Abstract
Professor Jack Chambers at the University of Toronto, also known in publications as J.K. Chambers, is an eminent Canadian sociolinguist, author of Sociolinguistic Theory (Wiley-Blackwell), co-author of Dialectology (Cambridge) with Peter Trudgill, co-editor of The Handbook of Language Variation and Change (Wiley-Blackwell) with Natalie Schilling, and many other books and articles.
My association with Professor Chambers began when I was an undergraduate, and I worked with him as Research Assistant as well as his student. I later served as regional director of the Vancouver survey in his Dialect Topography project. The interview was carried out on e-mail and in person.
Hi Jack, it’s a pleasure to speak to you about your career and thoughts on language. You were a keystone in my early foray into linguistics, and later rekindling my love for sociolinguistics after I returned from McGill University. So, take us back to the beginning. How did you get into Linguistics?
It seems like a circuitous route, looking back on it, but at the time it just seemed like a series of natural steps. I majored in English Literature and Language at the University of Windsor, where the “language” component consisted of Old English, Middle English, and Greek Literature in Translation. Then at Queen’s University for my M.A. it was all literature, no language. I started my Ph.D. at the University of Minnesota, but I only stayed one term. It now seems like scholarly arrogance—or maybe just immaturity—but I had the feeling that I had already mastered all the skills that I would use in literature. So I returned to Canada and took a job teaching secondary school English while I decided on what to do next. Jobs were plentiful then, and I had had terrific experience as a teaching assistant at Queen’s and Minnesota and made some money as a substitute teacher in Kingston, Ontario, schools, while I was at Queen’s. After a year and a half, I became head of a department, and I was pretty content. By then we had two children, but one day my wife Sue said to me, “This is not your destiny.” She was absolutely right, and we made the bold decision to move to Edmonton, more than 2000 km away, where the University of Alberta was starting a graduate Linguistics program.
Why Linguistics? How did that come into the discussion?
Well, by chance, I happened to take a course at Minnesota—I think it was called “English in Context” or something—and I found myself among dozens of Education majors. But the course was taught by Harold B. Allen, whom I had never heard of, and one of the textbooks was Noam Chomsky’s Syntactic Structures. Now this will be mind-blowing to people who know Harold as the author of The Linguistic Atlas of the Upper Midwest (1973), one of the jewels of the Linguistic Atlas of the United States and Canada (LAUSC), the gargantuan project that occupied American dialectologists for decades. But this was 1962 and Chomsky’s ideas were still being absorbed in the field. It didn’t take long for those dialectologists to despise Chomsky’s linguistics, but at that moment it was so new that Harold simply told the class that it was a new view of language studies. I don’t remember him ever referring to Chomsky’s ideas in the class, and my term paper for him definitely didn’t include any Chomsky references—it was about applying linguistic principles in classroom lessons, or something. But anyway, when I took up my teaching job in Ontario, I had this pristine copy of Syntactic Structures, and I spent lots of good hours figuring out what it meant. My first article in a professional journal was called “Linguistics in the English classroom” (1966), written while I was very much in the English classroom.
So Chomsky became your mentor? Or was it Harold B. Allen?
Well, Chomsky really. Though I got reacquainted with Harold Allen a few years later when I discussed LAUSC in my book Dialectology, with Peter Trudgill. My Ph.D. thesis at Alberta was called Focused Noun Phrases in English Syntax, and it was solidly in Chomsky’s transformational framework. The timing was perfect. I was hired at the University of Toronto in their new Linguistics program specifically to teach a course called “Introduction to Transformational Grammar.” Before the ink was dry on the thesis. And I taught it annually from my first year, 1970, until 1975 and sporadically after that, until 1982.
Ah, that’s why I don’t remember you teaching it. By the time I arrived at the University of Toronto in 1988, you were teaching courses called Canadian English, and Dialectology [which later became, more accurately, Urban Dialectology].
Yes, I went through a transformation of my own when I read an article by Bill Labov called “The linguistic variable as a structural unit” (1966b). I had just published a paper called “Canadian Raising” (1973), about the use of raised diphthongs /ʌj/ and /ʌw/ for /ɑj/ and /ɑw/ before voiceless consonants, the vowel that Americans often comment on when Canadians say words like “wife” and “about.” It was in a generative phonology framework, but I became acutely aware of variability in the real world. The old dialectologists said Canadian Raising stopped at the Rockies, but that was because no one had listened for it in British Columbia, and when we listened it was obviously there. So we also had the mystery about how such a specific phonological rule could spread so uniformly over thousands of kilometers. I called it “Canadian Raising” because of its salience in Canada but we knew from the start that it was a general phenomenon. And the name stuck. Since then, we have had studies of Canadian Raising in Georgia, Pennsylvania, northern California, the English Fens, as well as all parts of Canada (Chambers 2006).
So did you gravitate from generative linguistics to sociolinguistics by woodshedding? On your own?
For the most part. But I had a nice boost along the way. For my first sabbatical, 1976-77, I uprooted my family, now with three school-age children, to go to Reading University in England, where my classmate at Alberta, Paul Fletcher, was on faculty. He had introduced me to his colleague Peter Trudgill the summer before, and so I worked with Peter on variation and change. We signed the contract for our book Dialectology with Cambridge at the end of that year. It merged classical dialectology with sociolinguistic variation for the first time, and we’re proud to say the 1998 edition of the book is still in print.
So you went through a few years teaching “Transformational Grammar” and “Urban Dialectology,” a kind of split personality.
For sure, but this kind of schizoid specialization gave me, I think, a peculiar perspective. I stayed interested in Chomsky’s work, though less on his syntactic models than on his perspective, on the idea that our linguistic studies should ultimately shed light on the language faculty—on what our language behavior reveals about the processing mechanisms, on what is innate and what is learned. My research and teaching focused on Labovian concepts like change in progress and stylistic and linguistic variation. I naturally gravitated toward William Labov, applying his methods and amplifying his ideas (for instance, Chambers & Schilling 2013). One of my proudest publications is a kind of oral history of Labov’s accomplishments, “An Appreciation,” that came out the year he turned ninety (2017). My Chomskyan orientation gave me my own small niche, I think, in studying the social uses of language.
Chomsky famously or notoriously has not shown much interest in language variation, or in speech as utterances. I know you have been interested in the limits of variation: how far can people vary in their phonology or syntax before comprehension breaks down? Is that the kind of language variation issue that you consider to be rooted in Chomsky?
Any linguistic trait that appears to be essential rather than accidental will be rooted in the language faculty. One of the most productive areas of my research turned out to be what I called the Ethan Experience. It is named for the son of one of my first grad students. After graduating, he became a long-time colleague in the language section of the Anthropology department at the University of Toronto. He was born and raised in Czechoslovakia, now Czechia, and so was his wife, Ethan’s mother. Their English, naturally, is accented. They tend to have a tap /r/ intervocalically in words like “hurry” and tense vowels so that “red” sounds like “raid.” Ethan’s father, my colleague, happened to mention to me that Ethan at no time had features like those in his speech. I couldn’t help but take note because this [observation] came from a trained linguist and a meticulous scholar. I began looking into it and discovered another, even more mysterious fact. Ethan, a scholarly and articulate boy, did not realize—or better, was not consciously aware—that his parents spoke accented English, that is, had an accent different from his own English, until sometime around puberty. Though the evidence of their accents was copiously available to him from birth, abundantly available in his household, it only registered in him in his late grade school years. Well, these two observations—the absence of parental accent features even in isolated words, and the childhood ignorance that his parents’ English was accented—those two observations suggested that children come equipped with some kind of an accent filter, an internal mechanism that screens out accent features that are odd or eccentric in terms of the accent they are acquiring, that is, features that are different from the accent of their playmates and classmates. The Ethan Experience appears to be the common inheritance of kids who are raised in immigrant families, and thus it appears to be a feature of the language faculty.
So you started with the observations from Ethan’s father. How do you go about researching the Ethan Experience?
