126 : CHAPTER SEVEN
We can go further. The presumed truth of an artistic representa tion of a social fact is an essential element in our appreciation of th" work as art. That is, art and truth do not work at cross-purposes go-that you can have one or the other but not both. In a lot of works, you can only have both, or neither: no art without truth. The truth of the work's assertions about social reality contribute to its aesthetic effect6' That's why the class got so angry at Tom. If the story about his aunt arid-father was true, it moved and upset us. If not, it was just a silly j0fce No truth, no art.
8
The Morality of Representation
Representing society raises moral questions for participants, for makers and users. These come in several varieties: misrepresentation as a moral wrong; the way common techniques shape our moral judgments; the related questions of assigning praise and blame for the results of action and of casting participants in social action as heroes and villains.
"Misrepresentation"
Sociologists in my tradition routinely seek understanding of social organizations by looking for trouble, for situations in which people complain that things aren't as they should be. You can easily discover the rules and understandings governing social relations when you hear people complain about their violation. Fields of representational activity undergo periodic violent, heavily moralistic debates over the making and use of their characteristic products. The cries of "It's not fair" and "He cheated" would sound like the games of five-year-olds were the stakes not so much higher and the matters dealt with so much more serious. The problem of misrepresentation invites us to begin our analysis by looking for these conflicts.
Anthropology students at the University of Papua New Guinea complained, in the Nova program "Papua New Guinea: Anthropology on Trial" (Nova 1983), that Margaret Mead's Growing Up in New Guinea was "unfair" because it repeated the derogatory stories her informants had told about the students' ancestors, for whom the informants'
j3o :
C HAPT EH EIGHT
THE MORALITY OF REPRESENTATION : 131
people had a traditional contempt. The students didn't complain that Mead had reported what had been said inaccurately; they agreed that those people had said such things. Nor did they complain that Mead had presented the stories as fact; she hadn't. No, they complained because their own ancestors, whom Mead had not studied, used to say equally terrible things about those other people, and Mead had not given them equal time.
These complaints exemplify the class of complaints that arise from self-interest: "You made me [or mine] look bad!" The first assistant physician of the mental hospital Erving Goffman studied and wrote about in Asylums complained (in the footnote Goffman donated to him) that for every "bad thing" the book described he could have produced a balancing "good thing": for the victimizations of patients Goffman reported, he might have told about the newly painted cafeteria (Goffman 1961, 234). Similarly, the citizens and politicians of Kansas City, Missouri, complained that the i960 U.S. Census under-reported the city's population by a few thousand, thus keeping it from sharing in the benefits state law gave to cities over half a million (a law that had been designed to help St. Louis out of financial trouble some years earlier). Almost everyone whose organization Frederick Wise man has filmed complains that they didn't realize they were going to end up looking like that.
The practice of more or less fictionalizing reportage, as practiced by Norman Mailer, Truman Capote, and Tom Wolfe, among others, provoked a more general complaint. The well-known journalist John Hersey (1980) pointed out that these writers not only made things up but insisted on the right to make them up in the name of a higher truth. He argued that an author can invent details and incidents in writing labeled as fiction, which carries on its license the legend "THIS WAS MADE UP!" but not in journalism. There "the writer must not invent. The legend on the license must read: NONE OF THIS WAS MADE UP. The ethics of journalism, if we can allow such a boon, must be based on the simple truth that every journalist knows the difference between the distortion that comes from subtracting observed data and the distortion that comes from adding invented data."
Hersey adds, interestingly, that distortion by omission is accept-
able, because "the reader assumes the subtraction [of observed data] as a given of journalism and instinctively hunts for the bias; the moment the reader suspects additions, the earth begins to skid underfoot, for the idea that there is no way of knowing what is real and what js not real is terrifying. Even more terrifying is the notion that lies are truths" (Hersey 1980). But many critics have complained that print and broadcast journalism (e.g., Molotch and Lester 1974; Tuchman 1978; Gitlin 1980) leave out exactly those things people need to be able to assess issues properly. And it's easy to imagine that many readers would "instinctively hunt out" additions in just the way Hersey goes after subtractions, if they knew that they should; many of Wolfe's readers, as well as newspaper readers and television viewers, probably did just that.
