen. It’s enough that it could happen. It’s the possibility that requires vigilance and that makes the identity preoccupying.” Interestingly, this doesn’t happen as a result of positive stereotypes, because getting preferred treatment doesn’t register with us. “I might not notice I had an advantage. I might assume I was evaluated the same as everyone else. And not noticing my advantage, I might not become much aware of the identity on which it was based.” This is where discussions of privilege become fraught, as beneficiaries of privilege rarely notice the benefits they have received.

5 thoughts on “Whistling Vivaldi, by Claude M. Steele

Leave a Reply

Your email address will not be published. Required fields are marked *