Good question. I had some doubts about it because my evidence was only anecdotal. To this day, I haven’t found a way of getting empirical or experimental evidence for it. But my doubts soon evaporated. At the time I was teaching a course in the Sociology department with about 150 students, and in a diverse city like Toronto you can be sure that more than half the students in the course will come from immigrant backgrounds. I actually knew that because my first assignment, in this class and even in my graduate seminars, was for each student to write a two- or three-page linguistic autobiography. Well, the first time I lectured on the Ethan Experience was in this sociology class, and I presented it much the way I have presented it here. And as I was going through it, I sensed the rumble of recognition, moving up the banked rows of the hall—a kind of physical stirring. For the students it was a kind of déjà vu, and for me it was just plain exciting. And when the students presented their year-end term papers, we received an inordinate number based on the Ethan Experience—many with ingenious titles like “Angelo’s Ethan Experience” and “The Maryam Experience.”
I know you have presented it many times since, and written about it.
Yes. Often with exhilarating results. In New Zealand, the professor Elizabeth Gordon made a fascinating comment at the end of my talk. She told us that she was sent at age six to a private school, and one day she invited two or three classmates to her house after school. Her father had an accent—not a foreign accent but I guess a working-class accent—and the little girls started making fun of his accent and laughing at the way he pronounced things. And Professor Gordon said, “Those snooty little girls robbed me of my Ethan Experience!” So we learned it works with class accents as well as interlanguage accents, and that it can be disrupted under excruciating social circumstances.
Did you work on other concepts that you deemed to be essential rather than accidental properties?
Yes. I published an article on them, a kind of summation, called “Sociolinguistics and the Language Faculty” (2005) in a volume celebrating the 50th anniversary of the Canadian Journal of Linguistics. I think of it as a sort of consummation of this aspect of my work, and for the next twelve years I used it as one of the texts in my graduate seminar. Besides the Ethan Experience, I talked about Sex-Based Variability and Vernacular Roots. In the former, consistent sociolinguistic differences between females and males have generally been explained as reflecting differences in gender roles. In places as different as working-class Belfast, where the Milroys did exemplary work (1978), and inner-city Detroit, where Wolfram did equally exemplary work (1969), men generally worked close to home and had local contacts, but women often worked further afield in domestic or service roles and their household duties gave them responsibility for contact with teachers, landlords, and other officials. The sociolinguistic differences were also sharply defined. The women commanded a larger linguistic repertoire than the men who were their husbands and brothers. So we inferred that it was the gender roles—the social distribution of chores—that gave women a wider repertoire of variants, greater command of styles, and made them leaders in changes that were standardizing. The men in both places consistently outscored the women in their use of local variants, like lookin’ and thinkin’ for looking and thinking, and double negatives, like We don’t need none. But at the same time, we were gathering similar results in middle-class communities where gender roles were not well-defined, in places like New York, Norwich, Boston, Vancouver, Philadelphia, Sydney, and North Toronto, my neighborhood. In those settings, women and men generally shared occupations, recreations, family outreach, and other roles. And sociolinguistic research showed almost identical results—here too women generally had larger repertoires of styles, showed more stylistic flexibility, and led in changes that were standardizing than the men who were their husbands and brothers. In New York City, for example, r-lessness was being replaced, so that heart and hard, traditionally haht [ha:t] and hahd [ha:d], were heard with [r], and women were the forerunners of the change. These results, coupled with the abundant evidence that women score higher than men in verbal sections of standardized tests—the differences are only about 0.5 standard deviation but they are persistent for huge samples of test-takers—it appears that women and men show innate processing asymmetries. That is, there is Sex-Based Variability. One of the theoretical implications is that gender roles, where they differ, do not determine linguistic differences, but rather that innate sociolinguistic differences determine the gender roles. If families have to make a choice about which parent will talk to the children’s teacher, it will almost certainly fall to the mother.
You are talking about sex and gender as a binary opposition, as female and male.
These results all came to light when social apprehension of sex and gender was fairly simplistic. But a happy result is that a recent grad student in my seminar started looking at transgender groups to see how they fit into communal norms (e.g., Baird & Scherz 2023), I am proud to say that they carried on the research in their Ph.D. and beyond, and that a community of scholars is forming with these more inclusive objectives.
You also mentioned the theory of Vernacular Roots as another candidate essential to the language faculty.