Hersey, whether or not we accept his judgments, identifies the sociological core of conflicts over representations of social reality. No report in any medium or genre, following no-matter-how-strict rules, will solve all problems, answer all questions, or avoid all potential troubles. As we've seen, people who create reports of any kind come to agree on what is "plenty good enough," what procedures should be followed to achieve that good-enough condition, and that any report made by following those procedures is authoritative enough for ordinary purposes. That protects professional interests and lets the work of the people who use those procedures proceed, guaranteeing the results as acceptable, believable, and ready to bear the weight that routine use for other people's purposes puts on them. The agreed-on standards define what is expected, so that users can discount for the shortcomings of representations made according to them and at least know what they are dealing with. Hersey's analysis accepts this state of affairs as normal, standard, and proper. It is what 3 had in mind earlier when I said that everyway of making a representation is "perfect," good enough that users will accept the result as the best they can get under the circumstances and learn how to work with its limitations. Critics claim that misrepresentation has occurred when someone doesn't follow the standard procedures and misleads users into thinking a contract is in force when it is actually not being honored.
Quarrels among documentary film makers often revolve around
132 : CHAPTER EIGHT
THE MORALITY OF REPRESENTATION : 133
methods whose difference from a previous standard seem to create the possibility of confusion about what the film alleges to be true Michelle Citron provoked a storm of criticism by including "fictional" passages in Daughter Rite (Citron 1979), an otherwise "factual" film Some more conservative filmmakers complained that viewers would be misled, "tricked" into thinking that what they were seeing had actually happened when it hadn't. Citron, not unreasonably, argued that her film displayed a more generic "truth."
Users and critics also claim "misrepresentation" when the routine use of acceptable standard procedures harms their interests by leaving something out that, were it included, would change not only the interpretations of fact but, more importantly, the moral judgments people make on the basis of the representation. That often happens when some historical shift makes new voices audible. The people Mead studied did not read anthropological monographs and so could not criticize them. But their descendants, studying at the University" of Papua New Guinea, can and do.
In either case, the problem of misrepresentation is a problem of social organization, a problem that manifests itself when a bargain once good enough for everyone is redefined as inadequate. Many "moral" problems that crosscut genres and media can be similarly analyzed as organizational products, including the ethics of representation and the problem of the authority of a representation.
"Insidious": The Moral Community of Makers and Users
Frederick Wiseman's film Tificutt Follies (1967) describes, in an unin-flected, nonjudgmental way, the day-to-day life of the Bridgewater (Massachusetts) Hospital for the Criminally Insane. No description will do justice to this complex work, but here's a short version. Mostly in very long takes without a cut, it portrays scenes in the life of the institution which, you come to believe, recur repeatedly for staff and inmates: meetings in which staff discuss patients and decide on their treatment; hospital personnel force-feeding recalcitrant patients via an intranasal tube; a patient shouting gibberish for minutes on end
^thout stopping; a holiday show featuring members of the staff and inmates; Vladimir, apatient, explaining to an apparently unreachable staff why he should be let out. It's easy to see how such a place would drive a person crazy, but it's also easy for most people to see that a lot 0{ the people in there were probably already very crazy when they arrived. The film leads almost anyone, however, to conclude that the institution is a terrible place that ought to be closed and that the staff are cruel and unfeeling. Unlike most documentaries of its era, Titicutt Follies has no titles or voiceover commentary telling viewers what to think. Nevertheless, just as in Haacke's "Guggenheim" piece, the selection and editing of the film leads any reasonable viewer to conclude that this hospital is a terrible place.
A student in the "Telling" seminar objected that Wiseman's film, which 1 had presented to the students as a wonderful piece of documentary work, was "insidious," meaning (she said, when I asked her to explain what she meant) that it used all sorts of film devices ("tricks") to get viewers to believe that what they saw was "true": the lighting, the harsh, unceasing noise, the men's frequent nakedness (not commented o n by anyone in the film), the very long takes, which led viewers to think that this material was not simply a collage of cleverly edited short moments that might hide a larger and different reality. She wasn't clear about why that was "insidious," but I thought then and still think that it was a wonderful word.
