Richard Ohmann

Two interviews are the starting point of this chapter. I got them from a video project called "The Unemployment Tapes,"' designed to explore through talks with local people the human costs, the sources, and the possible cures of unemployment in an old industrial area of Connecticut. The Connecticut Council for the Humanities funded the project, and I was a consultant to it. At the time (fall 1978) I was also reading and thinking about class and language, and it occurred to me that interviews like these might be helpful. No interview can ever be a "natural" context for the speech of the person interviewed, but at least these respondents had no reason to think that their language was being observed, except in the incidental ways that we all observe one another's language. The interviewers were not linguists, psychologists, or sociologists. They were friendly, casual, young, not personally intimidating. The interviews followed no fixed schedule; in this they were more like conversations than are many experiments and interviews in sociolinguistic research. Yet the subject and the goals of the questioning remained fairly constant throughout.

So here are the transcripts, with some names and places disguised.

A COUPLE AT A SHOPPING MALL

Interviewer: I'd like to ask you if you have jobs right now.

Respondents: Yes. I: Have either of you ever been unemployed for any length of time?

R: No. 1: Well, would you say there was an unemployment problem in this area?

Man: Well we're new in the area. We just moved in a couple of months ago. From what I've been reading there is unemployment in the area.

Woman: I would say so. There are an awful lot of people going to Oakfield and Hill County to get jobs. They're not staying in the valley.

I: Do you have any ideas about what causes that problem?

M: I have no idea.

W. Not enough industry up here. A lot of industry is just leaving the area.

I: How come?

W: Taxes are too high? There's no rebate or anything else for them.

I: So if we give a tax break and some other breaks to business, then-

W: I would say that there's no reason for businesses to stay in Connecticut. They're not getting any benefits from it. It's cheaper to go down to the South and get cheap labor now.

I: What happens, when labor in the South matches labor up North?

W: They're going to have a problem.

I: Go overseas?

W: Possibly, yeah.

I: Then when labor overseas matches labor in the United States-

M: A vicious cycle.

I: What's the solution?

M: I don't know. If I knew I wouldn't be standing here.

I: Have you other thoughts on the subject?

W: I just wish they'd do something about it, that's all.

1: Who?

W: The government.

I: Could the government solve the problem?

M: I think they could make it a little easier. I don't think they could solve it. It's just going to-you're going to stop it here, it's going to start somewhere else. You're not going to be able to stop it. It's impossible. Like trying to stop war.

I: So it's part of the system?

M: I think so, yeah. I think it's part of life, but I think the government could make it easier.

1: Some people think a different system would solve the problem-

Both: I don't know.

1: A different economic system.

Both: I don't know.

THE MAYOR OF MILL TOWN

Interviewer: How do you think the high rate of unemployment has affected this community as a whole, in terms of its self image, in terms of its ability to deal with problems?

Respondent: Well, you know a very high percentage of unemployment is never a healthy condition, whether it's in Mill Town or anywhere else, and this lower Mill Valley region here has been pretty much plagued by high amounts of unemployment for at least fifteen to twenty years, and probably the greatest contributor to that would be the fact of how automation has taken over so much of the factory process that was once the main employer.

I: What are the other causes of unemployment, besides automation?

R: Well, I believe that automation is perhaps the chief cause of unemployment. Secondly, if we delve with other causes I would say it would be the lack of opportunity for the number of people that you have. We have a very densely populated area here, and like Mill Town with 6.2 square miles and you have over 21,000 people cramped into them, doesn't leave much space for industrial growth. In other words, we need to put our people to work. We need more facilities. We need more concerns here operating, businesses operating here, and we don't have the place to put them.

I: Whose responsibility is it to see that industry comes to, like, stop the high rate of unemployment? Do you see that as the responsibility of the government? Do you see it as the responsibility of business? Who puts pressure on business to do that? Whose responsibility is it?

R: Well, I don't think there is any one segment of society which, you're trying to point out, that is responsible. Like if it isn't there, that this is part of the responsibility of this particular segment. I think that it is very conducive to government to encourage industry in their area. I know I myself, as Mayor, am very anxious, and we have been working very hard, to fill up these remaining parcels we do have because it serves basically two purposes. It expands our tax base which makes life a little more comfortable for our citizens in terms of their tax bills, and secondly, it also in some effect provides more jobs which lowers that unemployment rate at least somewhat.

