Artificial intelligence theorist Eliezer Yudkowsky defines rationality in one of two ways: epistemic rationality and instrumental rationality.
The first concerns itself with what we think of as the truth — making sure that our beliefs about reality map well to some collective or objective definition of the truth. This is generally thought of as the realm of science, but many people would argue that there are truths that science hasn’t yet realized, or can’t realize, and those matter, too. In a sense, it’s about making sure that our beliefs align with facts and their supporting evidence.
The second concerns itself with our ability to get what we want. It is about how effectively we are able to live in accordance with our values, generally as measured by how well we can meet the demands of the goals we set for ourselves based on those values. It is more pragmatic in nature. Once we pick a north star, this kind of rationality is only really measured by the degree to which we are able to meet the demands set by that north star.
Philosophically speaking, rationality is a way of thinking about the world. The first of the two — epistemic rationality — is a kind of open-ended reasoning, where there is a direct feedback loop between how we think the world works and how it actually works, and these coevolve over time, with no clear end in sight, with no clear foundation, unless we force ourselves to be certain about something that is otherwise uncertain. The second of the two — instrumental rationality — is a kind of closed reasoning, centered around a single theme or goal or value system. It, too, evolves in an open-ended way, but it is deeply anchored around one thing and built around that.
But what if rationality isn’t really about thinking at all? If we bring the experience of our body down to its core constituent to understand our relationship to the world, what we have to start with is action or movement, not thought. Although we are capable of thinking before we act, for the most part, we act before we think. And we often do so unconsciously, in small ways that we don’t consider meaningful at all.
Before I can strategize about epistemic rationality, I have to breathe, and I have to eat, and I have to orient my body in space. Before I can define my specific values and look to instrumental rationality to guide me, I have to make sure that I am out of harm’s way, that my life isn’t going to end.
From this vantage point, rationality, at its deepest core, is about survival — about life and death, not right and wrong — because our actions in the world are predominantly guided by the deep, unconscious urge to maximize our fitness. The thinking, the philosophy — that comes after, and even when it does come, it is still influenced by these deeper urges to protect ourselves or to exert ourselves — physically, psychologically — and these deeper urges cloud our subsequent thoughts and abstractions in big and small ways. In this sense, our actions are always rational; thoughts just justify them.
To add some color to this, let’s look at this from the point of view of economics. Traditionally, classical economics has assumed that people will act rationally in the market. It assumes that we are indeed deeply self-interested, concerned with maximizing our odds of surviving and thriving. Here, rationality and survival are connected to one another in a way that philosophy doesn’t consider when it talks about rationality as thought.
And yet, over the past few decades, many psychologists and economists have found that this just isn’t true. People act in seemingly irrational ways all the time. There is, for example, nothing self-interested about going deep into credit card debt, paying 20 percent in interest, just to consume something that doesn’t produce value and instead puts you in a difficult long-term situation financially. But we all do many things like this, in our own ways, over the course of our lives. Behavioral economists have all sorts of interesting names like herd mentality, narrative fallacy, and loss aversion for the many biases that cloud our judgments and make us act irrationally in the world.
They say that our brains take a host of shortcuts when making decisions about the world because, to our ancestors, these shortcuts were useful in a fast and dangerous world. Today, they are no longer useful, but they still guide our decision-making, and thus, our actions. This, of course, contradicts the classical economic definition of humans as rational actors. And if you were paying attention to the earlier definition of rationality as that which guides toward survival, where our actions are our measure of rationality, this contradiction stands, because we act in ways that harm us all the time.
Except, there is a way in which these two models can be reconciled, and that’s by taking a slightly different frame in understanding human behavior: All of our actions — including taking on needless credit card debt and other harmful behaviors — are always rational, and the only reason they look irrational is that both we and the observers judging our actions often don’t know how we have convinced ourselves that these actions are tied to our survival.
According to our unconscious individual mental model of reality, every action we take is rational, even if we feel a conscious conflict for having taken that action. If I subconsciously judge my worth according to how I compare to those around me, then it is entirely rational to get myself into crippling credit card debt to buy a bigger TV or a new car at the present moment at the expense of the future. And it’s rational because some deep part of me has associated that need for approval or validation or keeping up with something that affects my ability to survive and thrive in the world.
If we bring it back to Yudkowsky’s definition of instrumental rationality, where rationality means the ability to live in accordance with our values or to reach our personal goals in an effective way, then all of our actions are guided by an invisible value system deep in our subconscious mind, shaped by a combination of our genes and our past experience, and this is why the rationality that we associate with conscious philosophical thought patterns is almost always a post-hoc judgment of our actions rather than a calm and lucid guide navigating us through reality towards some truth.
This, of course, isn’t to say that the kind of hard thinking that aims at what Yudkowsky calls epistemic truth doesn’t have a grip on some important part of the truth. It just means that when we speak of rationality, we have to first ground it in the reality of the physical body and its survival and work backward from there, which is a bottom-up process, rather than the top-down approach of thinking things and then bringing them down to reality.
On a personal level, the only rational thing about your sense of reality is your actions because they are based on what you have unconsciously learned to associate with your survival. Those are your true values. And if those actions don’t align with what you say you believe, or the stated values that you list as being important, or the orientation it takes to look out for your own and other people’s long-term future, it means that, somewhere in your mind, there is a reason that you are unaware of making you act in self-defeating ways. And that reason can either be a conflicting value, or it can simply be a lack of knowledge about what it takes to act effectively in the world.
One of the reasons, for example, that mindfulness and many similar Eastern meditative practices work in helping people figure themselves out is because most of them are focused on watching the body. That is all meditation really is. You sit somewhere, and in stillness, you watch the subtlety of how your body moves. And over time, as you do that, you start to more clearly see the chain of causes and effects that lead from your actions back to your thoughts and emotions, and how your actions, even when harmful, are completely rational given the background process it has taken to get to them.
At its core, rationality is about identifying this chain of cause and effect. To do that, we have to accept that the source or foundation can’t be some abstract thought, but rather, it’s how we model our very survival as an organism.
Join 40,000+ readers for exclusive access to my newsletter: