Language provides a framework that helps us think. The way we make sense of the world comes from our ability to think in categories, draw distinctions, and create structure. Thus not only does it help us derive facts and truths about the world, but considering the amount of sense data we have to process it does so in a pretty timely manner. But it’s not perfect as we found out when the world couldn’t agree on whether the dress was white and gold or blue and black.

It’s clear how language helps us categorize, but so too does reading and writing. Which, to us is almost as natural as language, and similarly important in parsing the world. Literacy is a cognitive tool, not something you’re born with. Language, on the other hand, you are born with some degree of innateness (an idea championed by Noam Chomsky). One of the cognitive breakthroughs for Homo sapiens was forming language and art. It’s not just important because it means you can communicate. But because it’s also a form of distributed cognition. You don’t have to hold terms in your mind, you can leave them somewhere. You can connect a thought from your past self to your future self. You can link ideas with other people and form a network. A network that can travel over time. Think about how significant this would be in the progress of humanity and getting us to where we are today. And along the way, we developed a framing that fit our evolution and characterized our thinking.

Categorical thinking was one of the first philosophical problems. Before Aristotle, before Socrates, Parmenides explored the topic bringing up discussions like the following: Consider a chair. There’s only one thing you can say about it, it is. You could also say the corollary, it is not a table. But then we get a paradox where something that is not, is existing. So Parmenides decided whatever is is and whatever is not cannot be. But what part of this discussion in metaphysics is relevant to me? We’ll get there but first back to the metaphysics. Imagine a yellow chair. The thing is a chair and it is yellow. Really there are two things: a chair and yellow, so how can there be two things that are one thing? This is the sort of problem pre-socratic philosophers got hung up on and we don’t. But that’s because Aristotle gave us the categories necessary to understand such a sentence. Our Aristotelian thinking is so ingrained in fact it’s hard to see how there could be any problem at all with such a sentence. The difference comes in how we’re using the key word ‘is’ in two different ways. There is the ‘is’ of essence (the chair) and the ‘is’ of attribution (yellow is a property of the chair). There’s nothing obvious, no distinguishing characteristic about the two ways ‘is’ can be used here. And so it is a complexity that Aristotelian thinking solved to the point that it’s just fact to us now. Who says no progress has been made in philosophy?

Biology has the same problem in classifying life. If you try to define different animals purely by a set of characteristics your list grows tediously long and still you might find counterexamples. With a short list we run into problems too. Plato quickly found this out after defining humans as ‘featherless bipeds’ when Diogenes plucked a chicken bare and said “Look! A human!” But going through a checklist of attributes is not how our brain thinks. We can somehow answer the question of whether something belongs or doesn’t belong in a category, but we have trouble defining the category itself. Why are you easily able to identify a bird, even though not all birds share those characteristics? Besides, what image comes to mind when you think of a bird? And why not a penguin or a duck? This categorization is intuitive to us, but it isn’t easy to model. A neural network might run through thousands of pictures of dogs and become decently accurate at identification. But for humans, even with a less familiar object, the amount of ‘training’ a human needs is significantly less. But we’re far from perfect, the process is often messy and imprecise. Often there are not necessary nor sufficient conditions for items to be included in a set. For example, soccer, hopscotch, chess, League of Legends, and hide ‘n seek are all games, yet have little to nothing in common.

One theory of mind, prototype theory provides that we think in terms of a typical example rather than a range of options or a list of criteria. And that is why, when considering games the first thing that comes to mind might be checkers, not Cyberpunk 2077, or maybe the other way around. Depends whether you’re over 40. Similarly if you’re asked to think about a bird, you will likely conjure up what a common generic bird is to you. For me, some plain finch.

Alfred Korzybski coined the phrase “the map is not the territory.” Our mind creates a map when we build categories, think in prototypes and use heuristics. It’s important to remember we have a model of the world, but it’s not the world itself. That doesn’t mean we should throw out maps altogether. Maps are useful; they simplify and encode a lot of information, but they are not the territory itself.

The human ability to find patterns is powerful and adaptive but also susceptible to error and self deception.Take color as a simple example of a category (forget the dress). Color is a continuum, a spectrum. Language divides color into categories so we can talk about it in a meaningful way. We can say that the sky is blue and know what that means (that it is not green or yellow). But of course the lines we draw between colors are arbitrary. Color categories are not universal. English did not have a distinct word for orange before 1500 – when orange trees were brought from Europe to Asia. Also many languages don’t have a different word for the color blue (it’s all shades of green, just ask Miles Davis). Interestingly enough colors aren’t that easy for children to learn (but that appears to be mostly due to confusing English syntax – placing color before the noun. Color may not seem to be the most poignant example on the issues with categorical thinking, but it does point to a bigger issue of how different people might perceive the same world that you’re in. One problem with thinking in categories is we overestimate how different two things are when there is a boundary between them. Our attention shifts its focus toward the line we draw that separates the facts. In reality blue and green are not so far apart (in comparison to say blue and yellow). In focusing on the border between objects we miss seeing the bigger picture.

So what’s the upshot here? What heuristic can I apply to frame things accurately? Unfortunately it’s not as simple as “pay less attention to the arbitrary rules that divide things up”. Because the sibling problem to the one above is underestimating how different two facts are when they fall into the same category. One manifestation of this; people have trouble differentiating their response to 1% risks, 0.1% risks and even 0.01% risks. The recent pandemic has been a big social experiment in many ways, one being how different people perceive the same risk. Still after having more information than when it started people’s perception of risk runs the gamut.

The categories we come up with have purpose. They form the basic framework that sorts the sensory data we receive. In general categories help our system 1 thinking and processing of the world. They give us a single frame of mind, a reference point, which is useful in communication because others can share that. But for better critical reasoning, one place to start is noticing where the lines fall and reframing so you can see from a variety of reference points.