Why? "Insidious" implies that an effect was achieved by means that you, the viewer, weren't fully aware of and therefore can't be critical about. When a voiceover in a film tells us something, we know a voice is speaking to us in intelligible sentences, and many, if not most, of us have learned that, most of the time, we should suspect authoritative voices. But we may not understand in the same way that when a camera points up at someone to film them from below, they will look bigger and more awe-inspiring or scary and, conversely, that someone filmed from above, by a camera pointing down at them, will look smaller, less authoritative, and more childlike. When we know what's being done we're on our guard, we look for reasons not to accept the idea urged on us, we recognize the tricks and are wary. When we don't know what's being done, when it's insidious, we aren't on our guard,
134 : CHAPTER EIGHT
THE MORALITY OF REPRESENTATION '. 135
don't take proper intellectual precautions, and are likely to be "fooled" or "lulled" into accepting a statement or idea we wouldn't accept if We had had all our antennae up to detect trickery.
People vary in what tricks will fool them. These insidious fore may be less likely to affect professionals than amateurs or the general public. We can guess, not unreasonably, that people who make films -for a living know what's up and take care not to be fooled. Some presentational tricks are so well known that they aren't "tricky," whi may be what Hersey had in mind in distinguishing the common journalistic practice of not putting everything in a story, which he thought was OK because "everyone" knows newspapers do that, from inventing dialogue that never took place, which ordinary readers might not - -be accustomed to watch out for.
The distinction is important. Inaccuracy or corner cutting or other : "illegitimate" practices presumably do not fool users who know that makers routinely use such practices. These alert users discount for "~: the distortions introduced by such expectable routine activity and are skeptical about conclusions based on material produced in ways that contain these routine "errors" or "distortions." But people who don't
know about routine distortions or omissions may accept conclusions---------
and ideas they would never accept if they only knew the routine tricks that were leading them that way.
If these naive users knew how the trick was done, they would know that these "invalid" methods do not produce "real evidence" that would withstand crucial tests. And then they would know that the conclusion was "no good," having been "justified improperly." I put all those words in quotation marks to indicate that the informed readers I invented in the last paragraph might take this view, not that I accept all those criteria and all that reasoning myself.
Which suggests a generalization. For every means of telling about society, there will be some group for whom that way is justified by a moral pact between makers and users, which specifies permissible ways of persuading users that what's alleged is valid and therefore can be publicly acknowledged as acceptable, and which identifies sneaky and unacceptable ways. People who use sneaky means will be seen by parties to this pact as cheating, violating the moral agreement that
makers and users have entered into. Users who are parties to this moral pact will be knowledgeable, within the limits set by the agreement, and so will not be easily fooled; they will expect makers to abide by the agreement and avoid means of persuasion not already agreed t0. ("Insidious" implies what might not be true, that users would object, if they knew, to being persuaded by means they're hardly or not at all aware of.)
We needn't imagine that this pact has been agreed to in some self-conscious, document-signing way, or even the way agreements are invoked when you buy computer software (by opening the envelope containing the disk, you accept all the terms of some contract). We can just suppose that people agree to accept it the way so much is agreed to and accepted in ordinary social activity, by continuing to participate in the activity, even as one becomes aware of all these tacit understandings. (Keep in mind the standard etbnomethodological warning: participants often honor agreements after the fact by figuring out, on every occasion, what they might or must have had in mind when they said they agreed to whatever it is.)
Other users, not parties to such a pact, may not know what to look out for and could therefore be easily fooled by the unscrupulous. But we might say of users like this that they have no business using stuff they don't know enough about, that its not the maker's fault if they insist on fooling with what they don't understand and can't properly evaluate.
We can say all that—if we insist on taking sides in such a potential dispute, which we needn't. I'd rather avoid taking sides on such issues and instead just observe who disagrees with whom about what—treat it as a sociological phenomenon to study rather than as a court case for us to decide.
For every form of telling about society, we should look for (as a possibility, not an inevitability) a moral community of makers and users, whose members know and accept some standard methods of communicating ideas and conclusions about society and of persuading others of the validity of what's communicated, even though those methods are riddled with faults and flaws. Users know all about what makers do. No "insidious" persuasion is going on. Makers are not do-
136 : CHAPTER EIGHT
ing secret things to fool users; there are no secrets. (The highly professionalized and esoteric world of mathematical models, discussed in chapter 9, is like that. The only people who consume these models regularly are people who themselves could make them, and maybe do.)
