What About Mozart? What About Murder?

Howard S. Becker

Three fields which might seem on the surface to have very little to do with each other--the sociologies of deviance, art, and science --have a deep underlying similarity, which surfaces as a chronic fight over definitions. Investigating those fights and that similarity tells us something interesting. I'll begin with two examples from my own experience as a participant in such discussions.[1]

When I published Outsiders in 1963 I was living in San Francisco, quite near the University of California at Berkeley which had, among its many appendages, the Center for the Study of Law and Society, then directed by the deservedly well-known scholar Philip Selznick. Selznick harbored in his center a cluster of students of "deviance," as the study of what had been , variously, called "social disorganization" or criminology was then coming to be called--Jerome Skolnick, David Matza, Sheldon Messinger, and others--although he himself had grave doubts about the whole thing. In the natural course of events, I was invited to talk at the Center about the "labeling" theory of deviance, of which Outsiders served as a prime example.

I remember the occasion vividly. I laid out the ideas contained in the first chapter of the book--that deviance was not an innate or "natural" quality of someone's act but rather resulted from the joint activity of such an actor and of the people who responded to his or her activity by labeling it as "deviant." ("Deviant" was a general term some of us had begun to use to encompass all kinds of negative categories that arose in specific areas: "criminal," "crazy," "abnormal," "perverted," "unethical," etc.) This ran counter to the common idea that these categories were things that people actually were, essential aspects of their being, and thus easily defined. A criminal committed crimes, an abnormal person did things normal people didn't do.

When I finished, there was the usual time for questions and discussion. And I remember Phil Selznick, standing in a doorway at the back of the room, smoking a cigar, looking at me quizzically, and asking, mildly, "Well, Howie, I see what you're getting at, it's very interesting," and then coming in for the kill: "But, after all, what about murder? Isn't that really deviant?" He settled back, convinced that he had made a devastating point. I didn't think so, and countered with familiar counterarguments: that reasonable people differed over which acts of killing were murder and which weren't, that these differences varied depending on what kinds of people were involved, the historical era, and so on. He didn't think I had answered his question. I thought I had.

Before analyzing this example, I'll add another. Some years later, when I was teaching at Northwestern University, the then Dean of the College of Arts and Sciences, Rudolph Weingartner, inaugurated a series of "Dean's Lectures" and asked me to give one of them on the subject of my new book, Art Worlds. Although I hadn't put it this way in the book, Art Worlds presented what you could reasonably call a "labeling" theory of art, as one of the components of its general approach to art was that it was a collective activity, something people did together. The labeling component had to do with the problem of defining art as an object of study, a problem which had perplexed aestheticians for millenia and showed no signs of solution. I had solved it somewhat cavalierly by refusing to attempt a definition, being interested instead in studying those occasions when people did define things as art and argued over the definitions.

As I presented these ideas in my lecture, the Dean increasingly became visibly upset. He was ready with the first question: "Well, Howie, that's all very interesting but, after all, what about Mozart?" I was ready for this question with an appropriate though unresponsive answer: "What about Mozart?", I countered. The Dean looked surprised (he had thought the meaning of his question was clear enough to a reasonable person) and said, "Well, isn't Mozart really a musical genius?" And I made the obvious answer: if you accepted all the premises of a particular approach to art, then Mozart certainly was a genius. But people often rejected those premises and, if they did, might not agree with that conclusion. The Dean, like Selznick, felt that he had made a devastating point I had not answered acceptably; I thought my answer was just fine.

I can't present an example from the sociology of science based in my own research experience, but the examples are can be found everywhere in that field. The killer question in that field takes two slightly different form. The positive version is "What about airplanes?" If you reply, in the form I used with the Dean, "What about airplanes?", the answer is, "Well, don't you believe in the science that says they will fly? If you don't, why do you ever get in one?" All the the people involved in such arguments are contemporary scholars who constantly fly here and there, so it's a serious challenge: why they do that if their own research and thinking imply that science is "merely" a matter of consensus.

The negative form of the killer question in the sociology of science is "What about astrology?" If you say "What about it?" the follow-up asks if, thinking that science is just consensus, you think astrology was true back then when all learned men believed in it? Sociologists of science respond that, being reasonable contemporary people, they believe that airplanes will fly and do not bother to read their horoscopes in the daily paper. But, they add, since scientists have often collectively believed things later generations of scientists no longer believe, they also think it likely that scientists in years to come will treat many things contemporary science takes as obviously true the way we now think of astrology. And once again, the critics think they have made a devastating point, but the sociologists don't think so.

