From Marburg to Miami:
Putting language variation on the map
So far on the blog, I’ve focused a lot of attention on the pre-history of the language and on the mostly 19th-century and mostly European scholars who conceptualized the study of historical linguistics into the approaches that many researchers in the discipline still accept and use even today. I’ve also alluded now and again to the relationship between language change and language variation, although these really have been only allusions rather than any kind of exploration of what I think is an interesting topic in its own right, one that I am looking forward to exploring in a future post. But what’s on my mind today is language variation as it occurs at a (relative) moment in time, how linguists have come to approach it as an object for analysis, and particularly how an approach known as linguistic geography came into being. This is a topic close to my heart for several reasons, all of which I think I will get to eventually in this post.
The idea of language variation as a thing to study is a fairly recent development, although the dates of some of the earlier work overlap with the heyday of German(ic) enthusiasm for historical linguistics discussed in previous posts. And the guy who usually gets the credit for pioneering research in language variation is yet another German linguist, a fellow by the name of Georg Wenker (1852-1911), who worked at the University of Marburg, which was the alma mater of one Jacob Grimm, also a major player in the development of modern linguistics, as discussed in more detail here. (Note: ‘Wenker’ is pronounced the German way, so just accept that it is not as funny as it could be.)
Beginning in 1876, Wenker sent out questionnaires with 40 or so sample sentences to 50,000 teachers throughout Germany and asked them to collaborate with their students to rewrite the sentences in the local vernacular. Wolfgang Näser, Wenker’s much younger colleague at Marburg (much younger in that Näser is still an active scholar today while Wenker went on to his reward a hundred years ago), has claimed that Wenker’s sentences contained “all relevant phenomena regarding phonation and morphology where variation of any kind could be expected.”
Some 45,000 of Wenker’s questionnaires were eventually completed and returned to him, a rate of response that I probably don’t need to point out would be pretty much impossible to replicate. It also made it impossible for Wenker to analyze all the data during his lifetime. But he was able to publish some of his findings and is usually credited for creating the first modern linguistic atlas, literally documenting Who Said What Where, with the publication of his first Sprachatlas in 1881.
In collaboration with Ferdinand Wrede (Wenke’s successor as Sprachatlas project director) and Emil Maurmann (about whom I could find out practically nothing), Wenker also used the data to create original hand-drawn maps, thereby putting the ‘atlas’ in Sprachatlas and illustrating the locations where particular linguistic features had been attested. The pioneering linguistic geographer and his colleagues eventually created over 1,600 such maps, a phenomenal feat of cartography, not to mention an equally impressive artistic achievement.
Wenker’s reliance on the self-reports and observations of the research participants rather than first-hand observations by trained linguists open the data to challenges from later researchers, but his project was an influential model for linguistic geography into the 20th century.
Another key figure in modern language variation studies is Jules Gilliéron (1854-1926), a Swiss linguist who hiked around the southern Rhône valley region to study the speech of the locals in this French-German-Italian contact area for his 1880 Petit Atlas phonétique du Valais roman (sud du Rhône) and later trained Edmond Edmont (1849-1926) to collect linguistic data in France. The lucky Edmont’s job combined two of the best things in life – biking and linguistics – and he cycled around France from 1896 to 1900 in the process of interviewing 700 speakers using a 1,500-item questionnaire. He and Gilliéron published the results along with nearly 2,000 maps in the 13-volume Atlas Linguistique de la France (1902-10).
The work of Gilliéron and Edmont was influential, especially their methods of direct observation (versus Wenker’s reliance on indirect reporting). Several more European projects were soon underway, most notably the Sprachatlas Italiens und der Südschweiz (Linguistic Atlas of Italy and Southern Switzerland), published in 8 volumes (1928-40), by Karl Jaberg and Jakob Jud, who had both been students of Gilliéron’s. In 1931, Jud traveled to the U.S. with Paul Scheuermaier, who had also worked on Jaberg and Jud’s Sprachatlas, to train American students in linguistic field methods.
The Americans who trained with Jud and Scheuermaier would soon head out to begin fieldwork (1931-33) for the Linguistic Atlas of New England (LANE), under the direction of Hans Kurath (1891-1992). Kurath, then a professor of linguistics at Ohio State (later Brown and finally the University of Michigan) had been appointed in 1930 to head up a new project, the Linguistic Atlas of the United States and Canada. LANE was to serve as the pilot study and was published in three volumes (1939-43), none of them available online. Linguistic Atlases of the Middle and South Atlantic States (LAMSAS), North Central States (LANCS), Pacific Northwest (LAPNW), Gulf States (LAGS, directed by one of my all-time favorite people, Lee Pederson), and other regional Atlas projects would follow.
Raven I. McDavid, Jr. (1911-84), then a graduate student at the University of Michigan, joined the Atlas as a fieldworker in 1940 and proceeded to conduct interviews of 278 LAMSAS participants, most of them in his native South Carolina. When Kurath retired in 1964, McDavid, by then a professor at the University of Chicago, assumed the directorship of LAMSAS and LANCS. Fieldwork continued on both projects through 1978, which was just around the time that a young doctoral student at Chicago, Bill Kretzschmar, joined the project. When McDavid died in 1984, Kretzschmar, who had recently joined the faculty at the University of Georgia, was appointed to succeed him as project director.
Kretzschmar brought the Atlas into the 21st century by digitizing substantial portions of LAMSAS data, including the linguistic data of speakers of Sea Island Creole, or Gullah, an endangered contact language spoken on islands off the coast of the southeastern United States, collected and transcribed by the legendary scholar and linguist Lorenzo Turner (1890-1972). Kretzschmar has also launched new community language projects and continues to work on new ways to make more Linguistic Atlas data available online.