Yes. The evidence for Vernacular Roots is Anglocentric but that’s because English vernaculars have been studied way more widely than other vernaculars. The simple observation of six or more decades of sociolinguistic studies is that a subset of vernacular variables recur in all parts of the English-speaking world (2012). So, for instance, consonant cluster simplification, aroun’ for around, pos’ for post, copula absence as in Mama sick today, multiple negatives, Billie don’t want none, and a few other processes occur in the speech of the least educated speakers in almost every community that has been studied carefully. So, for instance, they show up in the stratified speech in Harlem, Appalachia, Newfoundland, Norwich, Sydney. Any attempt at explaining their global occurrence in terms of physical contact is doomed to failure. No explorers or adventurers carried these features from place to place in all parts of the globe. Schreier (2002) even found these features in the vernacular of Tristan da Cunha, probably the most isolated settlement on the globe. And to top it all off, these variables also occur in pidgins and Creoles, in interlanguage varieties, and in child language. Parents of loquacious children bewail the fact that their children go through a phase where they produce double negatives. I don’t want none, Nobody doesn’t want to play with me, and so on. The simple fact is that children from middle-class families in complex societies almost never hear multiple negatives. Or, of course, the other vernacular variants. They simply come out naturally, without a model. This is not a new observation. We have had acquisition theories that maintain that the child’s task in mastering adult language norms consists of suppressing these “primitive processes,” Braine’s (1974) term. Creolists, notably Bickerton (1984), explain the ubiquity of these features in English-based pidgins and Creoles as the result of a “language bioprogram,” an innate property of the language faculty. The theory of Vernacular Roots merely adds sociolinguistic evidence gathered over a period of about sixty years, when we worked on communal variation in urban as well as rural communities all around the globe. Those studies were inspired, of course, by Bill Labov’s Social Stratification of English in New York City (1966a), the foundation of our discipline.
Are there projects now looking further into these innate factors?
I receive comments and inquiries from people working on or interested in the Ethan Experience but not much about Sex-Based Variability or Vernacular Roots. For better or for worse, communal studies have given way to much more circumscribed projects. I am hearing a lot about teenage slang, occupational argot, incursions from internet talk, and so on. Broader communal studies seem to be fairly rare at the moment.
I imagine you see your Dialect Topography of Canada as part of that initiative for communal studies. What inspired the original Golden Horseshoe study [in 1991]? How did it differ from dialect studies back then?
The main difference was in the detailed personal information that we gathered as correlates of the linguistic variation we were finding. Labov’s studies in Martha’s Vineyard (1963) and especially New York City (1966a) showed that we needed more information than traditional dialect studies had provided. We called it sociolinguistic dialectology and outlined it in our Dialectology textbook. Besides age and place, the traditional staples, we asked for sex, occupation, place of birth and place of formative adolescent years, education, occupation, and parents’ occupations. The Golden Horseshoe covers 125 km around the western tip of Lake Ontario, and it houses one-fifth the population of Canada—an amazing fact when you consider that Canada occupies the second-largest national landmass in the world. We also surveyed the New York state border that abuts with the Golden Horseshoe. We gathered data on lexicon, phonetics, phonology, morphology, and syntax. The methodology at the time was a written questionnaire using what I now recognize as ingenious fill-ins and variant choices. It is a real storehouse of information.
How did the second Golden Horseshoe survey [2000-2001] come about?
Well, it was really because of the success of the first one. Colleagues across the country got interested, and I got grants for surveys in Montreal, Quebec City, the Ottawa Valley, the Eastern Townships, New Brunswick, and finally Vancouver when you settled there. With local linguists as directors, administering the Dialect Topography questionnaire, it took time. The directors were well-intentioned but busy, and the surveys were completed as time permitted. I realized too late that I should have applied for a million-dollar grant and covered the whole country in maybe a three-year span. By the time you joined my project as a director in 2000, the Golden Horseshoe data were about a decade old, so my research assistants undertook a new survey of that region to provide a temporal comparison. It enriches the data exponentially. The online Dialect Topography Survey (https://dialect.topography.artsci.utoronto.ca/) (Chambers & Pi 2005) which you made user-friendly when you became the technical director, finds users not only in Canada but further afield, especially from linguists looking to school their students in the correlation of social categories with linguistic variation. It has been a real boon.
I really enjoyed being a part of that project! Just thinking back, I remember fondly the times our team was sending out paper questionnaires, opening and recording returned surveys, and entering data in your office. What are some of the major findings from Dialect Topography?