We can ask all the standard sociological questions about these representational communities. How do they recruit members and socialize them into the way they do business? Which participants know all about the persuasive means makers use? Where did they learn that? Which users know less and are more likely to be taken in by insidious means? What selection process divided users into knowledgeable and not so knowledgeable? Did those who don't know have a chance to learn but not take it (as I can imagine many people reading this book might decline a free ten-week class in mathematical modeling)?
In many of these worlds a small group of makers produces representations viewed by a large group of not-very-knowledgeable users. Most people who see movies in theaters or on the small screen wouldn't know how to make one. Not knowing how to make a film, of course, is not the same as not knowing how to watch films critically. On the other hand, the statistical tables and charts presented in newspapers and popular magazines may well fool people not trained to-spot trickery. They may know that statistics lie but not know what kinds of lies are told and how to detect them, something only experts know. (Which produces books like Damned Lies and Statistics [Best 2001], which aim to set them straight.)
Are all the less knowledgeable participants in these worlds being duped? Many people might not care much that "insidious" means of communication had fooled them. Suppose you told viewers that the makers of Titicutt Follies had manipulated their emotions and conclusions by the cutting and pacing of shots, so that they had come to believe what might or might not be true. Many of them might say (might not, of course) that they didn't care, that they believed the evidence of their senses, what they had seen and heard, independent of such influences; that no such influences could chajige their judgment of the doctor's failure to take Vladimir's logic as seriously as we end up taking it; that no instruction on the artful use of montage, camera angle, lighting, or sound recording could change their conclusion
THE MORALITY OP REPRESENTATION : 137
! that the treatment people receive in hospitals like this one eventually kills them; that an awareness of sequencing and editorial decisions oin't take away from the inhumanity of the lockup procedures and the way the guards tease inmates they saw in the film.
So "insidious" implies what might not be true: that users would object, if they knew, to being persuaded by means they're not aware of. That points to another level of moral agreement involved in maker-user relations. I'm going to speculate about possibilities here, not report research results.
Some users might well be mainly interested in the "big" conclusions of the work, for which there is plenty of what seems to be straightforward evidence, for which the insidious means are only "incidental," as the mood-establishing background music of a documentary film might be. These users could say that all that incidental stuff just helps them grasp the message; they aren't fooled by it, they welcome it as a reader might welcome an easily read typeface. Readers might welcome a graphic device they're minimally aware of which emphasizes some element of a table more than it "deserves" (by using a device that professional statisticians think is misleading) because it helps them see what's important to them. Critics might say that that just shows how fooled they really are.
Who gets to decide that someone else doesn't know enough to make judgments on serious questions for themselves? We routinely assume that is true of children below a certain age, probably without thinking much about why we're entitled to assume it. Can we similarly assume that we know better than adults, who know less than we do about the matter at hand?
Questions about who can and should protect less knowledgeable users leads us to consider varieties of social organization surrounding the making and using of representations and the learning of the morality that surrounds these activities. One way to learn what that variety is would be to ask about different methods and organizations of socialization into the making and use of representations.
We learn about some representations as part of growing up: how to watch movies or read books, for instance. Others require specialized training: learning to read a complex statistical table or a technical
138 : CHAPTER EIGHT
THE MORALITY OF REPRESENTATION : 139
map. Many representations come in a variety of forms, some readable by any ordinarily well-socialized member of a society, others only by experts and the specially trained. The difficulty is not an intrinsic property of any representation, it depends on what people have been trained to do. If eveiyone in a community learns, as a matter of course, to read complex weather charts (as might happen in a seafaring community or at an airbase), that's ordinary socialization, although in other places only the highly trained know how to do the same thing. This varies historically too. What was esoteric in another generation is grade school stuff now. Conversely, fewer people now have skills ~~ how to make a dress from a pattern you buy at the store—that were once more common.
Hersey argued that we needn't worry about users' being fooled by the journalistic practice of not telling you everything you need to know, because readers know how to protect themselves against that form of trickery. They will, he thought, do the work of protecting" themselves, reading carefully, thinking about possible other materials the journalist might be leaving out, assessing what those materials might contain, and deciding how what might be in those materials could alter their judgment on the issue being discussed.
