Diversity Without Silos

Summer 2014

By Scott E. Page

“For the group, as well as for the species, what gives an individual his genetic value is not the quality of his genes. It is the fact that he does not have the same collection of genes as anyone else. It is the fact that he is unique. The success of the human species is due notably to its biological diversity. Its potential lies in this diversity.”
— Fran├žois Jacob, Nobel Laureate

Our students drink from two separate streams.

When studying diversity in political science, sociology, and history, they learn about demographic trends, changing notions of identity, and attempts to overcome historical injustices. They learn to view human differences through historical, legal, moral, and ethical lenses. These teachings stress normative themes: fairness, equality, justice, and inclusion.

The schools that they attend “celebrate” diversity with special days, weeks, and months dedicated to particular people and peoples and their accomplishments in spite of innumerable constraints. Such events have the stated goals of promoting a greater understanding and appreciation of differences and helping to build a more inclusive and just society. These celebrations align with the lessons learned in most social science diversity curricula.

When these same students confront the topic of diversity in science classes, they approach it from a different standpoint. In biology, they learn how genetic and phenotypic diversity within evolutionary systems underpin adaptation and speciation. In ecology, they read about how species diversity promotes ecosystem robustness. In chemistry, they learn how fewer than 100 stable elements combine to produce an endless diversity of compounds — from the water we drink to harmful concoctions such as DDT. In mathematics, they learn the value of diverse problem representations — why, for spherical problems, one should use polar coordinates, but in more rectangular domains, one should stick to a Cartesian representation. And in engineering classes, they learn how the specialization of skills and knowledge speeds learning and drives technological growth.

These two streams can become one. Indeed, we should encourage a confluence of the normative historical and social lessons of diversity together with the more positive scientific curricula. By creating this confluence, we will contribute to the advancement of our society and change for the better how students measure and value themselves. (As a by-product, we may also alter how their schools admit students and hire and promote staff.)

Allow me a didactic digression. One hundred years ago, scientific management, what many call Taylorism (after mechanical engineer Frederick Winslow Taylor), was all the rage in American business. Advocated by industrial giants such as Ford, Carnegie, and Rockefeller, Taylorism was based on the idea that “if you could measure it, you could manage it.” Through measurement and trial and error, one could determine the optimal size of a shovel, the optimal speed of an assembly line, and even the optimal temperature for work.

Taking this logic to its natural conclusion, one could even estimate the value of a worker. Imagine workers placing rivets in sheet metal, pounding spikes, or turning bolts. Time-study experts could monitor workers and calculate a worker’s piece rate. This rate translated directly into a value added for the company. Making these calculations required only counting the rivets nailed, spikes pounded, or bolts turned.

For separable tasks, one could then compute the total value of all employees as the sum of the values of the employees. This approach proved useful (and therefore took on legitimacy) because, in the Industrial Age, we lived in an additive world. The whole was, to a first approximation, the sum of the parts.

Let’s now return to the present and imagine a group of scientists working on a perplexing problem in biochemistry or neuroscience. Because we have developed methods for measuring aptitude, we could assign a number — an SAT or GRE score, or an IQ — to each scientist. But these ability numbers would not explain much about the success of the group. Unlike the work of riveters in an early 20th-century factory, the productivity of a team of scientists today is not additive in the team members’ abilities. Instead, the scientific team’s collective ability depends on the analytic and technical tools that they possess, their background knowledge, and whether they can work together effectively.

By the same logic, an individual’s contribution in almost any field today increasingly depends not on some abstract ability measure, but on his or her specific individual skills and knowledge, how those skills and knowledge contribute to the collective diversity of the group, and whether he or she can function as part of that group.

More and more, work is something that is not done alone. It is done in teams and groups. So when members of the current generation head out into the world eager to make a difference, they will find that their contributions will most likely not be individual accomplishments but as members of groups. This will be true whether they become chemists, marketers, social workers, system analysts, or teachers.

The movement toward working in groups has two root causes. First, the challenges that we confront have become more difficult and more interdependent. The fact that we’ve already solved most of the easy problems and that machines now do much of the routine work explain the first trend. The growing interdependence is caused by the increasing complexity of the modern world. What was once primarily a straightforward business or governmental decision may now include serious ethical, ecological, political, and sociological dimensions. What was once thought of as a problem in neurology may now require knowledge of chemistry, physics, and networks.

The second driving force behind the growth of group work is that groups, more often than not, outperform individuals. Empirical evidence demonstrates rather convincingly that, when it comes to solving hard problems, teams of people with diverse skills outperform individuals. To take one example, Brian Uzzi and Ben Jones from Northwestern University have found that team-authored academic papers are more than four times as likely to have a large impact on a given field of study than individually authored papers.

