This piece is about why I think the Deutschian definition of rationalism is wrong, and what I would propose as an alternative (but admittedly more vague) definition. I’m posting this first draft without edits as part of an attempt to get feedback while working on a longer piece about what I think is missing in the Deutschian worldview. Any feedback, especially from critical rationalists about what I’m missing, would be appreciated!

David Deutsch defines rationalism as the view that knowledge comes from the mind; or the view that the source of knowledge is the mind. Deutsch asserts that there are empiricists who think the source of our knowledge is the senses, and there are rationalists who think the source of our knowledge is the mind (or our reasoning faculty). Deutsch proceeds to dunk on both views, asserting that there is no source of our knowledge—knowledge arises out of ongoing conjecture and criticism.

While this dichotomy between rationalism and empiricism did exist historically, is there anyone today who seriously believes that our knowledge just “comes from” the mind or from our senses? If not, what use is there in continuing to define these words like this?

They’re very simple, and clean definitions, which is to his credit. But they’re misleading. How I might define rationalism? Rationalism is the stance that rationality is our primary mode of engagement with the world. (And I think this is a stance that Deutsch, knowingly or not, subscribes to.) I acknowledge this is much more vague. It’s a definition I would have scoffed at two years ago. But I genuinely believe it’s a better definition.

So let’s expand on the definition: a rationalist believes that most of what we do in engaging with the world is engage in rational thought. Rational thought refers to: systematic, logic-driven thought. Creating precise categories, with symbols to represent them, and performing transformations on them via logical rules. Example: mathematical proofs.

Is this what we do all the time? Or most of the time? No, it’s not. What are we doing most of the time instead? As Jake Orthwein might say, we’re engaging in meaning-making:1

But as Chapman argues, our basic way of interacting with the world is better understood as meaning-making rather than as belief-formation and plan-making to achieve goals. The modernist conceit that all meanings can be reduced to rational ones misunderstands human beings, and religion by extension.

Let’s go back to how I think rationalism views human behavior: “Creating precise categories, with symbols to represent them, and performing transformations on them via logical rules.” I grant that we do create categories in our head, but it’s not via conscious or systematic mental processes. It’s more spontaneous. You look at a chair and your mind computes “chair”, although not based on some definite checkbox of chair properties, but based on a vague resemblance to chairs you’ve seen in the past. The categorizations of the world are not perfectly rational: they cannot be perfectly rational, because this would require them to be objective. But chair-ness arises as a category not out of some objective imperative for rational correctness, but rather out of its usefulness to us as embodied beings of a certain size and shape.

An example to demonstrate this: you can imagine being shown a picture of a….chair-cat thing. Like, something that looks like a mesh of a chair and a cat. And your brain would look at that, and it would not have a clear answer about whether it’s a chair or a cat. And a critical rationalist might be inclined to say: “well, we can’t figure out what it is, but objectively it’s either a chair or a cat”, but objectively it’s not either of those things!

Image by Dall-E A chaircat, by Dall-E.

Perhaps I’m strawmanning: a critical rationalist would not necessarily say “it’s objectively either a chair or a cat”, but they would say something like “it’s objectively a specific, precise thing. it either fits into the category of cats or not; it either fits into the category of chairs or not. the categories of the world come in binaries (truth and falsehood, chair and not-chair), and although we can’t be sure that we aren’t applying the right set of categories because we’re fallible, there are some categories out there that are objectively correct and precise and which do definitively chunk the world into discrete pieces.”

But then we ask: are there really such categories? Tell me about a category—a way of dividing the world into objects—that can be defined very precisely, without edge cases, without blurry boundaries, with a specific set of objective properties that any old rationalist can assess and check off, that any two observers—regardless of their shape and size and physical make-up—would agree on.

I will make this (hopefully unnecessary) caveat: what I’m espousing is actually not relativism, and it’s not nihilism. There’s a tweet by Jake Orthwein where he says: there is an objective world, but does it necessarily follow that there are objective truths about the world?

There is an objective world, but it is not necessarily the case that there are objective, perfectly definite categories in the world. Instead, I’m more compelled by the view that categories are tenuous. They are not arbitrary: they arise out of interaction between the observer and the world they exist in. But they are not totally observer-independent either.

  1. I have realized in writing this that I really have absorbed the distinction between rationality and reasonableness that Chapman proposes, for better or worse. ↩︎