That's a heavy responsibility for an ordinary reader, and it returns us to the question of the division of labor. Do people actually do that work? Average newspaper or magazine readers probably aren't that careful or skeptical about what they read (something for a researcher to nail down). They might be more like the students McGill interviewed, who didn't think they had to read the tables in scientific articles because the editors have already ensured that the tables say what the text says they say and support the author's argument.
Praise and Blame: Who and What Are Good and Bad
Social science and historical analysts almost always, explicitly or more or less covertly, make strong moral judgments about the subjects they write about. Historians don't just argue about whether the Civil War was inevitable; they want to establish, say, that it wasn't in-
evitable and so the people responsible for it happening are guilty of causing it. Had they behaved differently, it never would have happened, and all those lives would have been saved. Or they want to establish that the war was inevitable, given the configuration of forces and events at the time, and so those same people are not guilty.
In the late twentieth century, sociologists and anthropologists, and others as well, argued about whether poor black people in the United States—who no one denied were not as well off as other people, in many different ways—contributed in some way to their own troubles (just as people argued about whether European Jews had done something that made them complicit in their own deaths in the Nazi camps). Scholars and others were arguing about "the culture of poverty" or, in another form, "black culture": Do poor (or black, or poor and black) people more or less willingly participate in a system of understandings and practices that makes their victimization by a system of exploitation, repression, and oppression inevitable? Or could they, by somehow not participating, improve their situation?
Even though social scientists may seem to argue about specific factual findings and specific technical problems, you can almost always find behind the arguments a desire to show that something is just the way it should be or not at all the way it should be, although the "should" is generally left unargued and unsupported. When the student in our Performing Social Science class read the article on race and educational expenditure "with feeling," he made that animus evident.
The Rhetorical Value of Being Neutral
Most ways of telling what we know about society try to appear neutral and avoid looking like they're just raving and ranting in a way that would convince only the already convinced. They present facts and let users arrive at conclusions.
Some makers keep their moral beliefs to themselves. They run the tables, present the materials germane to a problem they have made a serious moral judgment about—racial discrimination, for instance— and then let readers come to their own conclusion: a standard scientific stance often recommended by authorities (e.g., Ogburn 1947).
140 : CHAPTER EIGHT
Makers who do this count on all or most of their users sharing their moral position. Most U.S. social scientists (sociologists, certainly^ maybe this is less true in other fields) are more or less political liberals as Americans use that expression, more or less left as the rest of the world understands these things. They can, as a result (or so [hey think), take certain premises for granted. If I demonstrate an income disparity between blacks and whites, I don't have to say it is a bad thing. Almost everyone who reads what I write will agree it's bad. The moral conclusion follows automatically from the statistical result (which, nevertheless, does not logically lead there).
Such disguised judgments appear in other ways of telling about society. It's not only social scientists who assume an ostensibly neutral stance. Haacke's "Guggenheim Project" and many of his other works use the same strategy, presenting more or less well known facts, arranged to lead users to a moral conclusion he expects them to work out for themselves. Wiseman s films have a surface air of simply presenting what you could see if you'd been where he's been.
In twenty-first-ceiitury Western societies (and many other places as well), being scientific means being neutral. When you pursue almost any public goal, the strongest ally you can enroll in your cam-paign is science, precisely because everyone thinks it is neutral and therefore not influenced by what we would like to be true, but only by the results of impartial, objective research. Opponents who disagree with your religious beliefs and question your moral imperatives have a tough time arguing with science, which, everyone thinks, just tells it like it is. Which is to some large extent true, despite all the critiques and social constructionist arguments. I accept most such arguments, but I still trust a neutral scientific study more than an argument based on religious revelation or deduction from a moral imperative that I don't accept (or even from one I do accept).
By presenting my findings and analyses neutrally and objectively, then, I can communicate my moral judgments effectively. As long as users share my moral premises, simple logic will bring them to my moral conclusions.
