The Corruption of our Social Learning System

why we need a new scoring system

When I propose an eight-category scoring system for the content of public intellectuals, am I trivializing a serious enterprise by turning it into a game? My answer is that we already have incentive systems for public intellectuals, and those systems have been corrupted. It’s a game now, and it works badly. No game will be perfect, but a different game would be better.

Let me start with an outline of topic sentences.

  1. We learn socially, so that most of our beliefs come from other people.

  2. This makes the problem of choosing which people to trust the central problem in epistemology.

  3. What Eric Weinstein calls our “sense-making apparatus” can be thought of as a set of prestige hierarchies, at the top of which are the people who are most widely trusted.

  4. Our prestige hierarchies are based largely on credentials: professor at Harvard; writer for the New York Times; public health official.

  5. The incentive systems and selection mechanisms in the credential-based hierarchies have become corrupted over time, allowing people to rise to the top who lack wisdom and intellectual rigor.

I think of electrons in an atom as occupying orbits relative to a nucleus. I have never observed this. I have never done any experiments that would verify this. I believe it because that is what I was taught fifty years ago by my high school chemistry teacher, Dr. Frank Quiring. I have not kept up with chemistry or physics since then.

In The Secret of Our Success, Joseph Henrich drives home the point that almost all of the knowledge that we possess comes from culture rather than from personal experience. Similarly, in Darwin’s Unfinished Symphony, Kevin Laland points out that the human species is unique in the extent to which we learn from one another. Learning comes both from informal copying and formal teaching. Formal teaching, and in particular teaching that passes new knowledge from generation to generation, are why humans are the species with culture.

Philosophers typically view the problem of knowledge, or epistemology, as one of aligning the beliefs in your mind with the “reality out there.” But because our beliefs about reality come from other people, I think that the choice of which people to trust is the core issue in epistemology. I have made this point to academic philosophers, and they blow me off, insisting that the issue of aligning beliefs to reality is the nub of the problem. Relying on “testimony” (other people’s beliefs) is just one method for trying to solve it. I think that they would say that I choose a person to believe based on how well I think that person’s beliefs align with “reality out there.” But I would counter that I choose who to believe first, and then I choose what to believe.

Children are inclined to believe what they are told from adults who care about them. I believed that Dr. Quiring cared about me, and I believed he knew something about chemistry, so I bought his story about electrons occupying orbits around a nucleus. He in turn could cite Nils Bohr, a Nobel Prize winner, and Dr. Quiring could also cite the textbook that we used. Many scientists reported experiments that supported the Bohr model. All of these testimonies gave me good reasons to believe it.

But then we run into the fact that different people tell us different things. Then we have to decide which person to believe. Suppose one economist tells me that a 20 percent increase in the national minimum wage will reduce employment by 2 million, and another economist tells me that it will not reduce employment at all. In theory, I could arrive at my belief on this issue by doing my own research. But I am not going to do that. At best, I will read a few survey papers and try to decide what the most reliable research says about the issue. The epistemological problem boils down to deciding on what basis I should trust economist X more than economist Y.

Henrich points out that humans have two types of hierarchies. In a dominance hierarchy, the people at the top gain authority by force, and the people at the bottom reluctantly obey. In a prestige hierarchy, the people at the top gain authority by earning respect, and the people at the bottom willingly try to copy and learn from those at the top.

Prestige hierarchies work through competitive mechanisms. In a democracy, people compete to win political offices. Under capitalism, firms compete for customers. In a legal proceeding, lawyers compete for the verdict. In science, researchers compete to obtain discoveries and verify hypotheses.

The value of competition in choosing the people to trust was driven home to me years ago by David Brin, in his essay on Disputation Arenas. In that essay, he offered a proposal for structured competition on the Internet to improve what I call social epistemology. Better ideas would win.

Prestige hierarchies are robust to the extent that the people who reach the top have wisdom worth sharing. These hierarchies are corrupt to the extent that people reach the top by “gaming the system,” getting ahead through ruthlessness rather than wisdom.

The term “sense-making apparatus” makes me think of the set of hierarchies that is primarily founded in academia and has its largest impact through the media, including legacy 20th-century media that have survived as well media that emerged in the present century thanks to the Internet. One sign that they are breaking down is that we see a rise in the use of dominance moves, such as censorship. Elites resort to dominance moves when they can no longer rely on prestige.

The process of getting ahead in a prestige hierarchy is analogous to the process of earning a bonus in a firm’s compensation system. If the bonus criteria align with the firm’s goals, people who do productive work will earn bonuses and the firm will be successful. If the bonus criteria are not well considered, worker who are not particularly productive will obtain bonuses, and the firm’s performance will suffer.

Bonus systems are like a game. The firm wants to get the most (useful) effort from its workers for the least compensation. Workers want to get the most compensation with the least effort.

My observation is that the longer a specific bonus system is in place, the better workers become at figuring out how to get more compensation for less effort. Incentive systems naturally degrade over time. Management has to revise the bonus systems every few years if the firm is to prosper.

Most incentive systems use a combination of formal measures (“metrics”) and informal judgment (“what your boss thinks”). Neither is perfect. Jerry Muller’s The Tyranny of Metrics describes how the formal approach often goes wrong. Informal judgment can be used as a corrective for imperfect metrics. But judgment also can introduce bias and cronyism.

Our prestige hierarchies of academia and legacy media rely heavily on credentials. Think of the process of obtaining tenure as a professor or the process of obtaining a prestigious position for a newspaper or TV network. Such credentials are awarded on the basis of judgment by incumbents. They reward conformity rather than excellence. Why this has emerged as a problem now more than in the past is a question that I am still pondering for a subsequent essay.

In any case, popular trust in our sense-making institutions has fallen dramatically over the past 70 years. The relationship between elites and the public at large in 2021 is somewhere between troubled and dysfunctional.

Many elites cannot understand why people do not trust “the science,” government officials, leading academics, or the news as reported in legacy media. But more detached observers, such as Martin Gurri in The Revolt of the Public and Yuval Levin in A Time to Build, understand that elite misconduct contributes heavily to the problem.

Our 21st-century media, as typified by Twitter, introduce metrics into the process of shaping the intellectual hierarchy. Likes, shares, and the number of followers are now the relevant metrics, and the survivors of legacy media now also compete using similar metrics.

But these metrics lead to terrible outcomes! What ends up being rewarded is tribalism of the worst sort. Nastiness, bullying, disrespect for those with differing views.

I think of Twitter as the intellectual equivalent of a village with no manners, with people constantly getting into fights and punching one another. Even worse, the rewards—the likes, the shares, the followers—go to the people who fight the dirtiest.

Imagine a tamer village, in which people are given approval for saying “please,” “thank you,” and “excuse me” instead of for eye-gouging and groin-kicking. I am proposing metrics in order to try to both tame the village and populate the sense-making apparatus with wisdom rather than bias and cronyism.