There are regional differences, of course, but also a kind of basic homogeneity that pervades the vast territory. Both the variation and the homogeneity are copiously documented (for instance, Chambers 2006-2007). And we got lucky with the timing. I got interested in a class dialect called “Canadian Dainty.” Rex Wilson, a traditional dialectologist, was the first to mention it. He called it “Fredericton Dainty,” because his wife was from Fredericton, New Brunswick, as he was too, and her speech had certain features that were considered “genteel” or “polite.” She pronounced /ju/ in words like “news” and “Tuesday” where most Canadians dropped the /j/, “when” and “where” and “whale” with /hw/ where most Canadians dropped the /h/, and lexical items like “schedule” with /ʃ/ instead of /sk/, “tomato” with /ɑ/ instead of /ej/ (Chambers 2015). I had noticed that accent across the country and broadened the name to “Canadian Dainty.” In the 1990s, when we launched Dialect Topography, that accent was dying out and so we were able to document its demise across the country, feature by feature. By then, it was a relic of some of our oldest subjects. Historically, those features had come into Canada from a vast wave of British immigrants in the mid-nineteenth century. Those immigrants became teachers, bankers, clergy, and politicians, and they superimposed those British features on the native North American accent of the original settlers. They became markers of gentility and education. My parents remember their school teachers in the 1920s and 30s correcting students’ pronunciation of “tomato,” for instance, and promoting “rather” with /ɑ/ instead of native /æ/. It didn’t work on my parents, but obviously it did on a few of their upwardly mobile classmates. So Dialect Topography captures the death knell of one of the vestiges of Canada’s colonial roots.
You mentioned a “traditional dialectologist.” So did you find a community of scholars when you came along?
An interesting sidelight. There was some opposition from the older scholars. Dialect writings about Canada had been dominated in the early days by British-bred professors, and they sometimes lamented what they saw as the “Americanization” of Canadian English in the loss of those dainty features. But of course, that was exactly the opposite of what was really going on. Canadian English originated in the same waves of immigrants that had settled the United States. The political boundaries that created the two nations came more than a century after the original settlement. In shucking off the dainty features, Canadians were simply reverting to the native accent features and getting rid of the ones that had been superimposed on them by a much later immigration wave.
That accent sometimes gets noticed in old sound recordings by politicians and others, but is a museum piece now.
Right. Of course, a few features did persist, but we don’t perceive them as dainty features. I mean, like, the name of the last letter of the alphabet, “zed” in Canada and the rest of the world, but “zee” in the United States. A schoolteacher sent a letter to a Kingston newspaper in the 1850s complaining about Canadian children saying “zee.” Now it’s a shibboleth here. Canadians say “zed,” like the British. And “tap” for a faucet or spigot, which presumably came from British English. Our data shows how well-embedded those forms are. Generally, the data looks at dozens of variables that show up in English everywhere. It presents an open invitation for language scholars to compare their regional and standard accents and dialects and their social embedding. We have used the results to verify claims about the global spread of variants, including the replacement of past tense sneaked by snuck, a change in conjugation happening in Australia, England, North America, and everywhere. Interesting because it changes weak forms to strong, contrary to the trend in English for the entire millennium of its existence, when strong forms have become weak or “regular,” since time began.
I remember a splendid gala retirement conference for you [in 2005] but then you kept on teaching courses and publishing and traveling to conferences for several years after that. How did that work?
At the time it just seemed natural. I kept teaching my seminar Urban Dialectology, which had a dual credit for fourth year undergraduates and for grad students. It covered topics that otherwise didn’t get studied, and I always assumed that was why the department didn’t put me out to pasture. And I was still producing research, of course, and sitting on committees, and all the rest. Strange fact: the year after my official retirement, I became acting chair of the department for six months.
What were the topics that kept you involved?