I: Do you think that the federal government should play a major role in bad economic times, as it is doing with CETA?

R: Well, certainly. I think that if you look at the entire history of our country, that it has always been the federal government that has come to the rescue. Take the great depression and all the federal programs that we used to bail it out. What you are really doing is, you stimulate the economy by priming up the pump and throwing money into the economy. That's-but by giving these people salaries and positions and all, they are going out and spending money, which gives business, the private sector, more of a stimulus, because they've got money coming in, they have the cash flow, and you hope for expansion.

I: Does that ever make you think about the economic system that we have, that it always has to be fed?

R: I think unfortunately it will always have to be fed. The government the government-the federal government-the government in general are big partners in the private sector. I think they really prime a lot of money into them that, you know, makes things happen.

I: And you think that is the way it should be?

R: I don't think it's the way it should be. It would be wonderful to have private enterprise exist on their own, without any regulation or any help from government, but I don't feel that it is workable.

I: Why not?

R: Oh, for many, many reasons. I don't think first of all that private well, you just take private enterprise as it is, and what if it wasn't regulated? I mean you take, again-going back historically, Standard oil and all the great trusts that were brought together there in the early 1900s where a few people were making millions and millions of dollars-and which were like trillions today-and the majority of the people in the country, the standard of living was very, very low. It was when government came in and started to regulate the amount of profits that these people could make and to really decentralize the main business interest that people started to get a better standard of living. The unionism thing was all part of the entire movement, I believe, which created a better standard of living, and this was all done through government legislation.

Now I want to present in schematic form some rather sharp contrasts between the way the mayor talks and the way the man and woman talk, leaving aside all judgments about effectiveness, clarity, and intelligence.

A. Length and Complexity

1. The responses are much shorter in interview I. So are the sentences: there is no independent syntactic unit of more than sixteen words in I: six of the sentences in II are longer than thirty words.

2. There is little coordination and almost no subordination in I, except in sentences beginning "I think," "I would say," and so forth. There is much of both in II. For instance, in the sentence beginning "That's-but by giving. . . " the main clause is preceded by a gerund phrase and followed by a relative clause with an embedded appositive, then an adverbial clause that contains another appositive-like structure (" they have the cash flow"); there are several coordinate constructions along the way.

3. There are few explicit causal or logical connections in I, and many in II.

B. Modifiers

There are few adjectives and adverbs in I, and those mainly of degree. Modifiers are many and various in II, including derived adjectives ("industrial") and nouns used as modifiers ("unemployment rate").

C. Abstraction

There are few abstract nouns in I, many in II. Those in I appear mainly in simple constructions with the verb be, and are unrelated to one another: "There is unemployment in the area"; "Taxes are too high"; "There's no rebate. " The abstract nouns in II appear in a variety of syntactic positions, and are often related syntactically and conceptually to one another. For example, in his fourth answer the mayor connects all the following nouns within a single sentence: "economy," "salaries," "positions," "business," "sector," "stimulus," "cash flow," and "expansion."

Reference to Context

The man and woman refer only a few times to the context of the discussion: "Oakfield," "Hill County," "down to the South," "Connecticut." The mayor not only anchors the discussion geographically to Mill 'own with its 21,000 people in only six square miles, but also gives it a context in the social system (the economy, the government, etc.) and in history (the last fifteen to twenty years, the depression, the early 1900s). Note also that interview I includes one exophoric pronoun (a pronoun with no antecedent in the discourse): "I just wish they'd do something." There are none in II.

E. Reference to the Discourse Itself

There is virtually none in I, other than expressions of uncertainty, like "I think" and "I don't know." The mayor uses such constructions, and also refers to the discourse in at least four other ways:

1. He comments on the interviewer's question. For instance, when he begins his first answer, "Well, you know a very high percentage of unemployment is never a healthy condition," he in effect says, "That's a silly question," by stating a general principle that covers the situation and that should be obvious to anyone. Compare this to the beginning of his fourth answer.

2. He implicitly rejects the question: when asked who is responsible for reducing unemployment, he denies the presupposition that some one part of society is. When asked if the need for Keynesian measures makes him "think about the economic system," he simply reiterates the need for Keynesian measures, declining to answer the question (answer 5). When asked, next, if he thinks that is the way it should be, he does respond, but then goes on to show that the question is infelicitous-you cannot properly ask if X should be the case when X must be the case.