* * * * *
It was 1989 when I first ventured into 317 Park Hall at the University of Georgia. I was looking for a professor I hadn’t met before whose name contained an astonishing six consecutive consonants. I was trying to put together a committee for my master’s thesis on how speaker gender is perceived to influence linguistic behavior. I had a thesis director and was now looking for a few other interesting linguistics professors to work with.
The sign on the door of 317 read “Linguistic Atlas of the United States and Canada,” and when I peeked in, I saw a large room divided up into a maze of file cabinets and bookcases, loaded down with books, journals, folders, and box after box of what appeared to be (and were) reel-to-reel tapes. We sat in the middle of the room, surrounded by 60 years’ worth of linguistic data in every imaginable format, although I had no idea at the time what it was. We talked about linguistics, my thesis idea, and whether he might consider serving on the committee. After about 20 minutes of conversation, Dr. Kretzschmar said, “I’ll do it.”
The rest isn’t quite history. I finished my thesis and my master’s degree only after some fairly insane departmental drama that makes a lot more sense to me now, was completely horrifying at the time, and resulted in Dr. K’s having to take over as director of my thesis, which he did more graciously than I probably deserved and more competently than I had previously had reason to imagine possible since I had only my former director’s example to go on. When I wrote to him years later to tell him that I was planning to go back to school to pursue a PhD, Dr. K did not hesitate: “You should come back here,” his email said.
In 1998, I returned to the University of Georgia to begin my doctoral work and start a job in 317 Park Hall. Thanks to Dr. K, I got a university-wide research fellowship that funded an editorial position with LAMSAS. My job was to compile and edit headnotes, train undergraduate researchers, and especially work on getting the data of African American and Gullah interviewees online. This meant that I got to work with the handwritten field transcripts, the pages of phonetic transcriptions that the fieldworkers had made on the fly, documenting pronunciations, lexical variants, and grammatical forms as they interviewed LAMSAS participants. Lorenzo Turner’s were clear and beautiful, sometimes including notes about cultural customs and drawings of local plants; Raven McDavid’s were lightly pencilled and impossibly meticulous, sometimes nearly impossible to interpret, the phonetic symbols often marked up with diacritics to within inches of their lives, attesting to the slightest of pronunciation differences among speakers.
* * * * *
I have never been a believer in destiny. It makes more sense to me to look back after the fact and see what patterns might have emerged as things more or less worked themselves out. And that might actually be part of what is appealing to me about the study of language, especially variation and change: the interesting designs you can see when you look back over things that happened over the history of a linguistic feature or of a language, the opportunity for interpretation, for the analytical and creative acts of developing a narrative to explain how whatever it is got that way. Unlike the Neogrammarians, though, I don’t believe in a systematicity or regularity to the processes of change or the patterns of variation. I can understand the appeal of that idea — it has an almost spiritual awesomeness (in the traditional sense of the word) about it that is very compelling — but I just can’t believe that it really works that way. And even if it does, I imagine that the design, the system, will never be within human comprehension, although that isn’t a good enough reason to stop looking for it, if you believe it is there and can be found. But I like the idea of there not being one, or of it being so incomprehensible to us that there might as well not be one. That’s almost as good.
Anyway, I don’t believe that I was destined to become a linguist. I’ve always been fascinated with the way people talk, the way words sound, and how we make meanings out of collections of sounds. In that respect, I always wanted to be a linguist. Even before I ever knew there was such a thing. But I could have always wanted to be any number of other things, too, and had a few things gone even just a little differently, this would be a completely different story. When I look back now, after the fact, over the patterns that are there now for me to interpret, to analyze, and to create a narrative to explain, I do see kind of a design, an irregular one, not a logical one, that zig-zags and curlicues between the work that has become so important to me and my first language variation research project.
It was linguistic geography, a project in which — at age 7 — I attempted to determine Who Says What Where in classic Atlas fashion. And so now, in celebration of my 40 years as a linguist, I present that study here in print for the first time:
Miami Bathroom Announcements (1972)
When I was in second grade, a family from New Jersey moved into my neighborhood in South Miami. I noticed a lot of differences in the ways that they talked from what I was used to hearing, but I was particularly taken aback by their bathroom announcements, that is, what they said when they were going to use the restroom. The main feature I noticed was that while I was used to speakers focusing more on the going, the members of this family included disturbing information about what they were going to do when they got there. This startled and perplexed me.
So I started paying attention to what people said when they said they were “going to the bathroom.” I listened to friends, family members, strangers I was lucky enough to overhear (I had to try to guess where they came from). I had family members in New Jersey, too, so I also got to listen to how my cousins talked when I was with them. I wrote down what I heard on a chart I kept in my notebook for school, keeping track of who said what where. This project went on for months. I discussed it with no one.
In the table below, I reconstruct my findings to the best of my recollection. You will probably need to click to enlarge it in order to see it clearly. I make no claims as to the validity of the study’s methodology or findings.
The usual disclaimer applies: I am a professor of English linguistics, so I am like totally credible and everything, but this blog has not been vetted or peer reviewed and therefore is not to be considered a scholarly source for anyone out there who might be looking for information for a research paper. Also, these are my original words, and while some of what is posted herein is based on widely known and available information, that doesn’t mean you can take my words or my ideas and use them as your own. That’s plagiarism and it isn’t right, so don’t do it.
All images in this post are in the public domain, except for the Miami Bathroom Announcements data chart, which is my original work, and the photo of me, for which I own the copyright. Neither may be used without permission.