Language Log

« previous post | next post »

So began the abstract of Bill Labov's 1972 paper "Some principles of linguistic methodology":

Current difficulties in achieving intersubjective agreement in linguistics require attention to principles of methodology which consider sources of error and ways to eliminate them. The methodological assumptions and practices of various branches of linguistics are considered from the standpoint of the types of data gathered: texts, elicitations, intuitions and observations. Observations of the vernacular provide the most systematic basis for linguistic theory, but have been the most difficult kinds of data for linguists to obtain; techniques for solving the problems encountered are outlined. Intersubjective agreement is best reached by convergence of several kinds of data with complementary sources of error.

1972 was when I got out of the U.S. Army, after returning from Vietnam, and started graduate school at MIT. Leaving one war I didn't believe in, I found myself enlisted in another — because what Bill called "current difficulties in achieving intersubjective agreement" was a euphemistic way of describing the religious wars then consuming the field of linguistics. Some of the theological leaders had become established elsewhere (George Lakoff, Jim McCawley, and Paul Postal), but Haj Ross and Dave Perelmutter were starting guerilla campaigns within MIT, where Noam Chomsky and Morris Halle were the intellectual government.

I respected and even admired all of the people just named, and learned meaningful lessons from their work. Focusing on the analysis of prosody helped me to avoid the worst of the intellectual killing fields. But I saw no real way to stay neutral in that period of academic linguistics, any more than I could have in 17th-century central Europe.  Which is why I finished grad school in three years, and spent the next 15 years in an industrial research job.

As a dialectologist and student of Uriel Weinreich, Bill Labov began his career in a different tradition. When I was a grad student at MIT, work of that kind was covered in a course whose formal title (I think) was "The History of Linguistics", though everyone called it by the vernacular name "Bad Guys". In that context, Bill's ideas were not taken seriously, despite his attempts at engagement, clearly presented in his 1969 work "Deletion, and Inherent Variability of the English Copula". The abstract:

The following report presents some of the findings of several years' research into the relations between standard English and the non-standard English used by Negro speakers in various urban ghetto areas. The immediate subject is the status of the copula and auxiliary be in Negro non-standard English. The approach to the problem combines the methods of generative grammar and phonology with techniques for the quantitative analysis of systematic variation. The notion 'rule of grammar' is enlarged to include the formal treatment of inherent variation as a part of linguistic structure. Furthermore, a model is presented for the decisive solution of abstract questions of rule form and rule relations, based upon the direct study of linguistic behavior.

You can see more of this attempt at engagement in a paper on Bill's website, "Some Observations on the Foundation of Linguistics", about which he wrote

This is a paper that I wrote in 1987 but never published.  It deals with some general issues concerning the opposition between a mentalist (or idealist) approach to linguistics, and the materialist line taken by historical linguists, dialectologists and sociolinguists.  I would welcome any comments before revising it to fit the current situation.

That paper starts this way:

Linguistics is a relatively unified field of study, compared to many others.  Over the course of the long history of linguistic analysis, dating from Indian grammarians in the 4^th century B.C., we find emerging a high degree of consensus on the fundamental categories like sentence, phrase, noun, verb, vowel and consonant.  There is also a wide range of agreement on fundamental principles, and the concepts of language structure enunciated by Saussure (1922) and Bloomfield (1933) in the early part of this century are introduced to students in all elementary texts.  It is agreed that linguists are not interested in a given corpus of linguistic data in itself, but rather in the rules, system and faculty of language that enable speakers to produce that corpus.  It is agreed that language is a system of abstract categories that are mutually defined by their oppositions; that sentences cannot be understood by combination of the meaning of words, but only through the progressive combination of smaller phrases to produce a tree-like structure of immediate constituents.

At the same time, we can observe a profound division in the foundations of our discipline, that corresponds quite closely to the traditional philosophical opposition of idealism and materialism.  (In the linguistic literature, this opposition is sometimes described as 'mentalism' or 'rationalism' vs. 'empiricism').  The idealist approach is exemplified by generative grammar, as originated and developed by Chomsky (1957, 1965, 1981), and various other treatments that would account for the same data by parallel methods: generalized phrase structure grammar, lexical-functional grammar, and others.  The materialist position is exemplified by the practice current in phonetics, historical linguistics, and dialectology.  The principles of this position have been developed most explicitly in sociolinguistics, and in particular in the quantitative study of linguistic variation, which will be the basis of the discussion to follow.

