This week, from the 18th to the 22nd of March, it’s the Game Developer’s Conference. This is an event in which Game Developers from across the industry give talks and presentations on what they do and how they do it to their peer group. In honour of this, I’m presenting articles this week that seek to summarise and explain some academic concepts from my own readings to a general audience. In deference to my supervisor, I am also trying to avoid writing with italics in these articles outside of titles and cites.
How do you know what an apple is?
And do you need to know what an apple is to talk about what an apple is?
What I’m going to describe here is a basic outline of the idea of what’s called ‘prototype theory,’ but I’m not going to approach it from its native lands. Prototype Theory is part of cognitive science, and it wants to examine the ways people think in some really complex and mechanically observable, testable ways. Cognitive sciences involve a lot of questions about how brains work and how people react to how their brains work, while what I work in is communication and media studies.
I often glibly say that the work of media studies is just examining how every human being interacts with everything that does or doesn’t exist, everywhere, forever, and that makes it easy. I can’t tell you how a brain work but I can show you ways humans seem to tell stories to one another to understand things. Cognitive science by contrast is the kind of study where you’re going to bring together the work of multiple different experts to explain your theories.
Prototype theory is a way of looking at the ways we think about words and define them that lives much closer to the media studies stype of research because it looks at assumptions about how things are defined and goes ‘but what are ya gunna do, though.’
The story of prototype theory often begins with this question:
Can you define furniture?
and what follows is a sort of awkward set of pauses. This is because you have a cultural expectation for what a definition is, and what a definition is for. There’s a correctness, an absolute utility to definition – definitions are to some extent or another unassailable. That makes the difficulty in defining things a bit of a slippery point.
But you still know what you mean when you say ‘furniture’ right? You can point to things in your house and know they are and aren’t furniture.
things such as chairs, tables, beds, cupboards, etc. that are put into a house or other building to make it suitable and comfortable for living or working in.
Cambridge Dictionary definition
and
movable articles used in readying an area (such as a room or patio) for occupancy or use
Merriam-Webster definition
These are some pretty wobbly ‘definitions.’ Is a bookshelf furniture? Does it stop being furniture when it’s integrated into a wall, because now it’s not movable? Is a breakfast bar furniture or not, since it can’t be moved? But it’s got cupboards in it, so is it furniture again? Plus that first definition provided just a list of examples, that seems a strange way to define things, right?
Wittgenstein talked about how difficult it was to define things, and he used games as his example. This is why games studies always bring up Wittgenstein because they want to say ‘nuh-uh, it is so easy to define a game.’
It turns out the way we typically and culturally see the term of definition is literally derived from Aristotle, and is known as Aristotlean Categorisation. The idea is that to define something, you need a shortest list of features that is shared by all its valid members, in an attempt to distil a thing into its essence and create a simple pair of categories: Definitely the thing, and Definitely Not The Thing. A definition of furniture in this context needs to be able to explain how a tressel table is furniture but the stack of plates sitting on top of it aren’t, and then where the decorative fruit bowl next to them fits. This Aristotlean method wants to find the list of traits that are necessary for inclusion, and sufficient for distinction. As an example, you might start with a list of the things you think are definitely furniture and then say, look, a bunch of these are wood, but not all of them are made of wood, which means wood can’t be necessary for inclusion, and it can’t be used to distinguish a furniture thing from a non furniture thing. On the other hand, furniture is something humans interact with so is a couch furniture if nobody engages with it? Are the car seats orbiting earth right now furniture, or are they no longer furniture because nobody can sit on them?
(That kind of question winds up getting you into Foucault spaces.)
The complaint about Aristotlean categorisation I have is that it makes everything simple and binary and creates confusion in communication. If you ask me for some fruit juice, and I give you olive oil or tomato juice, we both know I didn’t give you what you asked for, even though a technical definition people are familiar with puts both olives and tomatoes in the fruit category. It paints a world of in-outs and dichotomies, and these categories tend to be used by people in positions of authority to exert power over others. You know, the way that some people want a definition of women that can be used explicitly to exclude trans women, and in the process wind up excluding a lot of cis women as well. It’s also a way in which our vision of strict categorisation tells autistic people the world is very tightly organisable, and they’re wrong and stupid for not intuiting the categories correctly. A lot of ‘weird autistic behaviour’ is where someone who perceives a categorical definition that works for them, and then being stressed at how other people neither perceive nor respect the category properly. Ask a noncisgendered autistic person sometime about what gender means to cis people, and you’ll usually be in for a treat.
Definitions, typically, are about power. They’re about the ability to limit possibilities.
(It’s okay, we all can use a little violence.)
The way we structure meaning isn’t actually from a series of definitions and information that integrates with those definitions. This is where we turn to prototype theory, which is a way of considering information differently. The idea in prototype theory is that our idea of a thing – which I’ll bold for reference in this sentence – is that we, through practice, construct an idea of a thing that is not necessarily any actual thing that exists, but which has the traits we think of as right for the thing. This idea is the prototype of the thing, and the more like the prototype, the more readily you can say another thing belongs to that category. If the thing is furniture, you can say that an end table and a bookshelf are both like the prototype of furniture, but a throw pillow isn’t really. This proximity-to-prototype model is also culturally constructed. Consider our fruit juice example: Some people consider olives and avocado and tomato ‘less fruit’ than apples and pears, for example.
And this makes sense! Because it’s not like when you’re a little kid learning about things, that when you ask a question the first thing you do is define it absolutely. You tend to ask what a thing is and you get a name for that thing, and then when you bring up that name, that thing gets brought back up. You see an apple, you point at it, you’re told ‘that’s an apple’ and later on you ask about apples, and you’ll get shown another thing that’s definitely an apple, but probably isn’t the same apple as the apple you were first shown.
There’s more! There’s a talk about how these definitions are socially formed – apples don’t tell you what apples are, a human tells you what an apple is and you tell that human what you think about apples, so you need apples as a reference point, but at no point did the apples get involved in explaining that. There’s how this all relates to the communication and media theory wing of things. You don’t need to know how an apple’s DNA or how to build an apple to know what an apple is, and you don’t need to know how an apple could signify sinful knowledge to know that it does. These are all ideas brought to bear through communicating about them.
After overturning Aristotlean categorisation on a level compared to Copernicus’ heliocentrism model, it feels a bit of a footnote that it looks useful to me as a way to describe the way that ideas in niche analogue tabletop design aerosolise across boundaries. There’s no strict ‘deck builder’ rules that don’t exclude some interesting cases, and there’s no strict definition of a ‘card’ that doesn’t bring in some confusing distinctions. Instead of perimiters, these ideas need to be treated as useful handles.
The source I’m considering here is the work of one Eleanor Rosch, except I’m not considering Rosch, I’m considering Rosch as described by Dan McLellan on the Data/Dogma podcast. Rosch’s idea builds on other scholars, too, like Roger Brown, about whom I know nothing, and Ludwig Wittgenstein, whose work intersects neatly with my work in games studies. I describe this to show that this is not just the work of grabbing some enormous book and reading it from cover to cover, but instead engaging with things that I find approachable, then following the threads in it to the more complex works, step by step until you realise that everything you know is anchored to a vast interlocking spiderweb of beautiful nodes of light that connect you, yes, you, reading this, at least as far back as the 17th century to the minds of thinkers like Isaac Newton and Michael Faraday. Though both of them were weirdoes, so you are part of a grand chain of weirdoes and should continue writing that The Terror/Avatar: The Last Airbender crossover fanfiction.