Adventures in computation lexicology. A physicist uses mathematical analysis and modeling to decide whether “expresso” is a legitimate variant spelling of “espresso” in English, and finds that it is. While “expresso” is used much less frequently than “espresso”, its usage has held up well over time; so it is not going to go away, and, given a piush, it might some day overtake “espresso” in popularity. Wonk alert: Quite a lot of mathiness, especially around the middle
pull down to refresh
57 sats \ 2 replies \ @OT 30 Nov
Isn't that how language works? If enough people say it, it becomes correct.
In Australia people pronounce Melbourne mel-bn. When you hear an American say mel-BORN is sounds so wrong!
reply
144 sats \ 1 reply \ @jakoyoh629 30 Nov
I don't know if that's the deal with your example, but a lot of the time, it's just a local accent. Doesn't mean it's wrong.
But yeah, there's this whole language thing where bigger crews start roasting the smaller ones for how they talk and for using some less common words. You see this low-key turf war that, long-term, just kills off diversity. I'm not sure if that's a good thing or a bad thing, though.
reply
21 sats \ 0 replies \ @adlai 30 Nov
it's a double-edged sword; dialect markers are also used for exposing outgroup members, e.g. how locals can recognize that some visitor is a tourist, possibly even from a foreign nation rather than just the next town over.
reply
0 sats \ 0 replies \ @adlai 30 Nov
I find the effect of LLMs on this interesting: there's probably reduced pressure on humans to copyedit their own words now that producing text which is "too perfect" will win you the scorn of people [and bots...] who reflexively accuse anything of above average punctuation and grammar quality to be superhuman; this reduced pressure means greater variance in the range of spellings and grammar produced and tolerated by people, possibly accelerating dialectical drift.
reply