3. He comments reflexively on his own terms and statements: "In other words"; "I mean"; "Like if it isn't there"; "business, the private sector"; "money . . . the cash flow"; "the government-the federal government-the government in general."

4. He makes new starts in the middle of a sentence, indicating that he has reconsidered and thought of a better way to proceed: "That's-but by giving these people salaries"; "I don't think first of all that private-well you just take private enterprise"; "I mean you take again-going back historically." (The man in I does this once: "It's just going to-you're going to stop it here, it's going to start somewhere else.")

Contrasts like these run through all the unemployment tapes I have studied. People on the street, picked out as ordinary workers or perhaps unemployed people, were asked the same kinds of questions about unemployment as were officials, businessmen, and specialists. Speakers from the first group did not elaborate, rank, or expand their ideas much, did not make many distinctions, made few logical and causal connections, did not develop abstract ideas, did not relate their words very explicitly to context, and referred little to the discourse itself in a critical or metalinguistic way. Speakers from the second group rated high on all these measures.

In the last two decades, the sorts of contrast that emerge in these two interviews have drawn a lot of attention, especially in Britain. There a group of sociologists and linguists inspired and led by Basil Bernstein has done very extensive research on differences between working-class and middle-class speech. And Bernstein's concepts of "restricted" and "elaborated" codes are now firmly planted in the center of this intellectual terrain-much respected and much criticized.

According to Bernstein and his colleagues, the elaborated code of the middle class runs more to subordination and modification than the restricted code of the working class. It includes more adjectives, adverbs, prepositions, complex verbs. It facilitates distinctions of all sorts, in particular logical ones. Elaborated code users distance themselves more from the immediate situation and from the content of their talk, through abstraction, through passives, through expressions of probability, through suppositions ("I think"), through questions and refusals to commit themselves quickly to definite interpretations of ambiguous experience. The elaborated code allows or encourages more individuation of response and more reflection on language itself. Restricted code users are more bound to the local, concrete situation. Much of their meaning is implicit-dependent on prior understandings of the context. (Hence they do not refer so explicitly to the context; exophoric pronouns are an extreme example.) In Bernstein's own words, the restricted code emphasizes "the communal rather than the individual, the concrete rather than the abstract, substance rather than the elaboration of processes, the here and now rather than exploration of motives and intentions, and positional rather than personalized forms of social control."' Again,

By now it should be clear that the analysis assigns profound social values to the two codes and that it has wide political implications. Bernstein himself does not dwell on these, but does hint at the depressing circularity suggested by his findings. For instance, "One of the effects of the class system is to limit access to elaborated codes." In another article he argues that "the genes of social class may well be carried less through a genetic code but far more through a communication code that social class itself promotes" (p. 143). Putting these statements together, we can derive this principle of social continuity: the class system sorts people into elaborated and restricted code users; the codes perpetuate the class system.

The moral is drawn more fully in The Politics of Communication by Claus Mueller, who draws on Bernstein's research as well as many other studies of class, child rearing, language, and belief. Mueller argues that in advanced capitalist societies, a social order marked by severe inequality and the powerlessness of most people is sustained and legitimated, not so much by coercion (the police and the army) or even by manipulation (propaganda, censorship), as by

He thinks this impasse especially intractable because the codes are passed on in the home to very young children. He agrees with Bernstein that class differences in child rearing are decisive, and that working-class parents block the development of linguistic autonomy in their children through strategies of teaching and discipline that call on authority more than on reasoning and exploration. If this is so, neither school nor "Sesame Street" could easily undo the damage, even if school, for working-class kids, were an open and supportive institution. Mueller concludes that the only likely challenge to the legitimacy of the political system in countries like ours will come, not from the traditional working class, but from the intellectual and cultural "strata."°

Now I find myself in one of those strata and trying to challenge the legitimacy of power in the United States. For people in that position, marxism has long been the richest source of political practice. Few marxists would, even now, join with Mueller in giving up on the proletariat as the revolutionary class. A more common marxian position is that, indeed, some intellectuals defect from the capitalist social order, but they do not become thereby a revolutionary class or group in themselves; on the contrary, their task is to work politically and educationally (the two are really the same) within the proletariat, which is the leading force for revolutionary change. Marxism itself is, in this view, the system of ideas that derives from the experience of the working class-no proletariat, no Capital. But intellectuals must help give it voice, as Marx did, and so play at least a small role in the articulation of working-class consciousness.