The two approaches, idealist and materialist, differ sharply in their approaches to the foundations of the field: definition of language itself, the methods for gathering data and analyzing it, and the goals of linguistic activity.

Yesterday's otherwise excellent NYT obituary, "William Labov, Who Studied How Society Shapes Language, Dies at 97", gets the history of empirical studies in linguistics somewhat wrong:

Unlike most linguists before him, whose work was largely theoretical, he insisted on the importance of field work […]

Although Bill helped pioneer new methods for collecting and interpreting fieldwork data, there were many others who practiced and taught fieldwork before Bill entered the discipline. Edward Sapir did fieldwork on Wishram Chinook in 1905, under the supervision of Franz Boas, who taught fieldwork techniques in the late 19th century. And shortly after the first portable (Nagra) tape recorders became available, linguists began using them, as documented for example in W.S. Allen's 1956 paper "Structure and system in the Abaza verbal complex". 

Bill's key contribution was the idea that similar methods should be used to study linguists' own languages, not just to learn about "non-standard" versions, but also to get past the inaccuracies in our intuitions about how we talk (much less in what we tell others about those intuitions). And his work helped to define and establish the empirical study of linguistic variation, as linguistic fact, as an aspect of individual and social identity formation, and as the basis of language change.

But his main point in the cited papers was that abstract disputes among purely-theoretical linguists could (and should) gain substance by being tested against evidence from the way people actually talk in ordinary life. From the 1969 paper:

The data that we need cannot be collected from the closet, or from any library, public or private; fortunately for us, there is no shortage of native speakers of most languages, if we care to listen to them speak. Without such empirical data, we are now in the process of producing a great many well-formed theories with nothing to stand on: beautiful constructions with ugly feet. The test of simplicity – some internal evaluation measure which is in the continuous process of revision – has not satisfied many linguists to date. It is reasonable to ask that alternative analyses of the data on hand prove their value by pointing to further data which can conclusively resolve the alternatives proposed.

And from the 1972 paper:

Among the other innovations which Chomsky brought with him was a note of high seriousness in this respect. He is clearly interested in the structure of human language and the capacities of the mind which learns it, not in different ways of looking at the matter. Since Chomsky believes that linguistic theory is underdetermined by the data (1966), he proposes an internal evaluation measure, hopefully isomorphic with the one that the language learner actually uses. But the simplicity metric has had hard going; it is frequently misused by those trying to prove that they are right and that someone else is wrong, and there is some question as to whether it has actually resolved any important issues (Lakoff 1970).

Here linguistics is in a position to benefit from the example of the developed sciences. Scientific methodology can be thought of as the reverse procedure: trying to prove to yourself that you are wrong. That is, methodology is careful and conscientious search for error in one's own work, following Karl Popper's principle that the best theories are the easiest to disconfirm (1959). To be right means that you have finally abjectly, hopelessly failed to prove yourself wrong. It is dangerous to assign this responsibility to anyone else, for no one will have the same vested interest in this pursuit as you do.

This kind of methodological self-criticism leads to a continual refinement of our methods, introducing safeguards, reliability tests, cross-checks, typical of the scientific attitude which can profitably be practised even at the pre-scientific stage where we now find ourselves.

These lessons have been thoroughly assimilated by the sociolinguistic region of the field, but not so much by the rest of it, despite the fact that the internet has been added to Bill's 1969 list of sources, greatly facilitating the collection (and sharing) of representative samples for empirical research in linguistics. It's increasingly common to see empirical studies based on internet-available interviews, readings, podcasts, and so on — as well as recordings explicitly collected for analytic purposes. As mentioned earlier, Bill's tape archive of tens of thousands of hours of sociolinguistic interviews has been digitized, and will be published in a series of sections after transcription and appropriate anonymization.

A final personal note: Bill played a key role in persuading me to return to academia at Penn in 1990, by inviting me to teach a seminar as a guest the previous year.

 

December 27, 2024 @ 10:59 am · Filed by Mark Liberman under Philosophy of Language

Permalink