In all three cases, a field of empirical research (a science) has replaced a field of philosophical discourse. (In what follows, I have no intention of, to use Heinich's phrase, "dethroning" philosophy, only of noting what has been happening to fields of philosophical work.) What was once a realm of argumentation based on canonical examples and logical reasoning has become an enterprise in which scientists arrive at conclusions based on systematic investigation of the real world, taking much less for granted than philosophical discussions usually take. This process has happened often in the history of philosophy. Physics and other natural sciences were once philosophical topics, but their discussions of physical and biological phenomena have long since given way to the findings and theories of hard-nosed empirical scientists. Beginning in the late 19th century, psychology changed from an introspective philosophical activity to an empirical practice carried on in a laboratory.

Many classical topics in psychology remain for philosophers to talk about, at least for the moment--the nature of consciousness, for example--but many others have been superseded by empirical investigation. And, more importantly, many of the older questions have been reformulated, sometimes quite drastically, in a new language, getting their meaning from a new paradigm, as Kuhn taught the world to see. So the question of "art" changes from a search for unalterable principles by which "art" can be recognized to an investigation of the way the term is deployed in the ongoing activities of an art world. Ethics turns into the sociological investigation of how moral judgments are made and enforced. This does not mean that the old questions of morality and ethics and knowledge go away. They don't, and aestheticians and epistemologists and ethicists continue to debate them. But much of their territory has been colonized by social scientists, covering the same ground from a different standpoint.

Since I have now appealed to empirical investigation as the touchstone with which we distinguish what to believe from what to reject, it's important to take a small side-trip and make clear my position on the vexed problem of relativism, which underlies all three of these examples. Are there "real" things out there to which we can appeal to settle our disgreements? Or is it just a matter of who can shout the loudest and find the most allies? Can we find out what is true, by appealing to that reality? That's obviously what Selznick and the Dean and the scientists in whose mouths I put questions about airplanes and astrology were doing. Here's what I think, and I believe this is what most sociologists and, in fact, most working empirical scientists think too, when they are not engaging in polemics aimed at securing turf and the like.

Something is "real"--and therefore has to be taken seriously as the touchstone which tells us whether what we think is true should be accepted as such or not--when what I say about it will withstand all the criticisms and questions people can bring up to discredit it. That's always been my idea of how a sociologist should work. You anticipate what serious critics--people who really don't want your conclusion about whatever-it-is to be true, people who have a stake in showing you are wrong in any way they can do it--would say. Then you do whatever is necessary to counter those criticisms, so that the critics can no longer make that criticism, because you have answered it so well they have to accept it. This is not quite the same as shouting louder or having greater political skills, but it does refer to the agreement between you and the critics that they won't make that complaint any more because, by their own standards, it doesn't hold water.

This position has two antecedents or predecessors in the social science literature on method (I’m sure there are others, but these are the ones that have influenced me). One is Donald Campbell's idea that working scientists establish the truth of they say by dealing with "challenges to the validity" of their hypotheses. Which I translate as "countering the things people could say that would stand in the way of accepting those hypotheses." He brought this up in the context of experimental and quasi-experimental design, outlining (with his collaborators Stanley and Cook) all the possible objections to various such designs in social psychology, and the ways you could counter them. He accompanied these analyses with the important proviso that you could certainly not, in research on human beings, ever deal with all possible criticisms, and so you had to choose the ones that were likely to occur and not worry about the rest.. This analysis can be fruitfully applied to all the kinds of work social scientists do--field research, historical studies, survey research, and so on.

The second antecedent of my position is found in Bruno Latour's discussion of how scientists validate their findings to their colleagues. They do two things that are relevant here. They subject the prospective finding to tests of various kinds, "trials of strength," which it must withstand. Other scientists accept a finding that has withstood all the tests that can be brought to bear on it, has taken the worst the opposition can offer and emerged victorious.

Scientists who want their colleagues to accept their results as real do a second thing. They control the potential critic's moves by cutting off promising avenues of criticism with convincing answers before the critic even starts down that path. (Latour calls this "captation," adopting a rhetorical term for the purpose.)