If Bernstein and Mueller are right, however, there is a barrier to this task higher even than those raised by bourgeois control of police, schools, and media. Marxism as a system of ideas abstracts a great deal from local contexts and immediate experience; it cannot be given voice in a restricted code. If I may exaggerate, a bit, the implications of Bernstein's and Mueller's position: the revolutionary class in advanced capitalist societies, the class with the experience of exploitation and powerlessness and with the motive for socialism, has been excluded from the concepts and the very linguistic structures that must be used to express that experience and develop the institutions that will lead toward socialism. This would make the job of the revolutionary intellectual truly herculean. As I put it a while back, in the form of a question to myself and other radicals: "When we try to communicate to workers a socialist understanding of things, must we think of our task as, in part, making up a cognitive and linguistic deficit?"5 I'll couch this discussion in marxian terms to make clear my own commitment. But marxists are by no means the only ones who should be concerned about the social implications of Bernstein's research. Anyone who favors social equality, democracy, and a politically competent people, and does not see much of these in our society, should feel in these questions of class and language an urgency. For if Bernstein and Mueller are right, those who have available only a restricted code can do little more than passively observe the shaping of the future. Worse, there is probably as much potential for fascism as for democracy in the working class, since people raised by rule and nurtured in restricted codes tend "to abide by the prescriptions of external authorities."

I want now to turn a critical eye on the picture I have just drawn.' Bernstein and Mueller, whom I have allowed to stand for many others, advance an argument that has a hypnotic power. Once its underlying concepts and premises are allowed, the research leads inexorably to the conclusions I have sketched and to the political pessimism they sanction. But those concepts and premises are extremely problematic. It is my own belief that they are so defective as to invalidate the conclusions drawn from Bernstein's research, as well as the political interpretation of those conclusions that Mueller and others have offered. The trouble begins right at the beginning, with the concepts of class and code. .

Take class. The idea of class that both Bernstein and Mueller deploy is drawn from mainstream social science. It is basically a heuristic concept, obtained by calibrating one or more such fac tors as income, education, and occupation, and selecting a cluster of them as convenient or experimentally handy. They may then be correlated with other variables: speech patterns, IQ, lifespan, child-rearing practices, beliefs, voting behavior, height, hair length, literally anything that can somehow be measured. Plainly there is no reason that any of these other factors might not be substituted for one of the original three, if doing so produced "better" correlations. Such a shift in definition would of course change the actual membership of each class, but that would not matter because the classes, within this framework, have no reality other than a heuristic one for the sociologist manipulating data. The unreality of this scheme is reflected in the fact that it can lead to three or six or any number (nine was the one in favor when I took sociology in college) of classes, which are no more than "strata situated along a continuum" (Mueller, p. 45), artificially segmented to the convenience-again-of the experimenter or theorist. Since this continuum of groups has no intrinsic relation to the structure of society or its historical evolution, correlations obtained within it do not much illuminate the way society works, but leave us within a closed explanatory circle where nothing has priority over anything else. There is no way to tell, for instance, whether occupational status explains speech patterns or vice versa-or both.'

A marxian idea of class is a much better foundation for discussion of these issues. Without going into the complexities of this subject, let me note that when we ground class in basic relations of production, the difficulties I've just listed disappear, and there is at least a chance of connecting class to something like language in a way that explains how society works, how it reproduces itself, and how it changes.

Note first that from this perspective Bernstein and Mueller are not talking of two classes, but mainly of two parts of the working class. Almost everyone included in both domains must sell his or her labor power in order to live, having no significant capital. (The exceptions: some independent professionals and small business people, apparently included in Bernstein's and Mueller's middle class.) The main distinction between the two is that most of the people they call working class sell their power to execute routine tasks at someone else's command-physical labor power, in effect-while those they call middle class sell their power of conception-mental labor power-as well. Bernstein's working class (let me use the shorthand, "physical workers," for the moment) is limited mainly to executing someone else's plan, while his middle class (I'll say "mental workers") has at least a small role in the planning itself.

