#composable language

it was a very useful word before software usage coopted it.

"This is bad", "this has flaws", "this has problems", "this is broken", "this will misbehave", "this is liable to confuse or mislead your intuition into using it wrongly", and so on.

Those are deprecations in the original, general, composable sense. And often, those are good reasons in order to make a new version or alternative, and encourage people to stop using the old one, and start using the new one.

But then software has a very justified need for another meaning, that is more general in another sense, and could also be usefully composable if abstracted from software: "there is a newer version or alternative, and this one will eventually be unsupported or otherwise become 'bad' or less useful in the future" and "this way will be supported and good to use indefinitely, but you are encouraged to use the new way unless you have good reason not to".

A decent amount of modern "deprecation" usage in software is more like that second meaning.

On the one hand, I don't like the muddying of waters that happens when two useful meanings get conflated under one word. And I do think that part of this is because too many people think they can just learn words from context without missing or damaging some important meaning.

But on the other hand, both meanings are useful. It's just that words have their usefulness damaged when they have to carry two or more distinct meanings that can be useful in the same context.

Cultural appropriation as political BDSM created by BIPOC people is certainly a take. I don't agree.

For starters, it's only BDSM if everyone involved consents.

Jokes aside, I understand how a lot of rhetoric out there can give this impression. But I think that's because "cultural appropriation" nuance decays readily into a rationalization magnet for having an issue with a huge variety of things. Anything that empowers people to say "this thing you like or do is bad (and you should feel bad) (and you should stop)" in a socially acceptable and reflexively supported way is going to catch on with people who are interested in making others feel bad or making others stop doing what they're doing.

So then the question becomes: is there any other useful idea that the term "cultural appropriation" could have been intended to cover? Because if there is, then that's much more likely to have been the origin. And I still stand by my "Cultural Appropriation 2" post from a few years ago where I describe one such idea. A good way to discover that idea, or maybe other useful ideas that might fit the term, is to treat the words as composable: throw away the term "cultural appropriation", start with just the dictionary definition of "appropriation", and then try idea-fitting possible ways to apply "cultural" to that meaning - if there's a combination of those concepts that seems like it is actually useful, then I tend to assume that's what someone meant when they came up with putting the words together like that. Because I have to assume someone somewhen was able to do at least as well as I can at seeing useful concepts and the right words to express them.

We can come up with timeless definitions of the left or the right in terms of cognition or ideology, and in many ways this can be useful for analysis. But the actual on-the-ground process by which people identify as (or are labeled as) "left" or "right" is going to be more tribal and freed from adhering to the strict definition.

And so the danger is that if we teach people something like "[ideology label] is inherently a [left/right]-wing ideology", what we put out is both a cognetic opening and a rationalization magnet along the lines of "I can't be [ideology label], because I am (identifying as the side/tribe labeled as) [left/right]".

We could prevent this by just as consistently and successfully reinforcing our timeless definitions of the political left and the right, and teaching people to spot the deviations between those definitions and what in practice is associated with the "left" or "right" by the culture at large, and raising people to prioritize changing their political labels to match some official definitions rather than how people around them are using the definitions in practice. But that requires a lot more people to engage in persistent pedantry that we can see too many people find neurotic and ignore. And frankly, the nuance decay properties are not in our favor: "[bad thing] is inherently an [other side] ideology" is more spreadable, appealing, and accessible to a much wider audience than "[your/other side] label actually refers [precise definition]", and there's a lot more social reward for the former than for any of this nuancing.

This is true for all political labels of course, but at least some labels (for example "liberal" and "conservative") have useful composable apolitical meanings, and gain their political meaning from combining that apolitical meaning with political context. This helps anchor and stabilize their political meanings across time, so long as people learn and appreciate the apolitical meaning, and the usefulness of composable language.

To thingify a word or concept is when we start thinking about the word or concept as a discrete "thing", as designating some concrete category or boundary from other distinct and different things.

As opposed to thinking of it whenever possible as an independent dimension of description, a combining force, a pattern that can overlap, or an area or direction in the space defined by one or more of the many factors that make up and ultimately contribute to discrete "things" being what and how they are.

Thingifying is the typical, casual, and normal way to think, but it's utterly awful for really understanding anything more complex or nuanced than, well... concrete, very different things.

Instead the incel communities rapidly spiralled into a cesspit of entitled and misogynistic resentment, and the people talking about them had no better term than what they self-identified as.

Part of it was idea-fitting insecurity: people might think I was trying to look clever by using terms from quantum mechanics, or that I was intellectually incompetent enough to take a vague layperson impression of quantum mechanics and just free-form associate my way to ideas from there.

