A tribesman from a hot place points at what you’re wearing. “What is that?”
“A jacket,” you say.
“What is a jacket?” he asks.
What he wants to know is the purpose for which the jacket is used, and so you tell him “It keeps me warm. It protects me from the sun. It is very fashionable.”
A computer compiling information about the world is trying to fill in gaps in knowledge. It scans you and asks “what is that?”
“A jacket,” you say.
“What is a jacket?” the computer asks.
What the computer wants to know is what it matches to most closely in its existing stored knowledge. You tell it, “It is like a trenchcoat, a sweater, a coat, or a hoodie.”
An alien artist is unfamiliar with the structure of your world. It gestures its tendrils at you and asks “what is that?”
“A jacket,” you say.
“What is a jacket?” the alien asks.
What the alien wants to know is what it is that gives rise to the jacket, what the essence of jacketness is. You tell it, “It is a bunch of pieces of fabric stitched together with some thread.”
These are three ways in which a word can be ‘defined’ – the role it plays in the world around it (the up-definition), synonyms (lateral-definition), and the parts which construct the thing (down-definition).
Generally speaking, up-definitions are the most commonly used and the most practical. What we want to know about an object is what we can do with it. The same is applied to concepts – Love is “the thing we have for our children or parents,” surprise is “the thing that happens at a birthday you thought everyone forgot about,” and “existence” is “all this stuff you’re looking at.”
Up-definitions is also one of those things that can ‘feel like’ a satisfactory answer when what you really need is a down-definition. Discussions about morality frequently fall into the up-definition trap, where everybody’s idea of ‘wrong’ is a strictly functional thing, and then people get into conflicts over why different functional ideas are clashing with each other.
I’ve seen a few discussions of free will that also fail to recognize down-definitions; the up definition of free will is something like ‘making decisions independently’ or ‘conscious choices’ – or lateral definitions like “agency” or “my soul.” To ask about a down-definition is to ask about the fabric and thread of free will, about what little bits that idea has been built out of. Generally the down-definition I like the best is “a specific subjective sense”.
Up-definitions are useful, but down-definitions aid in presenting a more cohesive idea of what your mind is doing when it thinks. With some concepts it’s difficult to put any down-definition into words, but paying attention to the feeling of thinking about the concepts can also suffice.
Probably all concepts we use are built out of many smaller concepts, and those built out of smaller still, and oftentimes we forget this so deeply that as soon as we identify an idea like free will, we view it and wield it as a solid unit, and our debates with others feature challenging how our solid units serve functionally in the world around us. It’s like knowing how to swordfight without any knowledge of what swords are made out of – it works just fine, but it’s not holistic, and might one day prevent advancing to an expert level.
7 thoughts on “Up and Down Definitions”
I’ve had only a few experiences with acid but they’ve been lovely and good for my wellbeing in the days that followed. I would love to take it again but I don’t know anyone who could provide a reliably pure source…
Hm, so I’m reminded of the old LessWrong “one level up, one level down” (which I can’t find a link for right now), but that notion of “up” and “down” is rather different than what’s here. Let’s compare.
In that context, “up” and “down” referred to the “ladder of abstraction”. So a “one level up, one level down” for “red” might be “red is a color; a stop sign is red”. Obviously, that’s not what’s meant here.
Instead the “up” and “down” here seems to be one of composition, of a sort? It’s a little odd because your up-definitions talk about function, but the connection with “down” seems to be that we’re thinking of the use of the thing in terms of it forming a component in some larger system? I’m not sure these go as well together as you suggest.
Like what you call “down-definitions” (definition in terms of composition) seem like they’d rarely be useful compared to “up-definition” (definition in terms of function)? There are some times when we care about composition but it’s just so much less than we care about function.
Of course I have to be careful here — I say “function” but I guess really above I meant “behavior”. Because that’s what we usually care about, the behavior of a thing more generally, not necessarily the particular function that might have been intended (frequently there is none; what’s the function of a rock?). Indeed I’d say we generally mostly care about composition to the extent that it determines behavior!
But OK — we do care about composition when we’re trying to construct something or otherwise break through its abstractions. You give some interesting examples of that. What I notice about the soul example is that it’s very specific; like we can contrast the jacket example. Jackets are a function-based abstraction. The composition of a jacket only matters to the extent that it helps the jacket fulfill its function. As such there is no universal down-definition of a jacket, because not all jackets have the same composition; that’s not what defines them. In contrast, we might very well expect a down-definition of free will to exist.
The morality example I’m less clear on. Like, what would even constitute a down-definition in this case? I’m not sure I get it.
I think not all of these “definitions” you mention are what would ordinarily be considered definitions, in that they’re not what you’d use to explain the concept; e.g. in the examples you use where you’re looking for down-definitions, well, you’re looking for explanations. Of course, definitions often change once we have a better handle on something! (Consider “planet” e.g.) So getting a better handle on something is sometimes a “new definition”, though I don’t know that that’s the best way to think of it when you’re still trying to solve the problem. And oftentimes it’s not up-to-down but rather up-to-up or down-to-down or up-to-up-and-down…
Anyway, not sure where I’m going with this. But interesting to compare.
In this case:
Down is reductionist, so down as in parts which tends towards an atomistic viewpoint but still relies of a LCD (least common denominator) viewpoint, aka “synonym” here.
Synonym is a sort of constructivist approach, which I am sure there is a better classification but nomenclature is more my field than colloquialisms.
Up then is a definition of value and perhaps a practical definition.
Try looking up ontology and epistemology for more fun.
I think the problem people have with “free will” can be best articulated by their problems with defining free and will independently. Perhaps once these are understood the latter will be of more consequence rather than merely an interesting factoid to bandy about because people said it was important.
Fun related quote on defining ethics which came to mind:
“Thus in ethical and religious language we seem constantly to be using similes. But a simile must be the simile for something. And if I can describe a fact by means of a simile I must also be able to drop the simile and to describe the facts without it. Now in our case as soon as we try to drop the simile and simply to state the facts which stand behind it, we find that there are no such facts. And so, what at first appeared to be simile now seems to be mere nonsense” – Wittgenstein, Lecture on Ethics, delivered in November 1929 to the Heretics Society.
I love these philosophical posts because they seem so naturally discovered. The concepts here have different names to other people who have discussed them, but your take makes it feel like you pushed through each tangent organically and picked it apart yourself. Which I bet you did.
The up/down definitions are a great way to differentiate the purpose of language to either “tact” (identify).or have meaning that has a greater depth of understanding of what it “is”. Hugely important distinction. Philosophers and linguists have tackled those points too, but you summed your ideas up nicely in a few paragraphs. Does that make this blog post a better down-definition? I’d say so.