You get a great rhetorical advantage from this roundabout way of doing things: you can present your moral judgments as the findings of disinterested science. But the moral judgments, hidden though they
[HE MORALITY OF REPRESENTATION : 141
may be, can cause analytic trouble. It's a problem of language. Scientists try to use neutral terms that collect things that resemble each other enough that you can find verifiable generalizations about the conditions that lead to them. They want their language to be precise and don't try to make it include a moral judgment. Medical scientists don't usually treat germs and viruses as requiring moral condemnation. They want to know how those organisms work, and what their reproductive cycle depends on, in order to interfere with it effectively. Of course, they think the germs and viruses are "bad" for us and need to killed off and gotten rid of. But they don't spend time condemning them and calling them bad names.
Why not? Because everyone agrees that tuberculosis and syphilis and measles are bad. The diseases and the germs that cause them have no defenders. (Although George Bernard Shaw made a pretty good case for them in his 1932 play Too True to Be Good, in which a germ is an important and sympathetic character.) As a result, scientists can describe them in technical language and no one will accuse them of moral irresponsibility. But if they discuss the causes of lung cancer and whether the manufacturers of cigarettes are responsible for cancers that develop as the result of a life of smoking tobacco, neutral language has a moral consequence. (And recall the discussions of "good" cholesterol and "bad" cholesterol.)
The language that makers use to write about social life is always engaged in a game of expressing moral judgments, trying to avoid them, or making them in a disguised way. There are serious reasons to avoid name-calling in social analysis, which I'll treat in chapter 13, in the discussion of ErvingGoffman s carefully neutral analytic terminology. Some representations come very close to a nonjudgmental, quite neutral recitation of plain uninterpreted fact. James Agee did it in Let Us Now Praise Famous Men ([1941] 1988), and novelist Georges Perec made experiments in this vein that Til discuss in chapter 15.
Causes and Blame
Social scientists routinely look for the causes of the phenomena they study; it's the most common way of describing what we do. Moral judgments frequently take the form of assigning blame. Social scien-
143 t CHAPTER EIGHT
tists routinely assign blame by announcing what caused something to happen. If we know what causes something, we know what has to be changed in order to change some social consequence we disapprove of. If we don't like the situation of black people in the United States and want to change it, and if we know what causes that situation, then we know what to change in order to get the result we want. If we can identify X as the cause, we know we should do something about X so that it no longer produces the result we don't want. When you explain what caused something bad, you blame that cause for the untoward: result you've analyzed.
That's a misleading, and in the end mischievous, way of thinking; The justification for that harsh statement comes from an alternative way of thinking about how social phenomena happen. (For longer discussions of this complicated question, see Ragin 1987, 2000, and Becker 1998, especially 63-66 and 183-94.) Looking for causes misleads because it supposes an additive model of how things happen. It's mischievous because it leads analysts to assign blame in an incomplete and morally questionable way.
Suppose that the deplorable situation of black people in the United States is caused by a number of things: flat-out racial prejudice, institutional racism, industry's leaving the cities where black people live, the prevalence of the crack habit and trade in the neighborhoods many blacks live in, and so on. We could reasonably and correctly add many other things to the list, but the completeness of the list doesn't affect the point I want to make.
In conventional causal analysis, each cause influences the thing we are interested in. In conventional analytic language, the causal (independent) variables affect the effect (dependent) variables to some measurable degree. So racial prejudice adds (I'll invent the numbers) 10 percent to the bad situation, industrial flight from inner city neighborhoods adds 30 percent, and so on, until all the variance in the situation we want to explain is accounted for by a combination of these variables. Any of the independent variables could do the whole job of producing the unwanted result by itself, if it were strong enough, but none of them ever are. And any combination of them would do the trick if they were collectively strong enough. The causal variables are
THE MORALITY OF REPRESENTATION : 143
substitutable. To say it another way: you can add the causes' influence up, and any result that brings the total to the right number will produce the effect.
The alternative, a multiplicative analysis, looks for the "conjuncture" of variables that produces the result. What combination of variables has to be present for the effect we're interested in to happen? This approach says that each of those things is important. If any of them are missing, the effect won't happen, or won't happen in the way We want to explain, though something else unpleasant might occur. That's why it's called multiplicative. Remember grade school arithmetic. If you multiply some number, any number no matter how big it is, by o, the result will be zero. Analogously, if any of the conditions necessary to the final result is missing, you won't get the final result. With respect to the situation of blacks in U.S. cities, Mario Small's 2004 study of a community in the Boston area makes this clear.
