Let’s fix knowingness… but without empiricism

Look for good explantations before digging into the data.

The Dense Discovery issue 333 editorial comments on Brian Klass’ “Knowingness” and the Politics of Ignorance, a thought-provoking article with much to like—except its conclusion, which I think is misguided in its reliance on empiricism. Empiricism got us started on the path to progress—but it can’t take us the rest of the way.

In the post, Brian comments on an issue afflicting our politics and discourse that is more subtle and far more dangerous than the often-lamented polarization.

Much ink has been spilled about “polarization.” Most of it ignores a major cause: the widespread, misplaced faith that we already know that which we do not know.

Polarization is the entrenched and radical difference in opinions of sides that seem unable to find any common ground. Often, though, this inability to work together comes from a perception of already knowing everything there is to know on a given topic.

There are actually many disagreements that aren’t about divergent policy goals, but rather about the impenetrability of people’s minds. No matter the volume of ironclad empirical evidence presented, they think they already know. But you can’t solve problems if fresh facts never sway your viewpoint.

This is what Jonathan Lear calls knowingness, which Brian describes as:

a relationship to knowledge in which we always believe that we already know the answer—even before the question is asked. It’s a lack of intellectual curiosity, in which the purpose of knowledge is to reaffirm prior beliefs rather than to be a journey of discovery and awe.

Brian proposes a solution for the existential problem of knowingness.

We, as a society, would be better off if our politics could become re-centered around a collective process of searching for answers to solve problems, made possible by exploring evidence, and then learning from it.

The diagnosis—that a lack of intellectual curiosity fuels many modern issues—is spot on. But there’s one major issue in Brian’s proposed solution. You cannot solve problems by “exploring evidence, and then learning from it.”

One cannot derive correct predictions about the future from observations about the past. The classical example is that of a turkey who day after day explored the evidence and learned from it that when the farmer came he brought food—until the day before Thanksgiving.

This is the old problem of induction, which the philosopher Karl Popper discussed at length—see for example chapter one in Conjecture and Refutations.

Popper offers a better alternative to induction and empiricism for knowledge creation and problem-solving: conjecture and refutation. In short: identify a problem, guess at a possible solution, and use criticisms and experiments to refine or discard it. David Deutsch later improved upon this by identifying the role of good explanations in this process.

So our “collective process of searching for answers to solve problems” should not be an empirical exercise in looking at ever more data. Problem solving starts with identifying errors in our theories, conjecturing solutions, and testing them—not with data collection.

Evidence helps eliminate bad ideas, not create them. Only explanations can do that.

I wouldn’t be surprised if the obsession with “exploring evidence” and looking at facts contributes to the cultural issues Brian reports. When looking at facts alone, it’s possible to cherry pick those that support your thesis while ignoring the rest.

There’s much more to say about problem solving and good explanations. The best place to start is David Deutsch’s The Beginning of Infinity.

Let me repeat the Feynman quote I shared last week: “You are the easiest one to fool.” Knowing about knowingness does not make you immune to it. Being humble is a necessary requirement for problem solving, because the only way to identify and correct errors is through criticism—of our own guesses first and foremost.

So yes, let’s reframe politics as a process of acquiring knowledge, but let’s do it the way Popper taught us: through conjecture and refutation. If we want better politics—or better thinking—we must prioritize explanation and criticism over data accumulation.