Once the discussion is so grounded, and for all the immense complexities that remain, Bernstein's results make a good deal of initial sense. For instance, his account of restricted and elaborated codes:

-Restricted, context bound; elaborated, context free. At work the context is almost entirely provided for physical workers by their bosses; mental workers can do more to shape the context of their work.

-Restricted, concrete; elaborated, abstract. At work, physical workers manipulate things, while mental workers manipulate ideas, numbers, and so forth.

-Restricted code, predictable; elaborated, more individuated. Physical workers are not paid to vary from set routines; employers value to some extent the individuality and creativity of mental workers.

-Restricted code, few hesitations, expressions of uncertainity, or metalinguistic" references to the discourse; elaborated code, high on all these dimensions. Physical workers are limited to executing someone else's plan; mental workers have some responsibility for planning-precision and critical awareness in speech are important for them.

-Restricted code, simple in syntax; elaborated code complex, with much subordination, logical tissue, modification, and so forth. Physical workers are not asked to make many connections, see broad relationships, understand the larger processes in which their work is embedded; the reverse is true for many mental workers.

Of course, young children-the subjects of much of Bernstein's research-do not work in factories or law firms. For the hypothesis I am sketching out to have any plausibility, it would have to derive "socialization" practices from the total experience of classes and subclasses in production. Bernstein findings do point to such a connection, as a few samples will suggest:

-"Working-class" discipline of children stresses results, "middleclass" discipline, intentions. This corresponds to the distinction between execution and conception at work.

-"Working-class" parents teach skills; "middle-class" parents teach principles. This corresponds to what will be expected of the children later in their jobs.

-"Working-class" parents use "positional controls" (e.g., coercion: or, "Do it because your father says to do it"); "middle-class" parents favor "personal controls" (e.g., "If you don't clean up your room, your mother will have to do it, and she's very tired today"). Physical workers must learn to take orders without asking why. Mental workers need to know something of the rationale for what they do on the job.

From this pairing up of findings and causal hypotheses (overly schematic, to be sure), a clear picture emerges. A class builds its life on its role in production. The social relations it experiences there may be embedded in its linguistic codes, and carried over into the kind of training it gives its children at home. Now this is a very simple hypothesis, and may or may not turn out to be right. My point is that this approach to class at least permits us to work toward an explanation, in social structure and historical process, -of the ways people talk, rather than leaving us enclosed in a limitless circle or measurable attributes, none with causal priority over the others. It roots language and consciousness in material life.

But within a marxian framework, this is still insufficient. I have been using a notion of class that is structural and static. In this way of thinking, a class is defined by its relationship to the means of production and to other classes. The concept is incomplete unless joined to one grounded in the continuous movement of history. In this second view I do not simply and eternally belong to the professional and intellectual portion of the working class. Rather, in all my doings from day to day, I and the people I mingle with and am affected by constantly create my class position. As for instance, I confirm it by writing in this way to these readers , by continuing to work with my mind and my mouth more than with my hands, by failing to get rich enough so that I might if I liked stop working altogether, by sending my children to college so that they can work with their minds and probably also not get rich, and so on. From this perspective, class is not a permanent fact, but something that continually happens."

As soon as we look at it in this way, a still different relationship of class and language comes into focus. My way of talking, whether "caused" by my class or not, is one of the important means by which I, in my relations with other people, re-create my class, confirm it, perhaps alter it. When I talk I mark myself, for others, as some kind of intellectual worker. Learning to talk that way was of course one prerequisite to securing myself a place in intellectual work. I might add that although my father did similar work, I don't believe I learned my code at home so much as in various acculturating institutions along the way to professordom.

Just as my father did not talk like an intellectual in the nursery, neither do I talk to my children as I talk to my colleagues. And I speak differently again when I'm lecturing in class, when I'm trying to explain at the electrical supplies store what kind of switch I need, and when I am a witness in court. To follow this line of

thought is to call into question the second main term-"code"-in Bernstein's equation. He does allow that speakers of the elaborated code also can use the restricted code. I think it's more complicated than that. I don't "have" a code the way I have my Ford Escort out in the garage, to use whenever I go somewhere. If analogies are any use, a better one is probably to my wardrobe, from which I select in order to present myself in various ways on various occasions. Although there are clothes and ways of talking in which I feel most at home, sometimes I am not "at home," and I can confidently dress and talk to be comfortable in a variety of situations (though not all). But even that analogy won't do. Unlike a car or a wardrobe, a code has no material existence in history, except as it is ceaselessly re-created when people speak. The same is true of a class, seen in the second marxian perspective. And of course when we re-create a code by speaking, we almost always do so in collaboration with other people, and never in a setting that is socially neutral. Whenever we talk we do so within a nexus of power, status, intimacy or remoteness, family roles, institutional roles, designs r on one another, and so on. It is hardly an exaggeration to say that the whole of society as I know it is present in or impinges on my every verbal transaction.