In other words, I accept as "true," not just someone's opinion, a statement that no one has criticized in an unanswerable way, a finding that has withstood all the trials of strength it has been subjected to by friends and foes. This differs from Popperian practice, where scientists are supposed to always try to disprove something. Unlike Popper, I'm not proposing a criterion for what truth "really" is, simply saying what in practice has to happen so that other people who work in your field will accept what you say as true, or true enough.

Which adds another dimension to the analysis. Because what I propose involves a different way of seeing the question of truth. For me, that question is not as an eternal problem of epistemology, but rather one that arises for people in a multi-person, usually multi-organizational, setting, where what is true is a practical matter for working scientists to deal with. I don't ask whether they deal with it "properly," in some ultimate sense. I want to know if they deal with it well enough to shut critics up. The analysis consists of looking at a situation and asking who is criticizing who and with what results.

So I'll say that something is true when no one asks me a doubting critical question I can't answer to their satisfaction, forcing them to shake their heads and say, in effect, "Well, I hate to admit it, but I guess you're right." If all or most of members of a scientific community make that admission, it's good enough for me to treat the proposed result as true. There's a tricky aspect to this, because of course I wouldn't treat all questions from all people as worthy of that kind of serious answer. So: whose questions are accepted in the community to which the proposed idea is addressed? For me, now, it's the community of working social scientists, and maybe not all of them because, after all, this is not chemistry, where you might imagine that all the members of the community share standards of proof that aren't argued about. Remember that Kuhn's field observations showed him that social scientists spend a lot more time arguing about these epistemological difficulties than natural scientists do. (I've profited in this discussion from the observations of Ian Hacking.)

Similar answers to the charge of relativism can be made with respect to questions of art and problems of ethics. It's important to note that agreement on results in these fields is not as necessary to the continuation of work in them as is true of science. Ethicists and aestheticians can work on quite happily even though they disagree on fundamental principles. They have disagreed for two thousand years at least and will probably continue to do so, but the fields do not suffer from this.

We can say that art objects exist independent of any observer. But the sense in which that is true is not very interesting for a social scientist and the statement immediately has to be qualified. In fact, art works change and are changed constantly and the isolation of "the work" to talk about is difficult and can only be accomplished by agreement among discussants about some convention which says when the work is the work that we will talk about. Once that kind of agreement is reached, however provisionally and temporarily, discussion can go on perfectly well about "the work" thus isolated (of course, the changes in the work that do occur will create anomalies that will produce problems for the discussants).

The assessment of artistic value goes on continuously in the communities and networks of working artists. Since artists never agree on all these questions, the agreement on value really works best in smaller sub-communities, although even there such disagreements as whether Baroque music should be played on contemporary instruments or on instruments from the period when the music was written produce unresolvable differences. Disagreements on value do not, however, prevent working artists from agreeing to cooperate on more practical matters, such as the creation of venues for musical performance, or what works will be accepted, for practical purposes, as constituting an agreed on set of Acceptable Works, good enough to be exhibited or performed. But this possibility is often not realized, and sub-communities, which accept different standards exist side by side. Members of these communities usually have no trouble experiencing works in the same way and arriving at similar judgments.

So I'll say that, as long as I accept the premises and share the aesthetic experiences, I can agree that a work of art really is wonderful and that Mozart certainly is a genius.

Similarly with ethical questions and problems. Rooted as these are in forms of collective life, as a lot of sociological research (especially on so-called "social problems") has shown, they invarably lead to conflict and the inability to definitely resolve questions when the premises of the way of life are not shared among all the discussants. Once such premises are shared, it is possible for social science research to investigate the allegations of fact, of causal relationships and patterns of influence, that shape the application of the ethical standards in question. So I'll say that, insofar as I share the form of collective life in which the standards are rooted, I will inevitably share the ethical judgments that follow from them.

The comparison of empirical investigation and philosophical discussion--to return to the main road--makes clear that a fundamental aspect of the latter is definitional, and definitional with a twist. The enterprises I have characterized as philosophical want to find the rules which ought to govern definitions of value. The terms being defined are honorifics: "Art" is good. Non-art isn't. "Real science" is good. "Bad" or "pseudo" science isn't. "Law abiding" is good, "criminal" isn't. The application of these terms has real effects: whether you go to jail or go home, whether people believe your findings or make fun of them, whether what you've done is art or trash. When people argue over these definitions, more than logical precision is at stake.

