just listened to @pluralistic's How To Think About Scraping, which is a REALLY good summary of the (imo) correct view of scraping and training in relation to AI. you should check it out
https://craphound.com/news/2023/09/24/how-to-think-about-scraping/
when we're trying to work on language data, pure empirical work is functionally impossible
if we start from, let's say, written documents, whether it's formal writing or casual, we still just have a giant pile of text
which means the value of corpus based linguistics was EXTREMELY minimal
so how was, and is, linguistics actually done? how do we get around the problems of tractability?
we DONT rely on pure empirical modelling!
but, that's how most science is done, this shouldn't shock you. we don't merely gather up LOTS of data and then do a best fit etc etc.
no, we have some data, maybe even a very small amount, and we develop a hypothesis, and we then test that, and refine out hypothesis, etc etc
what this means is that theoreticians have the ability to investigate phenomena by looking at pretty much any kind of sentence
and that includes both extremely rare kinds of sentences
AND ALSO extremely informal sentences
most theoretical linguists work heavily in the domain of both at once
that is to say, because theoreticians are interested in natural language, we TEND to avoid formal language entirely because it's artificial
and because we can rely on judgments, we can actually do a LOT of work on JUST informal speech
a lot of it is actually the researchers themselves doing illicitations
still a few small, focused, specialized group of people, tho!
but that's ok, because big data actually isn't any help here anyway
this is how most theoretical linguistics was done in the past BEFORE the internet, and it's how most theoretical linguistics is done NOW even with the internet
very very very few theoreticians are out there collecting gobs of data and modelling it
the act itself was also very informal and jocular. here's something McCawley once wrote, under a pseudonym pretending to be a Maoist critiquing McCawley
"This note is concerned with a counterexample to the outrageous claim
made by the bourgeois imperialist linguist McCawley. . . . Consider the
idiomatic sense of shove X up Y’s ass. As is well known, Y must be coreferential to the indirect object of the next higher clause (including the
deleted indirect object of a deleted performative verb)"
here's some more McCawley:
Consideration of these examples makes it fairly clear that the "fuck"
of (12a)–(20a) (henceforth fuck1) and the "fuck" of (2) (henceforth fuck2) are two distinct homophonous lexical items. These two lexical
items have totally different selectional restrictions, as is shown by the examples:
(26) Fuck these irregular verbs.
(27) *John fucked these irregular verbs.
(28) Fuck communism.
(29) *John fucked communism.
and here are some more, of various origins:
1.
a. The fact that Max plorbed Betty did not convince Pete to caress her on the lips. (Postal 1988a [1969]:74)
b. Mary tried to give John a blow job, but she choked on it [ambiguous, depending on Mary’s success]. (Douloureux 1992 [1971]:48)
c. Let’s fuck. (R. Lakoff 1977:82)
2.
a. Hey, if John went to Chicago, that means we’ll soon have a big supply of
dope. (Schmerling 1971:249)
b. My cache of marijuana got found by Fido, the police dog. (R. Lakoff 1971b:154)
c. Fred does nothing but smoke hashish and play the sarod; John is similar. (McCawley 1976b [1972]:304)
3.
a. The M.C. introduced Mick Jagger’s penis as being large enough to amaze
the most jaded of groupies. (Borkin 1984 [1974]:18)
b. Paul is dead and I do not believe he is dead. (G. Lakoff 1975)
c. She left one too many a boy behind. He committed suicide. (Bob Dylan, cited in Zwicky 1976:683)
and some more:
4.
a. Amerika’s [sic] claim that it was difficult to control Vietnamese aggression in Vietnam surprised no one. (Grinder 1970:300).
b. *The shit that John took weighed 600 grams. (McCawley 1988 [1971]:96)
c. *I don’t want to kiss no gorillas. (Postal 1974:236)
its pretty clear from these that INFORMAL language was the focus of at least this body of research, but actually this is most of theoretical linguistics here
usually our examples aren't so vulgar -- the Generative Semanticists were trying to bring some levity to an otherwise boring tradition of boring example sentences and so they intentionally constructed examples that were obnoxious and funny and gross and whatnot because science should be fun -- but nevertheless, most theoreticians work in casual speech, not formal speech
and indeed most theoretical work is somewhat allergic to formal speech. we reject it as ENTIRELY uninsightful into language because it's FAKE
we theoreticians want to study the human mind's capacity for language, not what Strunk and White and the rest of the Rome larpers think
"formal" english has rules like Never Split an Infinitive
but that shit was literally invented because in LATIN you couldn't split an infinitive and people like Strunk and White wanted to model English on Latin because of Rome LARPing
that's horseshit and not science
so linguists don't use that shit
we use casual language
beka valentine
George Station
it's B! Cavello 🐝