Well, some of them were pet areas, like the Ethan Experience and Vernacular Roots, but others were exploratory, like the influence on language of social upheavals like Mass Media, Mass Literacy, and Cognitive Processing Styles. Some of the influence is sort of predictable, like the standardizing influence of mass literacy and the demise of regional accents, but now we had the wherewithal to demonstrate those effects empirically. And looking at mass media proved earth-shattering. This was at the dawn of the internet, and so we were mainly concerned with television and movies, and to the surprise of most people their influence on language proved negligible. As the years went on, evidence became overwhelming not only from the failure of TV accents in spreading different features but also from psycholinguistic evidence that infants absorbed no linguistic traits from TV instruction. Everywhere we looked, the conclusion became firmer that linguistic influence requires face-to-face interaction. Deaf parents, with the best intentions, for instance, sat their hearing child in front of the television set from birth expecting it to give him spoken skills, and the child absorbed absolutely no spoken language. He was adept at sign language from birth, but by the time teachers met him around age five, he had no spoken language at all, and his ability to acquire spoken language was seriously impaired.
Results like those go against common perceptions, and must have been a hard sell.
The evidence is pretty strong, but, yes, I shocked a lot of audiences spreading them around. I remember a talk in Stockholm when a young man put up his hand and said I was dead wrong because his impeccable English accent had been mastered by watching American movies daily all through his adolescence. But I pointed out he wasn’t watching movies, he was engrossing himself in a language laboratory. Not at all the same thing. Otherwise the movie screen is just a picture on a wall rather than something you interact with. Language learning requires interaction. Parents who jabber incessantly to their newborns are more likely to raise early talkers and highly social children, by every measurable test.
I remember that mass media seminar from the early days, but not the topic “Cognitive Processing Styles.”
I used that topic in my seminar starting in 2016, but I used it before that in a first-year course starting in 2013. I got lucky. My university invented general-interest courses called HUM199, for first-year students, nineteen-year-olds, with enrollment capped at twenty-four, so that incoming students could choose at least one course with small enrollments. Every department had to provide at least two of them. Some of my colleagues didn’t want to teach them because they were non-specialist courses. So I got to teach them, and they were a delight—bright young students with varied backgrounds and interests. One thing I did was give them a questionnaire developed at Cambridge called the AQ Test (Baron-Cohen, Wheelwright, Skinner, Martin & Clubley 2001). The students are asked to agree or disagree with fifty innocuous propositions such as “I am not very good at remembering phone numbers.” People who score high on the test are generally detail-oriented and those with low scores are broader generalists. So the scores really shed light on individual processing styles. I tallied their scores and showed the results for the class. They loved it. The results for my classes jibed with results for thousands of others. Men score higher than women, mathematicians score higher than literature students, and so on. My students shared age and education obviously, but their AQ scores over the years ranged from 6 to 36, and the differences are linked to personality traits such as extraversion, sociability, imagination, and the way they approach problems.
Interesting. What about academic implications?
To take a simple example, humanities students have lower scores than science students, showing that cognitive processing styles are a factor in the way students choose academic disciplines. Put simply, sciences are more likely to attract detail-oriented students. The results also have sociolinguistic implications. When we correlate women-in-science we find they have higher verbal fluency scores than men-in-science. This is not surprising because for over a century all kinds of standardized tests show that women outscore men statistically in all kinds of verbal tasks. What is surprising is that women-in-science also outscore men-in-humanities in verbal fluency. So here is evidence that Sex-Based Variability overrides cognitive processing. This result adds one more factor to the well-established sociolinguistic fact that women’s speech generally has fewer class markers than their husbands and brothers, and that they are leaders in linguistic changes that are standardizing.
Talking aloud about results like these is probably not your most popular lecture because there is a long trend among some social scientists to underplay or ignore results that show intrinsic differences between women and men.
You’re right, of course. Standardized tests such as the Graduate Record Exams stopped showing gender-related results sometime in the 1950s out of some kind fear that discrepancies might result in social or educational biases. The fact is clear, with literally millions of test-takers of GREs and a half-dozen other standardized tests, from all over the world, women outperform men in verbal tasks of all kinds and men outperform women in mathematical tasks of all kinds. It is not an accident of upbringing or environment but an innate cognitive bias.
How does that relate to linguistic issues?
In the grad seminar, we looked at ongoing sound changes and perceptual differences between low scorers and high scorers, and correlated the differences to the leaders of the change and the laggards. In the nonspecialist courses, we did simple tests of vocabulary size, and looked for correlates with cognitive processing scores. The results were consistently biased toward processing differences.