The philosophical investigation of ethics can be seen as (to be deliberately provocative) a primitive forerunner of the sociological study of deviance. The way people should conduct themselves and judge their own conduct and that of others has increasingly been reformulated by sociologists as the study of how people think others (and themselves) should behave and, importantly, as the study of the organizations set up to create and enforce these judgments, and the consequences of such organizational activity. Philosophers still write about ethical problems and search for defensible arguments for one or another ethical system. But much of this terrain is now taken up with sociologists discussing problems which are related, but formulated differently and meant to be answered in a different way, the answers judged for their empirical adequacy. Not "how one should behave," but "who thinks what about how one should behave and what they do about it."

Similarly, aesthetics has always been a field of philosophical inquiry, though early versions had an empirical character. But the great questions of aesthetics have mostly been "What is art ?" and "What is great art?", and the converse, what is not art at all. It's been a characteristically negative enterprise, designed to prevent unworthy stuff--which has its own (pejorative) names like "kitsch" or "mass culture"--from being mistaken for the real thing. The sociology of art, in the versions I prefer, avoids that question and instead looks at how the term "art" is used in the organizational life of art worlds. Who assigns this title, how those assignments are maintained and acted on, with what results--these are the standard questions in such an empirical inquiry.

Epistemology tells us what we should count as "real knowledge" and what as phony and not worthy of respect. The sociology of science doesn't tell us what "real" knowledge is, but rather what kinds of organized activities produce the results scientists prize as scientific. A good example is Latour's investigation of the way French soil scientists made the leap, in their study of the savanna-jungle margin in Brazil, from observational facts like an unmarked patchof forest to abstract ideas like the succession from one ecological type to another . This classic epistemological problem is difficult to solve if you put it that way--how do you get from A, way over here, to B, unimaginably far away over there? Latour shows that working scientists do it by taking very small steps--from a marked off piece of land, to a sample dug out of the ground, to a box full of such samples, to a chart based on this box, and finally to a journal article. Each step makes sense to the community the results are presented to, and the epistemological mystery is solved.

I don't want, in making these summary characterizations, to be unfair to the field of philosophy. Philosophy is a more varied enterprise than most academic disciplines and no summary statements of the kind I just been making will do justice to that variety. While many philosophers engage primarily in normative analyses of the kind I have criticized, plenty of others do what most social scientists would recognize as perfectly respectable social science analyses. I was taught, for instance, to see Aristotle's Poetics as an empirical treatise on the characteristics of those tragedies which had specific effects on their viewers, a kind of social psychology or maybe even, as we might say these days, a study in the reception of dramatic works. But in such philosophical subfields as ethics, aesthetics, and epistemology, the main thrust is not empirical, but analytic and normative.

Even when philosophers do discuss social organization, they do it in a hypothetical way, inventing examples to illustrate the categories they have developed in their analyses. So philosophers like Arthur Danto and George Dickie, when they trace judgments of aesthetic worth to the operations of an entity they call the "art world," do not discuss any particular art world in its full organizational reality. Instead, they use things that have happened or might happen in some field of art to bring out the distinctions they are making. So Dickie devoted himself to questions like this: suppose the keeper of an elephant in the zoo proposed his elephant as a candidate for a work of art. Does that make the elephant art? Well, it's a question, but not one that in any way engages the kinds of questions that get discussed in the contemporary worlds of, say, painting or sculpture or literature or theater.

When the operations of the art world began to appear in the discussions of aestheticians, it was not the art world observed by sociologist Raymonde Moulin, in which participants' judgments of aesthetic worth are inextricably mixed up with judgments of financial worth, not because the people involved are venal but because the organization of the world of contemporary painting makes that unavoidable. Nor was it the world of bargains and deals and very uncertain markets analyzed by the economist Richard Caves. Or the world of Big Hollywood Robert Faulkner described, in which a few top players in the major occupational categories make, between them, half the films that Hollywood distributes every year.

Aestheticians pursue the logic of how such worlds arrive at the collective judgments they eventually arrive at, but not the messy organizational reality of curators and dealers and collectors and critics, and their multiple and conflicting motives that social scientists describe. The art world is, for them, a logical device which helps to explain how a system of aesthetic judgments might work.

