Abstractions are snobbery

Almost all forms of abstractions are in their essence snobbery. The problems that they elevate can be solved without them[^1], only to a lesser degree of aesthetic please, which is subjective. Although objective measures can be introduced to argue for a specific abstraction, the relevance of those metrics are again products of opinions and they cannot be reduced beyond that. Attempting to do so can be useful for the sake of winning an argument, but when formalized into a system of logical statements such an attempt will break.

So is this an attack on abstractions? Should we give them up and always work at the most atomic layer? Absolutely not, on the contratry, the take-away is to embrace this form of snobbery. To indulge in discourse of the perceived gains of the abstractions we use. To realize that the value that abstractions bring is not in their correctness: in the real practical world there is no correct abstraction, but in their utility: "this abstraction helps us see X and Y clearly". In doing so you give in and accept that there will never be an ultimate abstraction, there are useful abstractions but you might have to take them out and invent completely new ones as your understanding of the world changes and new information emerge.

Be a vigilant snob, argue for your abstractions, and burn them to ground when they fall over.

[^1]: That's indiscriminately how they come to exist: by someone grinding it the crude way and not having it.

© Copyright Anton Agestam 2020-2022