Have you talked about these results beyond the classroom?
I presented my results for the gender discrepancies in cognitive processing in several talks, most notably at the NWAV conference in 2016. I was expecting to engage in an animated discussion with a contingent of feminist sociolinguists in the room, but it didn’t happen. The results, I think, are too powerful. And the idea of suppressing such well-established results really emanates from a misconception about statistical significance. Statistical results are not results about individuals. There are obviously great women mathematicians and scientists, and just as obviously great men historians and poets. Individuals are not defined by statistical norms. The push right now in institutions like my university and many others to get women enrolled in STEM disciplines is not really an attempt to eliminate the cognitive bias—that is never going to happen. I see it as an attempt to broaden the perspective of the people who engage in scientific inquiry. Our academic disciplines can only become more useful if they are opened up to broader cognitive styles. There can only be positive results by looking at the physical universe from different angles, that is, through the prism of all kinds of cognitive styles.
That brings to mind something about your scholarship that has always impressed me for as long as I have known you. You don’t believe in academic closets. I mean, you are as celebrated as a jazz scholar as you are a sociolinguist. How do you balance that passion for music with your linguistics career?
Well, you’re a good example of that yourself, as a writer of sci-fi and a semanticist and a computer nerd with administrative skills. For me, I got interested in jazz as a high school kid, and I had a group of friends who reinforced it. At the time, in the 1950s and 1960s, popular music just seemed like a wasteland, full of three-chord tunes with a lot of gyrations. Somehow I got the crazy notion that I would some day write the biography of Miles Davis so that everyone would recognize him as a great poetic figure on his trumpet, and I collected clippings wherever I found them, and articles from obscure places. I started writing about jazz as an undergraduate, and I started publishing about linguistics not long after. When I moved to the University of Toronto, I got associated with Coda magazine, a serious jazz journal run by two English expats who had a terrific influence on the music in Canada. And then I was invited to write a review for the Globe and Mail, our national newspaper, and various other places, like The Literary Review of Canada and the Cambridge Encyclopedia. 1 I was also publishing on linguistics, and when my editor at the Globe found out he recruited me for reviews on language. I kept the two areas separate in what was then called the ad-all, the one-line endnote that identifies the reviewer. I figured if someone saw me dabbling in two disciplines they would dismiss me as a dilettante. And it worked for a while. One of our star linguistic students, in the 1990s, was also an amateur jazz drummer, and he was dumbfounded when he discovered that both reviewers were the same person.
I remember you bringing me and a few of my classmates to a live jazz performance once. I didn’t know you were a jazz scholar until then, either. You did eventually write that biography of Miles Davis, of course, the two-volume Milestones (1983, 1985).
Yes. My wife took our three kids away for a couple of weeks one summer to let me finish the first draft. It was the last thing I wrote before I got my first Apple, and on the typewriter I suppose I rewrote the several hundred manuscript pages two or three times each. Before the internet, of course, there were no easily accessible archives for old news, and my file of yellowing items gave me uncommon access to past events. For many years, I stewed over the use of my book as a kind of source book by others, almost always unacknowledged.
When Milestones was published, you were also publishing linguistics.
Yes, it was kind of a pact I made with myself. Linguistics is my vocation. Music criticism is an avocation. No one ever suggested, at least not to my face, that I was giving too much attention to music and not enough to linguistics. This year (2025), for instance, I have a book on Duke Ellington, the twentieth-century composer, but I also have linguistic articles and talks, and I am working on a kind of memoir on the amazing things we have learned about the social uses of language in my time.
Do you have any advice for students on how to pursue a career and another interest at the same time?
Well, I think everyone does it, to some extent. If the second interest amounts to an avocation rather than a hobby or a recreation, I think you have to keep your priorities clear. When I first started writing for Coda, one of the more prolific reviewers was finishing his Ph.D. in sociology. As soon as he graduated and got an academic position, he quit the jazz writing. For years, I thought I was supposed to have done that, like him. But the one time I spoke to him in the years afterward, he said he only wrote reviews for the free records, and as soon as he could buy all the records he wanted, he gave it up. So I felt better about keeping my head in both sides because my interest in jazz was motivated by a heartfelt desire to draw attention to this highly creative, challenging and esthetically rewarding musical genre that was sadly, maybe tragically, undervalued. That desire motivates me to this day.
