More Labov remembrances
« previous post | next post »
Ximena Conde, "William Labov, ‘father of sociolinguistics’ who studied the Philadelphia accent, dies at 97", Philadelphia Inquirer 12/23/2024:
William Labov, the father of sociolinguistics who spent decades recording how Philadelphians talk, calling the city the “gold standard” for studying language patterns, died Tuesday, Dec. 17, in his Washington Square home at the age of 97. He died of complications from Parkinson’s disease.
Dr. Labov approached language as something that by its nature was variable, not governed by an ideal set of rules of grammar. His work changed whose dialects linguists saw worthy of study and dove into the socioeconomic politics of language. The way he saw it, dialects touched everything, from how you’re viewed to how you learn.
The whole obituary is well worth reading.
Here's a recorded lecture by Bill from 10/15/2008, "Language Change in America"; a 1/24/2013 interview "American English is Changing Fast", which was viewed 2,056,419 times; and a 2020 Abralin lecture, "Justice as a Linguistic Matter". A couple of Bill's other recorded talks that I've posted about over the years: "Labov's Haskins Prize Lecture" (1/15/2010); "Talcott Parsons Prize: Bill Labov's acceptance speech" (3/18/2020).
If you'll be in or near Philadelphia on January 9, you might want to attend the "Labov memorial event at the LSA".
The day after Bill died, Betsy Sneller posted about the "Bunny Paper", an unpublished paper based on a talk Bill gave in 1970 to the Hawaii Council of Teachers of English. In this paper ("Finding out about Children's Language"), Bill illustrates an important point about tests and surveys that is still under-appreciated by social scientists today. You should read the whole thing, but here's a taste:
The particular test that I'm going to discuss was designed to elicit the verbal response of kindergarten children in an experimental program. It was designed as a controlled test in which each child was subjected to the same stimulus and his response would be measured and compared to those of others. The data were supposed to give information on the children's cognitive capacities, their linguistic development, and the grammatical system of their dialect. It was hoped that this information would provide the basis for materials to be used in the first grade program. The test bas three main parts: five factual questions, a request to tell a story in response to a stimulus, and a series of probing questions about the picture.
The five factual questions were: (a) What's your name? (b) Who's your teacher? (c) What grade are you in? (d) Have you any brothers or sisters? (e) Do they go to this school? These are all "known-answer" questions: the teacher-tester bas the information in front of her. Their purpose was to test the children's ability to comprehend and respond accurately to simple questions.
The child was next shown a large photograph of children playing on a city street, and given careful instructions for what to do. I want you to look at this picture on the wall. Do you see it? I want you to look at the children in this picture and I want you to see if you can tell me a story about what the children are doing. Do you think you can do that? All right, whenever you're ready you can begin your story." We can call this a "request for display."
Responses to this request were never more than a few short sentences. The teacher-tester then asked questions about what particular children in the picture were doing, where the children were playing, about some litter in the street, etc. Finally, the child was asked if he had finished his "story" or if he had something else to say.
This test "proved" that some of the children were completely non-verbal.
Some were paralyzed into silence by the request for display:
Teacher: Can you tell me something about that picture, everything that you see in the picture, and everything that you see the children doing?
Eunice: [3 seconds silence]
Teacher: Will you do that, Eunice? … All right, go ahead. You just begin your story.
Eunice: [3 seconds silence]
Teacher: Tell me what you see in the picture, Eunice. Come on, look at the picture. Tell me what you see in it.
Eunice: [4 seconds silence}
Teacher: Hrrnmn?
Eunice: [10 seconds silence]
Teacher: What's this boy doing, Eunice? What's he doing?
To all these questions, Eunice preserved a stubborn resistance. Finally, she produced a minimal response to the teacher's verbal bludgeoning:
Teacher: What's he doing? What's he playing with?
Eunice: Ball.
Teacher: Hmmm?
Eunice: Ball.
Teacher: All right. What's he doing with the ball?
Eunice: [Silence].
Bill explains:
The adult-dominated situation which provokes non-verbal behavior in children gives us very little insight into the real abilities that they possess. Even if the adult is friendly, insightful, gentle and persuasive, we find that the language children use in his presence is essentially a response to him. To understand what children can do with language, we have to observe them talking to each other. Ideally, we want to record their spontaneous interaction with members of their own peer group, children that they see every day after school. The setting should be as remote as possible from the classroom or the testing laboratory. It should be dominated by the children, not by adults. They should be confronted with a difficult task, something which provokes disagreement and challenges understanding. And we will obtain the best results if the child we are studying has someone slightly lower in status than himself–someone that he can explain things to. Careful studies of the uses of language show that a great deal of speech is produced to raise the status of the speaker. No different from adults, children will talk the most when they have the most to gain from doing so. They will talk to those who look up to them more than to those who look down.
When I went to the mainland school where these tests had taken place, I wanted to see what I could do to get at the children's actual capacity. They were then being trained for twenty minutes a day in a method which presumes no linguistic competence at all.
Following the principles I just outlined, I came into the classroom with a rabbit under my coat. The children wanted to know: "What's that under your coat?" At first I wouldn't tell them, but finally I said, "Look, I've got a rabbit here. Can any of you help me out? Somebody's got to take care of him while I'm talking to the teachers." Everyone naturally volunteered. I selected one (in later trials we always picked the least verbal child in the classroom), and asked him to pick three friends. They took the rabbit into a little room where my tape recorder was set up and running–in plain view. They were told, "He's kind of nervous, so just keep talking to him. He's used to people talking to him: talk to him and he won't get nervous."
And needless to say, the allegedly low-verbal or non-verbal children engaged in active and vivid conversations with the rabbit and with one another.
It's worth considering what a similar perspective might suggest about the tests that show declining reading and math abilities. In addition to the social dynamics noted for children in the Bunny Paper, adults are paid $50 for participation in these 1-2 hour tests, with no necessary motivation to do anything other than get through the process as quickly as possible. So do declining scores tell us about changes in the distribution of literacy and math skills, or do they tell us about changes in the distribution of motivation to perform well on tests?