The same is true--here comes another sweeping generalization with a lot of exceptions--of epistemology, which looks for the criteria which distinguish real science from bullshit. Latour, approaching epistemological questions, doesn't look for the logical justification for a result, but rather at the way that question is transformed in the work scientists do into a series of small, uncontroversial steps, which get to the result in a way that is acceptable In the community of scientists it is addressed to. There's no ultimate logic of truth in such an analysis, only an understanding of the logic in use by people who do the science we all accept.

Often enough, these two ways of doing things--aesthetics and sociology of art, ethics and deviance studies, epistemology and sociology of science--just exist side by side, each minding its own business, speaking to its own disciplinary audience. But sometimes they conflict with one another and sometimes they create conflict within one of the disciplinary homes.

Inter-group conflict. This is what happened with me and the Dean about Mozart. He had been an aesthetician and could see that I was committing a heresy by suggesting that art and genius, and all the related concepts so central to an aesthetician's approach, were social conventions. (In a favorite ploy of philosophical discourse, he might have accused me of saying that they were "mere" social conventions.) He wanted me to admit that what people like him were interested in--art and genius and all that--were real, not just the outcome of some agreement between the relevant parties. Agreements can be changed and that would mean that those sacred things weren't, in some important sense, "real." To a sociologist nothing is more real than what people have agreed on, which we make central as "definitions of the situation."

Why should an aesthetician care what sociologists think? I think he wanted an acknowledgement of the reality of those concepts. He wanted me, and by extension social scientists in general, to admit that some features of art objects are not relative, not matters of opinion and consensus, but are rather inherent qualities of those objects, and that these inherent features have been verified as such by Social Science.

If social scientists agreed to that, then their research agenda would have to change. We would, if we did so agree, devote ourselves to answering such questions as "What are the conditions under which great art is created?", taking the adjective "great" as it generally or commonly applied, as a given of the research. We would not be so interested in the fluctuating reputations of works, one of the bases for social science relativism vis-à-vis the arts. That Shakespeare was once thought less of than he is now would be of no real interest and could be written off to the blindness, prejudice, and ignorance of earlier generations.

In such cases, what the dean and people like him appeal to is the common sense of informed art lovers and critics, what my father, a devoted believer in these ideas, called "the wisdom of the ages." My insistence that reputations fluctuate, that the features of a work arise from an interaction between an ever-changing object and a variety of constantly changing appreciators, seems like a willful ignoring of all that wisdom and knowledge, perverse and provocative. The same thing occurs in quarrels over the meaning of morally charged concepts like crime and deviance, and of the ideas of science. In the study of deviance, the professional groups who own the territory--police, lawyers, politicians, psychiatrists and other physicians--create the common sense understandings a sociological investigation makes the object of study. When my colleague asked me "What about murder?", he was voicing the common sense understanding that murder really is different and requires a different explanation than "less serious" actions that might be definitionally less clear. He wanted me to recognize this difference and agree that such a characeterization was scientifically founded.

With respect to science, scientists (many of them academics themselves) can't understand why fellow investigators of reality refuse to recognize the superior quality of the knowledge they produce, and are particularly infuriated by what they (incorrectly) take to be an implication of sociology of science: that science is just a matter of agreement among people, as though they could agree to anything and that would make anything science. They want us, sociologists, to agree that our own science, sociology, ratifies the claims made by the other sciences.

Intra-conflict. I have talked as though these attitudes and approaches were simply distributed along disciplinary lines but, of course, they aren't. The arts and humanistic disciplines are filled with people who are every bit as relativistic as the most dedicated social constructionist could ask for. This is very often for the good reason that they are personally acquainted with the ups and downs of reputation--it's what they study--and know how easily reputations change and on what flimsy bases these changes rest. Their own perusal of the evidence has led them to a sociological conclusion. You could say that they are sociologists without knowing it. I find inspiration and useful methods and results in the work of such people as Barbara Hernnstein Smith, Michael Baxandall, Scott Deveaux, Paul Berliner, and others from a variety of fields.

On the other hand, some sociologists accept that inherent features of events, objects, and activities exist, which survive all variations of social context and can only be interpreted and understood as possessing these unchangeable features. They agree that some art works are works of genius, that some science is the real thing while other activities are phony science, and that some activities are really deviant.