Now this position, which I have laid out very generally and will try to make more precise later, is a close neighbor t one of the cardinal principles of sociolinguistics. I mean the idea that for all of us speech is variable. Sociolinguistics speak of "variable rules," meaning, for instance, the frequency with which a worker will say "he don't" as against "he doesn't," in various situations. Along with our grammatical abilities, we also tacitly know what counts as a timely and appropriate utterance at different stages of a speech situation, as well as how to relate through speaking to people of various sorts (bosses, priests, kids), for various purposes (to buy a hamburger or get a job), and in various genres (story telling, arguing, answering questions). It begins to seem very hard to disentangle a single code from the dozens of ways that speech and society impinge on each other. The way one speaks at any time is strongly influenced by the whole surrounding network of social circumstance, more than by relatively remote things like income, the job status of one's parents, or the number of years one spent in school.

This perspective applies not only to such constructs as "code" and "vernacular," but even to the individual word itself. To quote V. N. Volosinov:

The sign, both in its form and in its meaning, is in Volosinov's view "ideological": to simplify, it projects consciousness on reality, and consciousness, in turn derives from the organization of society. Since different classes have different consciousness, and since for the most part they use the same signs in communication, in each sign different ideologies intersect. The sign itself is, in Volosinov's words, "an arena of the class struggle" (p. 23).

That is a rather dramatic way to put it, but I think the point is right. Many words have alternative pronunciations that carry a marker of prestige or class. There is Labov's example of postvocalic /r/ in New York; there is dese versus these; in My Fair Lady there is "the /rayn/ in /spayn/"versus "the /reyn/ in /speyn/."When teachers "correct" kids on such matters, they comment on the kids' class. When a speaker who normally drops the r or says dese or /rayn/ talks with someone from a higher class or in a position of authority, he or she may shift to the more prestigious pronunciation or defiantly stand his or her linguistic and social ground. (I'm as good as

r-you are, and I'll talk the way that's natural for me.) Such encounters may not be the heart of the class struggle, but they surely express conflict that is rooted in class.

As for meaning, consider the use of the word industry by the woman in interview I, and by the mayor of Mill Town. For her, industry is concrete (factories and machines), but also a remote and uncontrollable condition like the weather: "Not enough industry up here." Hence, not enough jobs-a fact of life. When the mayor says government should "encourage industry in their area," he speaks as a member of government who has some modest influence over the movements of industry. For him, industry is real people with interests that he can address, and with whom he is involved as other than just a seller of his labor power. Industry is the same sign in both sentences, but used in ideologically contrasting ways. For her, industry is an alien force, for him a set of valuable if explosive allies whom he wishes to help in their project of development. Both agree that it is good to have industry in the valley, but their political involvements in the matter are quite different. And they express their social situations in the ways they use the word.

It should be clear by this point, if you accept the argument, that Bernstein and Mueller ground their conclusions on a sociolinguistic method that in turn derives from damagingly static ideas of code and class, and of the links between code and class. In effect, Bernstein seeks to correlate two things, neither of which can be abstracted without distortion from the stream of social interaction, and both of which are incessantly re-created in every encounter.

In other words, we are dealing here with a phenomenon that is dialectical as well as dialectal. The power relations of a society permeate speech and shape it, while speech reproduces or challenges the power relations of the society. Please don't take me to be saying that class is only an artifact of the ways we talk to one another. But it would be equally wrong to say that the ways we talk are only an artifact of our class. The two are embedded in each other. Speech takes place in society and society also "takes place" in our verbal transactions with one another-which of course are,; inseparable from the economic and other transactions we enter. I have made this point a number of times now, insisting so much because it is important and, for social scientists at least, counterintuitive.

