Taylor Jones, known around the internet as Language Jones, has a twenty-minute YouTube video thoughtfully called “Are we WRONG about most FAMOUS LINGUISTICS experiment??” If I were modeling my style on his, I might have called this post “LINGUISTICS INFLUENCER is TOO WOKE — and WRONG about NAMES!!” But instead I went with the modest title he himself might have used if he weren’t so hungry for clicks and likes. Don’t get me wrong, I basically enjoyed the video, even though I dislike the snark-filled, overemphatic influencer style; Jones studied with the great William Labov (LH obit post), for whom he expresses great affection and respect, and clearly knows his subject. Still, I think he’s wrong about some stuff.
First off, and trivially, he says the name of Michael Lisicky wrong — he gives it initial stress, but Lisicky himself uses penultimate stress (as you can hear in the first few seconds of this video). No biggie, but I would hope that a linguist would take the trouble to get it right.
Now to the meat of the video. He discusses Labov’s famous paper “The Social Stratification of (r) in New York Department Stores,” and his basic claim is that it is fatally flawed because it does not take race into account: in 1962, when the study was carried out, the Great Migration of blacks to the north was going on, and whites were increasingly differentiating themselves from black speech — he cites Gerard Van Herk’s paper “Fear of a Black Phonology: The Northern Cities Shift as Linguistic White Flight.” Very true, of course, but the problem is that Van Herk is talking about the Northern Cities vowel shift, whereas Jones is talking about rhotic versus nonrhotic speech, and the fatal flaw in his argument is that nonrhotic speech is not a distinctive characteristic of New York Black English; to quote the very thorough Wikipedia article African-American Vernacular English, “The level of AAVE rhoticity is likely somewhat correlated with the rhoticity of White speakers in a given region; in 1960s research, AAVE accents tended to be mostly non-rhotic in Detroit, whose White speakers are rhotic, but completely non-rhotic in New York City, whose White speakers are also often non-rhotic.” Indeed, nonrhoticity is a notorious feature of old-fashioned white New Yorker speech, which means that the idea Jones is pushing, that the workers Labov interviewed were pronouncing r’s to show they were white, is absurd.
It is, of course, true that Labov’s very short (less than ten pages) paper does not prove some of the things it has been claimed to prove — it is more of a discussion-starter than a thesis — but it holds up better than Jones thinks, and I suspect there is a certain amount of slaying-the-elders going on. That said, Jones makes some good points, and it’s always good to be reminded of Labov’s work. (I should add that Craig, who sent me the link — thanks, Craig! — points out that Jones is an AAVE expert, so he would doubtless nitpick my nitpicking. As always, I welcome correction.)
That was fun! I listened to it with a specific filter very much in my conscious mind: did it confirm or oppose my own experience of NYC department store accents in the 1950s, 60s, and a bit into the early 1970s.
My father was a department store buyer/manager in Manhattan throughout those decades, at both ‘middle class’ and ‘upper class bordering on snooty’ retail establishments. Starting in elementary school “go to work with your father” (never with your mother in the 1950s) days, and on less formal occasions later, I was exposed to my father’s very conscious code switching when addressing customers—rhotic—and in negotiations with manufacturers who were mostly non-rhotic.
While I don’t have such precise recall of sales staff, I think they also adjusted their pronunciation and especially their diction when speaking with customers. When they talked to little boy me, they were absolutely non-rhotic, while with customers they suddenly remembered how to include the letter R.
In my opinion, Black migration had nothing to do with their varied pronunciations, as Black customers were notable by their absence at the middle class establishment and infrequently seen at the Fifth Ave. store.
It is a widespread phenomenon on YouTube that titles, and title pictures, are much more dramatic than the rest. Apparently people feel they have to do this to get people to watch at all, no matter what their personal style is.
Interesting stuff. His manner is a bit … intense, but not repellently so.
The idea behind Labov’s original interpretation is fascinating, even if that actual paper doesn’t bear it out (apparently.)
It made me remember a phenomenon in Kusaal that I mentioned in passing the other day: the use of the “animate” 3rd sg pronoun o “he/she” in place of “inanimate” li “it” when it’s referential (it doesn’t happen with “dummy subject” li, as in e.g. li tʋl “it [i.e. the weather] is hot.”)