Part of it was that I didn't want to add to the nuance decay of those words - I knew quantum mechanics is this complex subtle thing that I didn't fully understand and that many people misinterpret, and I wanted to not muddle the precise meanings, which I did not yet trust myself to fully understand.

So even though as a general rule I believe that words like "collapse" and "superposition", which are usable in general and composable ways, should never be reserved for narrow uses like that, I mostly kept these uses to myself, and even in my head there was some reservations about then.

Until I realized that the original meaning of quantum wave-function "collapse" (except they originally called it "reduction") is possibility collapse - they are the exact same concept.

"Superposition of states" was just an attempt to give a name to "all the possibilities not yet ruled out", or perhaps even just a translation of such an attempt.

Words like "good", "bad", "ethical", "evil", "useful", "important", and so on, do not actually mean anything objective or specific.

It's not that they are meaningless: what they really say is "I judge these things to be for or against [some implied priorities], because of [unstated reasons]".

We can make priorities and reasons explicit, even though we usually don't. For example:

It is "good" for fast and clear communication that these words work this way, because it makes these words composable. We can bind any values to them, instead of needing a separate word for every way of valuing things - they abstract the specifics.

The problem is that people often slip into thinking about these things as if they are absolute, or as if everyone fills in those blanks with the same unstated priorities and internal reasoning for determining value that we do. But minds project value onto the world, and all minds are different.

It only sounds constructive if you do bad-faith idea-fitting: if you assume that the person is not saying what's in their head in the way that is most natural and clear to them.

Because if you do put yourself into the mental state of saying things as clearly and naturally as you can, it should be obvious that if the feedback isn't specific, you have no way to figure out which of your many words and phrasings didn't click with someone.

To use a recent example: to me it is equally not intuitively obvious that these two things would cause confusion in casual conversation:

the sky is blue

escalation in violence isn't moving up discrete levels of maximal potential harm, it's sliding the scale of survival probability in your favor

To me they are only as complex as the underlying ideas force them to be, and no more. At the time of writing the thoughts behind them are equally clear and smooth and those words are the most direct flow of thought to language I have.

If I could intuit what might be difficult about either the language or the thoughts behind it, I would've worked to mitigate it. Clearly I failed, but how?

Like I really appreciate that even thesaurus Anon was giving me feedback, in their own way, but I can't do anything with it by itself. My mind doesn't have big words - it has words I comfortably know and words I don't.

Let's look past the false premise that I ever use a thesaurus to write (well, okay, I did eventually start occasionally using them to look up simpler or more common words that are close enough to words I would automatically use, after being accused of trying to sound smart or wording things in convoluted ways too many times).

"Don't use a thesaurus" is saying that somewhere in the space of all language I have sorta recently used there is too many words that are not smooth enough for others to parse that they stand out. Except to me that (inconceivably large) space is filled with just me comfortably using words that come naturally do me.

And if someone did good-faith idea-fitting on my words and the likely intentions behind them I don't see how they can conclude that it would be obvious to me which of the countless words that flow out of me naturally don't click.

At least when someone tells me a specific thing I said makes no sense to them I can start proactively interviewing people and thinking about how other minds work to figure out why that might be.

But the possibility space of all wordings I could use instead is also infinitely large. So instead of a giant range of spots to check an infinity of alternatives, I narrow down to a focused spot to check an infinity of alternatives.

And that helps (a lower-cardinality infinity is definitely substantially smaller), but if I had any strong evidence for any direction in that space being more likely to be the best possible wording, don't you think I would've taken it or at least explored it?

Of course sometimes I just overlook better wordings that I could think of. Mostly that happens out of habit or lack of time: some turns of phrase become automatic for certain mind flows and doing an "optimization pass" over them could collapse the wording into a simpler wording that still works, but only by taking advantage of that specific context.

If I use a convoluted phrasing it is almost always because it is actually more robust, more enduring. Instead of falling apart when looked at closely or moved, it brings enough cues that a careful and diligent thinker who does good idea-fitting can keep tracing over the details and asking "why this choice?" until eventually they find the right meaning, or get at least close to it.

But that requires readers to do good idea-fitting, and I wouldn't've felt the need to define a distinct term for it emphasizing that it is an active and iterative skill, one which composed well with abstract value words like "good" and "bad", if it was something that most people were already reliably doing well.

There is no conclusion here. No nice summary. I am still working on being more concise. And using smaller words. And simpler phrases. I won't stop anytime soon.