Minor debts, derisive remarks, a fight over a parking space, butting in line—these are the little things that nevertheless account for much of the violence in human society. But why? Roger V. Gould considers this intriguing question in Collision of Wills. He argues that human conflict is more likely to occur in symmetrical relationships—among friends or social equals—than in hierarchical ones, wherein the difference of social rank between the two individuals is already established.
This, he maintains, is because violence most often occurs when someone wants to achieve superiority or dominance over someone else, even if there is no substantive reason for doing so. In making the case for this original idea, Gould explores a diverse range of examples, including murders, blood feuds, vendettas, revolutions, and the everyday disagreements that compel people to act violently. The result is an intelligent and provocative work that restores the study of conflict to the center of social inquiry.
|Publisher:||University of Chicago Press|
|Product dimensions:||5.50(w) x 8.50(h) x 0.90(d)|
About the Author
Roger V. Gould was a professor of sociology at Yale University and visiting scholar at the Russell Sage Foundation when he died in 2002. He wrote Insurgent Identities and edited The Rational Choice Controversy in Historical Sociology, both published by the University of Chicago Press.
Read an Excerpt
Collision of Wills: How Ambiguity About Social Rank Breeds Conflict
By Roger V. Gould
University of Chicago PressCopyright © 2003 Roger V. Gould
All right reserved.
CHAPTER 1 - Conflict, Honor, and Hierarchy
One occasionally hears the phrase "senseless violence." People doubtless mean a range of things when they use these words, but the phrase almost always suggests the existence of another kind of violence: the reasonable kind, the kind that makes sense. Absolute pacifists aside -- and there seem to be very few people who would seriously claim that it is never reasonable to cause someone physical harm -- it would appear that some acts of violence are understandable, perhaps even just, whereas others are condemned as out of bounds, devoid of sense.
To be sure, people often hold different opinions regarding where to draw the line between reasonable and senseless violence. At the beginning of the twenty-first century, for instance, there are disagreements about when or whether parents should strike disobedient children, about what constitutes unnecessary use of force by police, and about when efforts at diplomacy ought to give way to armed intervention. But there is rarely disagreement about what the dimension is on which the line is drawn: in general, the higher the stakes are, the more willing people are to accept that physical violence might be a reasonableresponse. It seems to make more sense when urban residents riot after a police killing than when they riot after their team wins a major game. It is easier to understand why someone might kill to protect or avenge a family member than it is to understand why someone might kill to keep a mugger from taking the contents of a wallet. It is easier to see why someone might put up a fight when the wallet contains two hundred dollars than when it contains five dollars. And it is easier to explain why two friends might come to blows, or worse, over a dented fender or a romantic rivalry than over a disparaging remark.
At any rate, it seems easier in light of conventional ways of thinking about hostility, anger, and violence. The more intrinsically significant the event -- that is, the more someone stands to lose or gain -- the more reasonable we tend to think it is if he or she becomes angry, or violent, or deadly. This is not to say that our response is of the form "I too would react that way"; it is simply that we are less likely to feel mystified by the behavior or the emotion it expresses when the stakes seem high. In modern settings, employing physical violence is an extreme measure. It is therefore more readily understood, or at least regarded as understood, when the situation in which violence occurs is itself extreme.
Straightforward though this conventional way of thinking is, it has a hard time with the real world of violence. Specialists in homicide in the United States and other industrialized societies have known for decades that a very large number of killings are "motivated" by apparently insignificant concerns: arguments, verbal insults, small debts, parking spaces, accidental physical contact, and so on. Indeed, one of the first studies to look closely at the details of how and why people kill one another (in Philadelphia, as it happened) set aside a category for "altercations of relatively trivial origin." That rubric or a version of it continues to appear in statistical summaries of homicides all over the United States -- and consistently accounts for more than one-third of homicides in American cities. To be more precise, more than one-third of homicides cannot even in principle be attributed to conflict about a serious matter. Because the rest can be connected to something serious -- such as marital infidelity, property damage, or loss of a job -- they have been so classified. But that attribution is often itself guided by the conventional wisdom about violence: if one person kills another because of a gambling debt and an insult hurled during an argument about the debt, we assume that the murder occurred because of money. By default, the most significant issue, which is to say the issue that has the greatest impact on the disputants' interests, is presumed to be the primary motive for violent behavior. The apparent contribution of "serious" matters to violence, and conversely the amount of violence attributable to serious matters, is thereby maximized by the typical classification procedure.