This could well be a change in progress: Mooré, for example, has no animate/inanimate grammatical contrast. But the odd thing is that although I heard this often when eavesdropping on conversations or when people were not thinking about their speech as such (i.e. in normal life), if you asked people to repeat what they’d just said, they’d change the pronoun to li without noticing they’d done it, and written texts don’t do this at all; nor did people usually do this in elicitation contexts.
So it’s an “informal” usage, OK. But then what causes the “corrections”? There is no prescriptive grammar tradition in Kusaal, and as far as I know, no prestige sociolects …
I suppose it just means that a language can have different registers of formality quite independently of any social stratification, language ideology or influence from other languages. It just does …
[If I’d actually asked the speakers about this, they’d probably have told me it was due to Mooré influence. Unfortunately, this is deployed as a universal explanation for pretty much all dialectal and idiolectal variation …sometimes it’s probably even correct.]
I suppose it just means that a language can have different registers of formality quite independently of any social stratification, language ideology or influence from other languages. It just does …
Yes, this is fascinating stuff.
If one were doing Just-So Sociology Stories, one might well suppose that white employees at the lower-status store would have been at greater pains to emphasize their whiteness via pronunciation, i.e. the less class-based prestige they could muster the more they would emphasize race-based prestige as a substitute. Which would predict opposite results to what Labov actually collected, right?
But in any event, not all non-rhotic speech is the same, or even broadly similar. Consider Jean Stapleton’s non-rhotic portrayal of Edith Bunker. Whether or not you think she got the white-outer-borough-lady-of-a-certain-generation accent right, no one would take it for an attempt at a black accent. (Her native/neutral accent as used in interviews sounds nothing like Edith and is rhotic.)
Van Herk does speculate that racial shift could have been a factor in decreased non-rhoticity among white New Yorkers: “This local feature would have been highly salient in the speech of the Atlantic coast African Americans who migrated to New York City, and flight from a shared stigmatized feature may have accomplished more for white speakers than a vowel shift would have.” That said, if in fact as a matter of timing white rhoticity decreased after the black percentage of the local population increased, I don’t know how you confirm the theory that the latter caused the former because I don’t know how you would falsify it. Post hoc != propter hoc.
Another angle to consider might be to note the fairly rapid disappearance of the non-rhotic aspect of the https://en.wikipedia.org/wiki/Northeastern_elite_accent after WW2. That “posh” accent was quite different from a stereotypical blue-collar “outer borough” accent, but if both were still non-rhotic, non-rhoticity couldn’t have been the primary stigmatized feature of the latter that emphasized its lack of poshness. Indeed, non-rhoticity had to have largely vanished among the posher NYC demographics well before 1962 in order for Labov’s point to even make sense.
I agree that the Van Herk paper is uncomfortably just-so-story-like. (His brief mention of the New York rhoticity thing is actually something of an epicycle … and not the only epicycle in the paper, at that.)
USians are much better placed than I am to assess how plausible – or not – his actual overall hypothesis is a priori. But I’m not too taken with the syllogism: this is a very big thing; this is another very big thing that happened at about the same time; therefore, one of these very big things caused the other. He seems to take this as almost trivially valid, on account of the bigness …
As Albert Churchill said “When a headline ends in a question mark, the answer is no”. What does it mean when youtube headline ends in two question marks? It’s fun (in a way that horror/dystopias are fun) to think about youtube taking headline-writing on itself and tailoring them to individual viewers.
One thing I would say about Van Herk is that he’s very vague about timeline, and indeed one problem with NCS scholarship seems to be that everyone agrees the phenomenon had probably been going on for a few decades if not longer before people started trying to collect data about it with any rigor. But among the major NCS cities there are definite differences in the timing of the migration-driven increase in the black population where we do have data. For example, the upstate New York cities (Buffalo, Rochester, Syracuse) took until circa 1960 before their black population in percentage terms hit the levels Chicago and Detroit had already reached by 1940 and then quite substantially surpassed by 1960. So one might expect to see accordingly different timing of the emergence or strengthening of NCS phenomena that matches the different timing of the hypothesized causal factor if that was, indeed, the causal factor. But if anything I think the more common story about the spread of the phenomenon goes in the other direction with the shift manifesting a bit later in Chicago/Detroit than in Buffalo?