Good Guys and Bad Guys
The problem arises in nonscience forms of telling about society as well. Storytellers almost always, implicitly or explicitly, take sides. Stories have heroes and villains, and the storyteller usually lets us know who is who, either by explicit labeling or by providing easily read clues. In stories for grownups, we expect some subtlety. The villain doesn't always wear a black hat and have a long, droopy mustache, but by the end of the story we know who to cheer for.
Most people who do sociological research, or even just read sociology for pleasure or profit (that is, for some pragmatic purpose), think of it as one of the "social sciences," and the word science is often, though not always, taken quite seriously. As the previous chapter suggested, users imagine that what they read is not merely the expression of someone's opinion, consisting of and shaped by wishful thinking and pious hopes, but that it is in some way dependent on what is "actually happening" somewhere in the "real world." They prefer to think that what the report they are reading tells them rests on materials systematically gathered and analyzed, the "results" justified by something beyond the writer's genius or intuition.
144 : CHAPTER EIGHT
Users want to know all this because what they "really" want to know is who's to blame for this mess, who can be held to account f0r these injustices, whose fault it is. They want to sort the actors in a j cial situation—the participants in an organization, the opponents in— a political dispute, the parties to a quarrel—into Good Guys and Bad Guys, those doing the right thing and those acting badly. This rests on a simplified notion of cause: bad results are caused by bad people act-ing badly.
But you can't get from the results of a scientific study to moral judg- " ments like that in any direct way. You can sometimes show that these actions have these consequences (it's not easy, but suppose you can). But you can't derive the judgment that some people are good and have behaved well and others are bad and have behaved badly directly from the results of empirical investigation. You can decide, on the basis of a philosophical argument, that certain kinds of acts or consequences are bad and then show "scientifically" or empirically that those people have done those acts and that their acts have had those consequences. :
Many people find this troubling. They want to strengthen their moral position by demonstrating that what they disapprove of is bad scientifically. My own experience as one of the participants in the de- — velopment of the "labeling theory" of deviance (Becker 1973) gives an example. Labeling theory analyzed "deviance" as the result of complicated, many-stage interactions involving accusers, accused, and a variety of official and unofficial organizations. Such an approach generally cast doubt on conventional assignments of praise and blame, on the allocation of actors to the Good Guys or the Bad Guys, by showing that the process of accusation and proof of guilt was a social process, not a scientific procedure. Critics, appalled by such relativism, often asked something like this: "Well, what about murder? Isn't that really deviant?" They implied that while many acts might exhibit the definitional variation that was the key insight of the approach, some acts are so heinous that no reasonable person would ever define them in a way that excused the person or persons or organization that had committed them. It never helped, when this accusation was made, to point out that whether something was murder, as opposed to justified homicide or self-defense or acting on behalf of your country or sup-
THE MORALITY Of REPRESENTATION I 145
porting law and order, was exactly a definitional matter. This criticism, by the way, was made from both the left and the right, the right upholding "traditional values" and presenting such crimes as murder and incest as the killer counter-examples, while the left offered up crimes like "imperialism" and "colonialism" to accomplish the same end. (See Becker 1973,173-212.)
What was at stake was this: most people interested in problems of society want to say something more than identifying what they don't like as deviant according to community standards as applied in that community. They want to say that those community standards of badness were not just the community's standards but were standards science had shown, scientifically, to be bad. Critics didn't want the word deviant to have a simple technical definition as "something some participants in a situation call bad"; they wanted it to mean "bad, and science has proved it's bad."
"What about murder?" challenged me to deny what was obvious to any reasonable well-socialized member of our society: that something we all know is bad, like murder or incest, really is bad. When I said that I agreed with them and did think that murder is bad and was willing to say so, they were not happy; my agreement that it was bad didn't satisfy them. So I asked: Why isn't it enough to say that murder is bad and evil? What do we gain if say it's "deviant" too? What is gained is obvious: the authority of science. Because a judgment of "evil" can be justified only by theological argument and a judgment of "bad" by ethical argument. And even those who are firm in their own belief know that they cannot convince nonbelievers with arguments like that. They want an argument that works with nonbelievers too. That argument is science, which any well-socialized member of contemporary society presumably believes in.
