Recent searches
Search options
I was reading Bourdieu yesterday and he also observed this: “A striking and consistent finding in much language research is that lower‑status speakers have with great regularity led the linguistic charge in many of the innovations that have become well‑accepted parts of our language” - How Lower-Class Innovation, Like, Changes the Langwage https://lithub.com/how-lower-class-innovation-like-changes-the-langwage/ @linguistics #linguistics #accents #innovation #variation #language #academicchatter
@maitxinha @linguistics Related: Many years ago George Carlin noted that if you lock a Black guy and a white guy in a room for a while, the white guy will come out talking like the Black guy, and not the other way around.
@maitxinha @linguistics I've been thinking about these claims. I don't know if anyone's ever quantified them ("97% of changes come from below!"), but that seems like it could be hard to do.
Fridland alludes to the idea that there's something inherent in working-class language that promotes innovation (less uptightness?). But what if upper-class people innovate at the same rate, but there just aren't as many of them?
#linguistics #sociolinguistics
@grvsmth @maitxinha @linguistics It wouldn't be hard to quantify for the features that have been studied -- just count how many times the conclusion has been that it's a change from below vs not -- but it would be difficult to generalize to all #languagechange from there since #languagevariation/#sociolonguistic studies pretty much never use random sampling. But when the conclusion is the same for feature after feature in community after community, it becomes easier to say the claim is accurate
@joshisanonymous @maitxinha @linguistics It may feel easier after seeing the same conclusion repeatedly, but it is not any more justified scientifically.
The temptation to say "I'm seeing this everywhere I look, so it must be everywhere, even in places I don't look!" is one that we have to resist as scientists.
Another temptation to resist is confusing what we notice and remember with what's actually on record.
@grvsmth @maitxinha @linguistics I agree, but it's usually the best we can do in the social sciences. We're stuck almost entirely with observational studies and convenience samples when it comes to things like language change because anything better (for science) would generally be unethical to do
@joshisanonymous @maitxinha As a linguist who's spent most of their research career on this, I disagree in at least three ways:
First, if you can't do the research to back up your claim, just ... don't make the claim! Call it a hunch or a hypothesis, but don't report it as fact.
Second, in many contexts there are ways of sampling the entire population. I've done it. Don't just assume you can't.
I'll post the third in another toot...
@joshisanonymous @maitxinha The third disagreement is that the bias I'm pointing to is not that the sample of *people* may be unrepresentative. It's that the sample of *changes* may be unrepresentative.
It's really hard to make a list of all the changes that have happened in a language in, say, a fifty-year period. But to justify a claim that they "regularly" come from below requires an actual list of changes, marked by whether they came from above or below.
@joshisanonymous @maitxinha Multidimensional corpus analysis is an entire subfield that attempts to do this. I wrote a paper highlighting some of the challenges and I think even some of the multidimensional claims are not completely justified, but it's an important project and they've done a lot of good work:
https://brill.com/display/book/edcoll/9789401203845/B9789401203845-s003.xml
@grvsmth @linguistics Group membership plays a role as well. Once the ‘slang’ is wide and used by others, new slang emerges