Also the story about suburban Detroit white kids wanting to sound more “urban” but that somehow manifesting as wanting to sound only like city white kids (a rapidly shrinking population in Detroit during the timeframe in question) who had been assiduously differentiating themselves vowelwise from city black kids is in some tension with lots of instances going back to the mid-20th-century of linguistic innovations flowing from urban black kids to suburban white kids who wanted to Sound Cool. Although I think here he may be giving a quick summary of someone else’s previous study and maybe some nuance has been lost in the summarization.
FWIW I grew up in an area where starting in 1978 (the year I turned 13) suburban white kids who had previously been attending suburban schools with student bodies maybe 2 or 3% black suddenly found themselves in student bodies that were more like 25 to 30% black, due to one of those grand social-engineering projects that the federal courts were more enthusiastic about in those days. It would have been a great opportunity to do fieldwork on whether and how white-suburban-kid speech changed as a result over the following decade or two, especially as you got to younger cohorts that had had that peer group back to age 6. But AFAIK no one did the data collection. (We were barely 20 miles from the UPenn campus! Labov could have sent grad students!) I can anecdotally point to a few examples of slang lexemes/phrases that seemed to travel in the black->white direction* and can’t immediately recall anything that felt like self-conscious differentiation of white speech beyond what it had previously been, other than perhaps an unfortunate heightened interest in racial-slur lexical items among a few of my less constructive peers. But I was not a trained sociolinguistics fieldworker in my teens, so I can’t say there wasn’t something that I missed.
*Although that sort of transmission could well have occurred anyway via pop culture and mass media w/o the same level of daily cross-racial in-person contact.
Note FWIW that the Labov piece indicates that the data collected on each “informant” (i.e. unwitting store employee from whom “fourth floor” was elicited) included the informant’s race, but then in the summary of results nothing is mentioned about that. Perhaps the number of non-white informants was not enough to be material; perhaps something else. Perhaps the underlying raw data is still Out There enabling other analyses to be run? (And I have a sort of unreliable half-memory that maybe this isn’t the one and only publication based on that dataset?)
There’s also a sentence (on p. 172 of the version linked to) saying “As we will see, the ethnic composition of the store employees reflects these differences quite accurately,” but then the topic is never returned to and no further information on “ethnic composition” is provided. Editing glitch? Is this version formatted as a book chapter abridged from a longer original?
I found that odd as well, but was too lazy to investigate further.
The claim made by Taylor Jones that William Labov founded sociolinguistics is baseless:
“During the past four and a half decades, studies of the relations between language and society have coalesced to form the field of academic research known as sociolinguistics. In 1952 the late Haver C. Currie published a paper, first drafted in 1949, entitled “Projection of sociolinguistics: the relationship of speech to social status” (reprinted in 1971). It took some time for the term “sociolinguistics,” for which Currie claims priority, to take root, but by the early 1960s the first sociolinguistic conferences were being held and anthologies of articles dealing with properties of language calling for the inclusion of social factors in their analysis had started to appear. In the meantime, hundreds of research papers and books on the social organization of language behavior have been published, and sociolinguistics has become a recognized branch of the social sciences with its own scholarly journals, conferences, textbooks, and readers of seminal articles. The sociolinguistic enterprise has grown so much that it is difficult to keep up with developments in its various subfields. Written by leading researchers in the field, this Handbook offers an introduction to and an overview of the state of the art in key areas of sociolinguistic inquiry” (from Florian Coulmas’s “Introduction” to The Handbook of Sociolinguistics).
Language Jones! I used to really enjoy reading his blog years ago. I probably wouldn’t have followed him to youtube regardless (I greatly prefer reading blogposts to watching youtube) but the clickbait titles also do not entice.
Looking at his blog, i see he does still make long textposts (well, a handful over the past three years) but they seem of a kind with the LanguageLog posts that do phonetic analysis of “politician said x” or wading into some social media argument with a linguistic take. Which is fine, but less interesting (to me personally) than what I remember as his old approach.
He has made multiple posts about Yiddish in the past few years, which is new topic for him (if I recall correctly).
I greatly prefer reading blogposts to watching youtube
You and me both.