But it is time to turn from theory back to the interviews and show how they may be seen afresh from this outlook. To begin with, both interviews explore the same subject, and the questions asked are quite similar in content. Nonetheless, the interview with the anonymous man and woman is in significant ways a quite different event from the interview with the mayor. One takes place in the street outside a shopping mall: it is impromptu. The other takes place in the mayor's office, by appointment. He has had time to prepare his thoughts. The mayor is interviewed because he is who he is; the specific identities and positions of the man and

woman are of no consequence. They are selected precisely because they are representative, part of a mass. Again, the mayor is an expert on the economy of the valley. That is part of his job, while the man and woman suddenly find themselves on an intellectual terrain that is unfamiliar. Finally, the mayor is used to such encounters, and the man and woman are not. We may guess that the video equipment is at least a bit intimidating for them; it must make them feel that they are being observed, tested. Working with television is a more familiar challenge for the mayor. In a way, television is an extension of his office and his power, something he can use to his own ends if he is skillful. The television people are there by his sufferance and on his timetable: he begins the interview as in some ways their superior. So although the issues remain almost constant through the two interviews, social relations do not.

As you might expect, the participants also create their relationships differently in the two interviews, through the ways they talk to each other. For instance, the interviewer in I begins with four yes-no questions. This is a way of getting out some basic information, but it also establishes a particular social relationship. A yes-no question strictly limits the form of its answer. The questioner sets up a tight cognitive paradigm, asking only for some information to complete it. Of course the respondent may decline to play the game this way, but to do so requires a breach of decorum. By contrast, a wh- question frequently gives the respondent a kind of carte blanche as to how detailed and lengthy the answer may be. The three wh- questions with which interview II begins all accord the mayor that kind of freedom. On top of that, the first two questions to the man and woman request personal information. They do so in a respectful way; nonetheless, one condition for a felicitous question is that the questioner has the right to ask. If that right is not given by intimacy, it is usually given by virtue of some official purpose as to bank officers considering loan applicants or to census workers. However gently, the sidewalk interviewer assumes such a prerogative. (Note that when he shifts to an impersonal question about unemployment, the man feels constrained to preface his answer with more personal information, by way of excuse.) The first question to the mayor, on the other hand, is not only general and impersonal but assumes much knowledge on his part. It positions him as an expert, someone whose opinion is worth knowing, in detail and on a highly complex subject. It is an invitation to expatiate.

These differences arise from no bias of the interviewers. On the contrary, since I now them I am confident in saying what I believe is also implicit in the interviews: that their sympathies lay more with the workers and unemployed people they met than with the managers, officials, industrialists, and bureaucrats. The differ-

Ence stem from the speech situations themselves, and from moves the participants make that accept and confirm those situations. As a result of these moves the first interview proceeds somewhat like a quiz. When the interviewer shifts to wh- questions after a bit, it seems as if he is testing the man and woman. They respond like schoolchildren being drawn out against their will by an insistent teacher who is asking them to have opinions and ideas so that they may be judged. (Note especially the series of leading questions on cheap labor-a kind of catechism.) In interview II, by contrast, when the interviewer shifts to yes-no questions, their aim is to challenge and explore views that the mayor has already expressed. His position has itself become the subject of the discussion, and is in this way dignified. The interviewer is pressing him, as a serious antagonist.

Perhaps that is enough to establish my claim about the social dynamics of the interviews. One cannot know for sure how these people speak at other times, but the contrasts I have mentioned are certainly sufficient to have elicited a restricted code from the couple and an elaborated code from the mayor. Let me return, somewhat speculatively, to my initial analysis of the interviews, looking at Bernstein's categories from this new perspective:

The short responses and short, simple sentences of the man and woman are obedient answers of unprepared people who feel themselves tested and perhaps judged. Why not, with the camera looking on, a the questioner who clearly knows more than they about the subject? Their task, as I see it, is to avoid exposure or humiliation, to avoid the risk of saying something purely foolish. They take their leads from the interviewer, and try to sense from his reaction whether they have said the right thing. For them, "I don't know" is the ultimate defensive strategy, since it is at least not a wrong answer. The mayor is invited to dilate upon his subject; he does so, and in the complex (though often vapid) sentences appropriate to that task.