Maybe that example is enough to show that users of social science reports want a way to distinguish good and bad, good and evil, good guys and bad guys. And the people who make social science reports are, for the most part, not only willing but eager to supply that distinction. It doesn't take an acting student, reading a scientific report with great emotion, to show that either right on the surface of a social science research report, or barely beneath it, the makers are assigning
146 : CHAPTER EIGHT
THE MORALITY OF REPRESENTATION : 147
moral praise and blame, even when they profess "objectivity" and "SCj, entific neutrality." Historians do this openly and as a matter of course critics may blame them for not doing it. They assign blame, as I su£ gested, for wars. If Lincoln had done this or that, maybe the South-" erners would not have been so angered as to want to secede. They assess the moral character of historical actors. If Thomas Jefferson was' really the father of his slave Sally Hemings's children, and given the' undisputed fact of his slaveholding, does he deserve the respect we accord him as a founder of the nation?
Many social scientists will not recognize themselves or their work in the preceding paragraphs. William Fielding Ogburn, who brought. statistics to American sociology and sociology to the American government, thought that, since sociology was a science, sociologists should be neutral in a way that permeated their writing. He wanted objective and nonemotive prose that replaced evocative words with precise words with clear meanings (Ogburn 1947).
Most social scientists routinely follow Ogburn's advice, whether they know it or not. What they write still contains villains or heroes, usually disguised as the attribution of causality to variables. Take an excellent example of the genre, one in which the labeling of virtue and" vice is not hidden. Stanley Lieberson wrote A Piece of the Pie (1980) to answer this question: How come American blacks have not achieved the kind of individual and communal social mobility other ethnic groups managed? Why could the Jews and the Italians and the Irish and the Poles do it and the blacks couldn't? Is it because of discrimination, or does this failure reflect inherent differences in ability? Whose fault is the lack of black mobility and social success? Blacks', for not being good enough? Or whites', for not letting blacks have a fair chance? This factual question, to be sure, can be answered factually, with enough careful defining of terms and critical inspection of all the available sources of information. But it's simultaneously a moral question because, given the way Americans think about blame, if it's discrimination it's the white folks' fault; if it's not discrimination, if it's something about black folks, if it's "their fault," well, too bad, maybe we can do something about it, but it's not our fault.
Not to keep anyone who hasn't read Lieberson's excellent book in
SL!Spense, the answer, arrived at after inventive and exhaustive analy-sjS of a mass of imaginatively discovered data, is that the culprit responsible for the low mobility scores of blacks is indeed discrimination, no doubt about it. Though Lieberson's prose is as scientifically chaste as Ogburn could have wished, the moral animus of his argument is perfectly clear. The chaste prose, by the way, has an important rhetorical consequence: it helps convince readers who might not have fully made up their minds on these questions that the author who produced these results has no ax to grind. If the data had shown that it wasn't discrimination, he would have told them that just as forth-rightly, so they better believe this; there's nothing left to support any other conclusion.
Christopher Jencks routinely writes this way, taking seriously propositions that outrage most routinely liberal American academics and subjecting them to a rigorous examination. His prose is so antiseptic and his analysis so evenhanded that when he concludes, after a careful and systematic assessment of the available evidence, that Arthur Jensen's notions about the low intelligence scores of black Americans are hogwash (Jencks 1980), you believe him in a way you would not, quite, if he had begun with some conventional pieties about how reprehensible Jensen and his ideas are.
Most social science reports bury their judgments deeper than that. Perhaps it's better to say that they routinize them, so that the moral judgment is present simply in the choice of a problem. Why study the distribution of people of different races in the ranks of a large organization, if you don't think that there is some injustice going on? But once you've chosen the problem, no more moral talk, or not much. Your readers will supply it automatically.
Most users of scientific representations are content to let the moral go unsaid. Users of works in artistic genres seem more often to require that moral condemnation be expressed explicitly. I'll discuss the interesting case of Wallace Shawn's play Aunt Dan and Lemon, which provokingly refuses to make such judgments, in chapter 12.