A follow-up to yesterday’s posting “One of these things is not like the others”, in which my (now AI-enhanced) Google search for
“African American linguists”
produced a display of 9 people (plus a further display of 4 others) which was instantly remarkable because the person in the position of pre-eminence in the first display, Walt Wolfram, was not (unlike all the others) African American / Black, but notably German American / white. WW is an amazing, prolific scholar of African American English and its uses and of African American communities, and he is a champion of those communities, certainly deserving of huzzahs and celebratory parades and official recognitions with laurel wreaths and gold-embossed certificates and all that stuff, but he’s unquestionably white — as, in fact, the photo accompanying his name in that display makes clear. (I’ll add that he doesn’t “talk Black” either. His everyday variety of English is working-class white Philadelphia, tenaciously maintained throughout years of formal education; it’s one of his badges of identity.)
Just after I posted about Google’s “African American linguists”, it occurred to me to wonder about a rather long list of other names I would have expected the Google search to pull up: among them, Dr. G (Geneva Smitherman) and Alim (H. Samy Alim). Well, Google search is a fuzzy tool, maybe I should have asked a different question. So I tried:
“linguists who have studied African American English”
This elicited a list of linguists that (correctly) had WW on it, but no other white linguists, not even Bill Labov (William Labov, author of the crucial work Language in the Inner City of 1972, following on his field-defining The Social Stratification of English in New York City of 1966) or Roger Shuy (another trail-blazing scholar in the field, along with WW inspired by Bill Labov); and more names of Black scholars (now including Dr. G and Alim). I suspect that the search switched from searching about a hard question to searching on a related question it could find an answer to, namely:
“people who have studied African American English”
Here I note that Google searches are now, oh dear, managed by generative AI (with a note that it is still in development). This turns out to mean that they take strange turns, go down rabbit holes to recover tangentially related or entirely confabulated results, miss obvious stuff — and are constantly changing, so that two identical searches made half an hour apart can produce dramatically different (but still peculiar) results. Sometimes, the first search gives me that array of faces, sometimes a list of a few names, sometimes (cagily) just links to sites with “African American linguists” (or something very close to that) prominently on them.
There is deeper, some of it quite entertaining, bizarreness in my searching adventures. I now know why Dr. G and Alim didn’t come up in the first search, and I have an excellent hypothesis as to why the searches think Walt Wolfram is African American but Bill Labov is not — an idea that reduced my caregiver J (who is a computer software guy as well as a caregiver) to tears of helpless laughter.
In any event, I now can’t imagine why anyone would use a Google search to find an answer to a question they care about; you can trust it to take you to Wikipedia pages and the like, but on anything else you’re likely to end up in the Twilight Zone. Especially dangerously, it’s probably going to cough up some nuggets of truth (however it got to them) with bits of strange fantasy embedded in them. And then how do you investigate the investigators? And how do you find out about relevant stuff the search simply fails to tell you about?
The title. Invented Chinese pidgin English, intended as something you would hear in a Chinese laundry — conveying “I looked, but I didn’t find anything [in particular, I didn’t find your laundry]’, and also conveying a racial slur. From my 4/23/12 posting “Me no likie?”, about:
a formula no X, no Y, used as an ethnic slur for a long time, especially in the fabricated pidgin English of Chinese laundries: no tickee, no washee / shirtee ‘if you don’t have a laundry receipt [a ticket], I won’t give you your laundry [your wash / your shirts]’
In this posting, the slogan applies to search programs that look for easy stuff — how hard could it be to look for African Americans or for linguists? — but then fail to find obvious exemplars. While, incidentally, supplying the counterpart of an occasional dead raccoon instead of the laundered shirts you were asking for.
The first query. Asked for “African American linguists”. I’ll put aside for the moment the question of how a search program tells whether someone is African American / Black. While noting that the method you’d be likely to use, namely looking at a photograph of the person and judging their race from that, calls not only on visual abilities (something search programs don’t have), but also on background knowledge and reasoning abilities beyond simple programs; it’s a hard computational problem.
But where do you search for linguists? And how do you determine whether someone is a linguist? In the real world, these are immensely complex questions, as all questions of sociocultural categorization are. But search programs need something vastly simpler.
I went back to the puzzles of Dr. G and Alim; why weren’t they in there with John Rickford, John McWhorter, Lisa Green, and Sonja Lanehart?