If you don’t mind my saying so, I think that sort of proselytizing spirit sometimes creeps into your linguistics lectures.
You’re right, I’m sure. When I first started thinking about linguistics, I was amazed at how deeply into our being the language faculty is intertwined. We are born to talk. The urge to talk cannot be suppressed except by the grossest acts of criminal negligence. It is so embedded in our everyday experience that we fail to see it as a kind of miracle. We are full of admiration for people—legendary people—who devote their lives to becoming chess champions, or concert pianists, or nuclear physicists or stellar mathematicians. We think of those people as geniuses because to the rest of us even mediocre competence at chess, piano playing, physics, and mathematics is hard won. The learning curve is steep. But we all learn to manipulate the sounds and the syntax and the semantics of spoken language without any tuition whatsoever. We don’t take classes to learn our native languages. The complexities of language are every bit as complex as chess or music or physics or math—probably more so—and yet we master them as soon as we have the motor control to make sounds. We are born to talk. That is the language faculty at work. All we require is a stimulus, so that children raised in Japanese-speaking communities come out speaking Japanese, and those in English-speaking communities speak English. And when we discover the vacuity of mass media influence on speech, we discover that the stimulus has to be conveyed face-to-face, by human interaction. And then we discover that the language faculty provides a kind of filter so that immigrant children are steered toward learning the accents of their playmates without detouring through their parents’ accents along the way. The Ethan Experience seems like a miracle, and yet the evidence for it is all around us. Miracles they may be, but all these things are part of our daily experience.
Much of your research is rooted in your Canadian identity. You have even given some well-known concepts that name, like “Canadian Dainty” and “Canadian Raising.” Has that been a motivating factor in your work?
To some degree, I suppose it has. In my early days, I was disgruntled by the colonialism of some of the British linguists working in Canadian universities, and I suppose some of the Americans too. But really, in the end, my research involved Canadian accents and dialects in Canadian communities because they were what surrounded me. They were what I had to work with. When I spent three sabbatical years and many shorter sojourns in the south of England, I happened to find other Canadian families that had moved there. The children of these families came to my attention through my own children, who met them in their English schools. And so I could mount a long-term project on their progress in replacing Canadian English features like pronunciations of car and heart with [r] as in Canada or without it as in southern England. There were six children in two families, and when I first interviewed them they ranged in age from nine to seventeen, and I compared their speech to a group of native English classmates. The linguistic variables were completely familiar to me because of my time in the two places, and I suppose the discussion of those variants interested people in the two countries. But the real fruit of the research, as I saw it, was the dramatic effect of age on the results. The youngest subject acquired the features very quickly and the oldest almost not at all (Chambers 1992). But more fascinating was the only girl in the study, age 15 at the start, who was out of whack with the general trends—she scored much lower than the others in her age group. And when I finally asked her directly about it, she told me that she received tape recordings from her Canadian friends, and she practised talking like them. She was what I later classed as an “oddball,” and every sociolinguistic sample is likely to have some oddballs. In fact, I wrote about “oddballs and insiders” in my Sociolinguistic Theory (2009a), and I discovered that even oddballs fell into certain categories. Their presence is a nice reminder that we can discover statistical significance, but it does not account for individuals.
Technically, your Canadian samples were a matter of convenience, but you looked for results that would generalize.
Yeah, that’s a fair appraisal. One of the epigraphs in my book Sociolinguistic Theory is a curt statement from Robert Pirsig that says, “Data without generalization is just gossip.” And I was flattered when Bill Labov inquired about Pirsig because he found his statement a kind of mantra and often quoted it, he said, to his students. And so I told him it came from a novel by Robert Pirsig that almost nobody read (1991) but was a follow-up to Pirsig’s one famous book called Zen and the Art of Motorcycle Maintenance (1974). I like to think I gave Pirsig a niche in sociolinguists. His idea that data is only useful if it stimulates first principles is something that I have tried to live by. I hope that idea might have made an impression on my students—and on their students, and on and on.
That sounds like a kind of mantra of yours. And maybe a good place to stop. Thank you for taking the time with me.
Footnotes
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