B. Modifiers

The man and woman are not being asked to individuate their opinions, to shade, specify, qualify. "Do you have any ideas about what causes the problem?" The interviewer is asking them to take a stab at it. A short, tentative answer is the natural response. But the mayor is invited to discourse on the "community as a whole," its "self-image," its ability to deal with problems." He could hardly take on this huge and complex subject without qualifying his answer along the way. Also, because of who he is, his words are important. They will go on record. They had better be measured and circumspect.

C. Abstraction

For the man and woman, terms like "industry," "taxes," "rebate," and "cheap labor" are hand-me-downs from TV, the newspapers, casual conversation about distant matters out of their control. They produce these terms as part of their role in the quiz, but the terms are alienated. The man and woman have nothing to back them up with, no way to relate them conceptually to one another and to reality. For the mayor, abstractions about the economy are rooted in his daily work: in technical reports bearing on decisions he must make, in talk with advisors, the Chamber of Commerce, state and federal bureaucrats. This is not to say that his account of unemployment is better than that of the man and woman. In my own view, automation is a shallow cause, and the lack of acreage in Mill Town an empty one, while the woman is right on target in pointing to the free flow of capital in pursuit of cheap labor. One may talk flaccid nonsense in elaborated codes, and hard truth in restricted ones; as the mayor's speech well illustrates, an elaborated code may serve as a bureaucratic smoke screen. At the same time, abstractions are a verbal medium the mayor is used to and works within. He manipulates them freel and v luntaril rather than tentatively and with an air o tsi omeone~e's language, under pressure. They are an instrument of power for him in this situation, and a token of powerlessness for the man and woman.

D. Reference to Context

The subjecct of the interview's questions belongs to the mayor's field of action. They already have a context in his work and thought. For the man and woman, government, the movements of corporations, unemployment, and history in the large sense are distant forces and events, not because of any cognitive or linguistic deficit, but just in that the couple are connected to such matters only through activities like drawing a wage, buying commodities, and voting, which relate them to the historical context only in fragmented and isolating ways-always the mass media reinforce.

E. Reference to the Discourse Itself

The mayor's self-reflexive expressions, his comments on the interviewer's questions, his refusal to accept their premises, his new starts, all reflect the mayor's sense that he is in charge of the conversation. The questions are not, as he sees it, a form of power over him and a cage within which he must submissively remain. He can establish the terms and set the ground rules, up to a point. And what he says is important enough to warrant his

taking pains, finding just the right formulation. (It may also be relevant to mention that the questioner in this interview is a woman.)

In all these ways the interviews embed power relations and speech conventions that existed prior to the encounters. But this is not to say that the speakers' codes reflect only the social relations that previously obtained. Choice is available at every point: note, for instance, how the mayor takes over leadership of the interview by volunteering the chief cause of unemployment without being asked, how he changes the terms of the questions, and so on. No law prevents the man and woman from doing likewise (though the power relations they walk into have nearly the force of law). The participants create the social relations of each encounter, in addition to inheriting them. In so doing they reproduce society. By such tiny increments is class made and remade.

If my argument is sound, then, a Bernsteinian explanation of these interviews badly misrepresents the social forces at work in them, assigning to static "class" differences in speech that express dynamic and changeable power relations." More broadly, I have argued that this mistake follows from serious misconceptions of class and code. More the pity, because (1) Bernstein clearly meant it to serve the working class; (2) it has been highly influential, especially in Britain; and (3) the pedagogical inference drawn from it has generally been that we should teach elaborated codes to working-class kids, within the customary social relations of the school. Instead, I think the educational moral is roughly that of the 1960s reform movements, now much condemned: students should have as much responsibility as possible for their own educations. The habits of expressive power come with actual shared power, not with computerized instruction in sentence-combining or with a back-to-basics movement that would freeze students' language into someone else's rules, imposed from without. Respect the linguistic resources students have; make language a vehicle for achievement of real political and personal aims.

Finally, Mueller's political pessimism is justified only if we assume, as many leftists do (myself included at the time I first addressed these questions), that political consciousness is fixed, either at home in infancy and childhood or, even more deeply than that, by gross structural features of the society-if we assume that workers cannot become equal communicators and political participants step by step, and through action, but only by understanding, in a kind of conversion experience, the fundamental concepts of marxism. Movements toward worker self-management, co-ops, progressive credit unions consumer movements, union organizing, populist movements of many kinds, are all fertile soil in which elaborated codes (better than that of the mayor, I hope) may grow along with the habit of democracy.