Dr. G is, like Walt Wolfram, a very senior linguist (they are both just a year younger than me) who has done ground-breaking work on language use in Black communities: Dr. G in her work on the discourse forms used for various purposes within these communities (her ground-breaking work is the 1977 book Talkin and Testifyin: The Language of Black America); Walt in an enormous project documenting everyday language (of many sorts, but with plenty of African American English) in North Carolina, from one end to the other. Both of them have served as mentors to many Black linguists. Why does Google find him but not her?
Then there’s Alim, who is, with Sonja Lanehart, roughly one generation younger than those two. To my chagrin, I see that I haven’t written an encomium of Dr G., but I have written ones for Sonja (on this blog, two days ago) and for Alim (in my 2/13/17 posting “Books from Stanford”; Alim was then at Stanford, is now at UCLA). At Stanford, Alim was a professor of education, with courtesy appointments in linguistics and anthropology, and director of African & African American Studies (Alim is both African American and Muslim). One prong of his research is the study of how race and language are co-constructed, a field for which he coined the term raciolinguistics; another is the ethnographic study of hip hop culture. In any case, he is now David O. Sears Presidential Endowed Chair in the Social Sciences and Professor of Anthropology at UCLA.
Why does Google find Sonja but not Alim?, I asked. And then the penny dropped. Alim is, among other things, a linguist, but he’s in an anthropology department. Sonja is, among other things, a linguist, and she’s in a linguistics department. What about Dr. G? A lifetime career as a linguist in an English department, in Michigan. Meanwhile, Walt Wolfram is the William C. Friday Distinguished University Professor of English Linguistics at North Carolina State University. Is it actually about job titles?
Wouldn’t that be a dumb-assed way to go about finding linguists? Oh yes, especially given the great variety of departments linguists find themselves in. But you’re a human being, not a Google search program. Job titles are a snap, belonging to the category LINGUIST is hugely complex, and doesn’t translate easily into search schemes,
X State University. Could there, I then wondered, be some equally dumb-assed basis for Google’s persistent illusion that Walt Wolfram is African American? Not involving departments, but maybe institutions. I was aware that state-supported institutions of higher learning have a variety of historical purposes, and that one of those purposes was the education of African American students; another was the education of teachers; still another was education in useful skills, especially in agriculture (but also in mining, mechanical trades, STEM fields, and more). I know a good bit about institutions of higher learning in two states, Pennsylvania and California, and about the naming schemes for such institutions across the country (which are complex and vary from state to state), and I am tempted to display some of this knowledge, because it’s entertaining stuff, but let me cut to the chase.
What gets called X State University differs from state to state, in confounding ways. In Pennsylvania, Pennsylvania State University is the name of the state’s land-grant university (under the Morrill Act of 1862, signed by President Lincoln, providing grants that included support for agriculture); while TownName State University names more practically minded institutions (Kutztown State University, the one closest to where I grew up, began life as Kutztown State Normal School, then Kutztown State Teachers College, then (with broadened responsibilities) Kutztown State College, and eventually Kutztown State University.
In California, though, we eventually get the California State University system (with many institutions), against the more graduate research-oriented University of California system (with many institutions). California State University, Northridge and San Diego State University in the first system, the University of California at San Diego in the second.
Which brings us to North Carolina State University (in Raleigh), which is like Pennsylvania State University (in College Park) and Iowa State University (in Ames), the originally agriculturally related state university.
And then the second penny dropped, when I recalled that is some Southern states, there was yet another origin for X State University — for a state-supported HBCU (Historically Black College or University). So, just below the Mason-Dixon Line, we get Morgan State University in Baltimore MD, a public historically Black research university.
Then I discovered that South Carolina State University is an HBCU. North Carolina State University (where Walt Wolfram works), in contrast, was originally closed to Blacks and only relatively recently opened its doors to them; even now, the student body is only 6% Black (the state is about 21% Black).
It turns out that for 7 Southern states, StateName State University is a HBCU:
WV DE
KY VA
TN
AL
It follows that a StateName State University is probably a HBCU, and a professor there is probably Black. But mere probabilities are good as gold for an AI-run search routine. So Walt is almost surely Black. Well, as near as makes no difference, right?
Yesterday. My caregiver J asked me what I was working on — I gave him a quick rundown of the beginning of the WW puzzle — he said, astonishingly, that he’d heard that in the South, State University sometimes meant Black. (Where does he pick up this stuff? A little while back, it was songs from Candide.)
“BINGO!” I cried, adding a quick wrap-up note about NCSU’s deep resistance to the admission of Blacks. We both laughed until we cried.
My best Kharkiv Opera performance so far, I think.
Leave a Reply