Open any academic journal and the explosion in team-based problem solving will be evident. My University of Michigan colleague Joshua Berke recently published a paper titled “Control of Basal Ganglia Output by Direct and Indirect Pathway Projection Neurons” in the Journal of Neuroscience. He had four coauthors.

Four coauthors may seem like a bunch, but it may be fewer than the norm in his discipline. The November 20, 2013, issue of that prestigious journal contained 28 articles with a total author count above 150. A featured article — one that the editors highlighted as exceptional — had 21 authors!

The fact that many of the challenges we face now outstrip the abilities of any individual in no way implies that individual ability does not matter. To the contrary, in many fields, individuals need to be more capable than ever before. But individual ability has to be put to use acquiring a set of tools and a base of knowledge that will contribute to our collective diversity. Two heads can be better than one only if their contents differ. For that reason, our future challenges demand able and diverse minds.

These diverse minds will belong to people with diverse identities. Working groups of the future will be even more demographically diverse partly due to changes within the United States and partly because of increased globalization of the economic, governmental, and nonprofit sectors. Functioning in a group will therefore require being able to work with people who think differently and look differently. (Note that what I call a group need not be 10 people in a room or a lab that identify as a team. Many of these groups will interact over space and time using ever more sophisticated technology.)

Accept, if you will, that we are moving toward more group and collaborative work. I want to return to the core point of this article: the importance of the confluence of our two types of teaching about diversity. The study of diversity can be and is carried out on a scientific basis. Diversity improves system robustness. By that I mean, it enables systems to better withstand external pressures by providing multiple responses. Engineers teach this as Ashby’s Law of Requisite Variety. Put simply, if you have only a hammer, you cannot fix a leaky pipe. If you have a full toolkit, you’re more likely to be able to respond to any emergency.

Diversity also speeds adaptation by providing more immediate alternatives as well as more possibilities for recombination into innovative new solutions. I do not mean to imply that more diversity is always better. We can have too much, producing instability. And, we can have the wrong type of diversity for a specific function. Herein lies my main point: These scientific understandings also apply within the social realm.

Consider an example: We know that phenotypic variation contributes to species robustness by speeding adaptation. A similar logic suggests that cognitive diversity should speed learning in contexts that change quickly. Or, just as those 100 or so stable elements produce an incredible array of chemicals, so too can a few hundred basic scientific and engineering breakthroughs be combined to produce a near endless variety of innovative products.

Science therefore suggests the possibility of leveraging our differences for our collective benefit. To do so, we need to apply that science. One of the first lessons is that we need a new accounting. In the past, the additive world, numerical integration — Is each group fairly represented? — emerged as a natural benchmark. If one assumes that identity differences do not correlate with individual talent — be that at riveting, balancing ledgers, or solving math problems — then a lack of even representation suggests discrimination either implicit or explicit.

The new complex world demands deeper integration. Representative numbers alone do not suffice. The diverse minds need to be able to absorb the ideas of others. As in an evolutionary system, better ideas (a crude analog of species) need to be identified and selected with greater likelihood. And applying the logic of chemistry, group members must be able to combine those ideas to create a whole that is more than its parts. This requires more than equal representation or even inclusion. It requires deep engagement.

As mentioned, this deeper embracing of diversity disrupts current thinking about individual ability. At present, students confront numerous tests of their abilities. The resulting cardinal orderings implicitly reinforce the additive, ability-centric worldview. I’m a 1320. She’s a 1450. He’s a 1040.

What matters most today is how one merges one’s skills, knowledge, and abilities with the skills, knowledge, and abilities of others.

Yes, he may score 1040, but that is a mere reflection of a potentially useful set of skills, knowledge, and abilities. What matters most today is how one merges one’s skills, knowledge, and abilities with the skills, knowledge, and abilities of others. To extend the logic of Fran├žois Jacob (see quote at beginning of the article) into the social world: The child’s potential contribution to society depends not on his quality, but on his unique combination of talents and his opportunities to apply them in concert with the unique tools of others to improve the world. The future success of our human species will be due notably to our cognitive diversity. Our potential lies in this diversity.

Scott E. Page

Scott E. Page is professor of complex systems, political science, and economics at the University of Michigan and an external faculty member at the Santa Fe Institute. He is the author of The Difference: How the Power of Diversity Creates Better Groups, Firms, Schools, and Societies, Princeton University Press, 2007. He teaches a free online course called Model Thinking, offered through Coursera (www.coursera.org).