I don't remember everything I said to Philip Selznick on the occasion I described earlier, but this is what I would say now and have said on similar occasions since. I did remind him that people don't agree about what acts constitute murder, that murder under one set of circumstances is justifiable homicide, under other circumstances not,; that in many times and places murder or something it would be hard to tell from it has been in fact the only available way of settling disputes; and so on. That didn't satisfy him and it won't satisfy anyone who believes that murder is inherently deviant.

Nor would such weaseling satisfy the people who, on later occasions, asked why I was not willing to say that capitalism or patriarchy or homophobia were "really deviant," as many people were ready to say they were.

So I would, finally, have said that, since I shared many of those opinions, I was perfectly prepared to say that those things were evil, disgusting, or almost any other invidious word they wanted to use. And I would have asked why that wasn't sufficient. Because it would not have been. No more would it have been sufficient for me to say to the Dean that, being a musician myself, I felt the way he did about Mozart. And felt the same way about Dizzy Gillespie and Stan Getz and a lot of other jazz players (because I was and am a jazz player), opinions I doubt he would have shared. Just as my agreement (because, after all, I am a working scientist too) with all those who think astrology is not science would not have satisfied critics, because they would not have accepted my addendum that that was so as long as others shared my local definition of what constituted science.

Why wouldn't such answers have been acceptable? What I'm about to say has a very general bearing on many social science problems. These answers would not work because what these people ask for is not just agreement with their judgments, which I'm often enough glad to give, but a declaration that this is not a matter of personal opinion but a scientific finding, a result that carries the warrant of certified knowledge. Not that the Dean and I agree that Mozart is terrific, but that there is objective, irrefutable evidence that he is. Not that we agree that murder is terrible, but that its terribleness is a scientific result, not a shared emotion.

Why would anyone insist on that? Because science, and the certified knowledge it is thought to produce, is the only basis on which anyone can win an argument any more. If I say that my opinion on any of these matters is correct because, let's say, the religious truth that you shall not murder has been revealed to me in the sacred writings, many of my readers who do not accept religious revelation as a source of knowledge will not accept my conclusion. And I have nothing more convincing to say to them. If I say that my instinct tells me that Stan Getz is a better saxophone player than Kenny G ever will be, that only convinces people who already agree with me. These and similar subtleties do not provide the certainty and persuasive power that only science can now provide.

Why won't I accept these judgments, which I actually accept in my life as a citizen, piano player, and working social scientist? Because, if I did, I would be committing myself to a research program doomed to failure. An important step in research is establishing classes of phenomena about which to generalize. (You don't necessarily have to do this at the beginning, but you have to do it sometime.) If the members of the class aren't the same in ways that are relevant to what you want to generalize about, you won't find any interesting generalizations. This is what you do when you create a class defined by how other people (judges, police, psychiatrists) have reacted. It's what you do when you take conviction for a crime as being something about the inner nature of the person convicted, instead of something about the process of conviction. The people convicted might have some other things in common besides being convicted, but that isn't guaranteed. Or, rather, it is only guaranteed if the process infallibly singles out people who do have something else in common, such as having committed a certain act, which is nothing anyone can count on.

When Donald Cressey, wanting to study embezzlers, thought he could find them by taking a sample of people convicted of that crime, he soon learned that you couldn't collect people who had all done the same thing that way. Because prosecutors don't charge people with what they have actually done but with a crime the prosecutor can convict them of, and the definitions of financial crimes often made it impossible to prosecute them for embezzlement, the crime Cressey wanted to study. He had to devise other definitions and then find people among the available prisoners who fit the definition of his subject that he developed: "the criminal violation of financial trust."

This is the classic problem of good social science research. We can't create homogenous classes of activity for which we might be able to find reasonable causal processes if we rely on the conventional definitions available in the worlds we study. Those definitions are made, as Garfinkel long ago argued, for other purposes than social science and reflect a host of organizationally engendered compromises and expedients that can only impede our efforts to make some social science.

But, when we ignore "common sense," "conventional wisdom," or "the wisdom of the ages," we will surely run into opposition from the people who take those definitions as self-evident. That's our dilemma and there is no easy solution.

[1] Some readers may notice that this essay touches on some of the same subjects as Natalie Heinich's book Ce que l'art fait à la sociologie (1998:Paris, Minuit). I think she arrives at a somewhat similar position following a more philosophical analytic path.