This interpretive maneuver still leaves an enormous amount of mayhem that by most people's estimation occurs because of trivial matters -- "silly stuff," as more than one esteemed social scientist has said in my presence. There appear to be three characteristic responses to this difficulty. The first response amounts to theoretical neglect, although it is typically described in nobler terms. When available theories fail to account for the full range of human behavior, we often attribute the disparity to something like free will -- although it is more fashionable for social scientists to say "agency." Never mind that philosophers have struggled for centuries to agree on what the concept of free will refers to, not to mention whether there is such a thing. It is doubly comforting to attribute the mismatch of theory and behavior to a special property of human action: we simultaneously excuse the inadequacy of our models and glorify our status as human beings. There is no need to dwell further on this response -- it may turn out to be the best available, but it seems to me that the quest for understanding ends there. In any event, it is rarely invoked in the context of violence, perhaps because it is unpleasant to think that free will manifests itself in wickedness. People like to invoke free will to explain nice surprises, not nasty ones.
The second and more common response is, in essence, to classify as different from ordinary people those who do not conform to established notions about when it is understandable to harm others physically. Non-experts impose such a classification when they refer to anomalous users of violence as "irrational," "crazy," or "evil." Psychologists and others in the therapeutic professions do so when they define the people who engage in anomalous violence as mentally ill. Social scientists -- especially sociologists -- adopt a similar approach when they construct models of deviance that attribute departures from normative behavior patterns to environmental influences that teach or induce certain people to behave differently from everyone else. A person who shoots a prowler does not require special classification as crazy, ill, or deviant; but a person who shoots someone because the ice cream cone is the wrong flavor does. So violence that does not fit the conventional view is prevented from undermining that view by means of a classification of its perpetrators as unconventional themselves.
Sociologists have made a good deal of progress in identifying the mechanisms that sort people into conventional ways of life, on one hand, and into lives characterized by deviant practices, on the other. In fact, it is one of the unfortunate achievements of the discipline to have made itself unpopular because it appears so often to excuse antisocial behavior. By claiming to locate the sources of such behavior in the social environment rather than in the individual actor, sociological theories of deviance have challenged the self-righteous and punitive stance of law-and-order advocates who believe that people who behave badly are bad people. (Sociologists would likely have had this effect even without the inclusion of the radical perspective, in which the very categories "crime," "deviance," and "antisocial behavior" are denounced as ideological instruments ruling groups use to oppress others.) A large body of evidence suggests that in many people, deviance, including habitual violence toward others, is a response to a life of mistreatment, or thwarted opportunity, or at a minimum neglect and lack of socialization. There is also evidence that people identified as deviant by schools, courts, mental health professionals, and ordinary folk are likely to become more so as a result of this labeling.
Yet we miss an important issue if we approach the matter of violence the way criminologists approach delinquency -- that is, as a matter of distinguishing the violent people from the others and then explaining why they differ. Although it is surely the case that some people have a higher propensity than others to inflict serious physical harm, it is also the case that most of those who do harm others do not choose their targets randomly. Not counting warfare, the vast majority of the violence we find out about occurs between people who know each other. It is nearly certain that an even larger proportion of the violence we do not find out about, and of the conflicts that do not end violently, involves people who know each other -- principally because violence and conflict between close associates and relatives is much less likely to be reported than violence between acquaintances or strangers. Even people who by any standard have a special penchant for violence do not distribute it evenly: they nearly always direct their rage at specific people or specific categories of people. The same holds more firmly of those whose bouts with physical violence are rare: physical attacks by one-time offenders are particularly likely to concentrate on close associates rather than on bystanders or strangers.
These facts imply that there is something to be gained by thinking about interpersonal violence as a product of social relations, and of the conflict they engender, rather than as an expression of individual personality types. For one thing, it raises the question of whether there are certain kinds of social relations in which conflict is likely to occur, independent of the propensities of the people linked by such relations. It may be that one sort of person might respond to conflict by causing physical harm while another might withdraw or become self-destructive, but there is nonetheless a sense in which harm, when it is the response, might be attributable as much to the relation as to the individual. If it should turn out, for instance, that a disproportionate number of domestic homicides in the United States involve men killing their brothers, we might conclude that there is something about the male sibling relationship that recurrently produces more strife than the brother-sister relationship, the father-son relationship, and so forth. Recalling that many brothers are also fathers, sons, husbands, uncles, and brothers-in-law, we might have to acknowledge that it is as brothers that the offenders in such cases committed their acts of violence -- in other words, that it was in connection with a specific social relation, and not as a personal disposition, that they were capable of extreme behavior. The idea of deviance conceived as an individual propensity hides this unevenness in the way people interact with others.
It also hides the possibility that violent interactions, occurring as they do within certain kinds of social relations, are emblematic of the less deadly but more frequent battles of will that occur in tens of thousands of comparable relations every day. If it is true that violence is distributed in systematic ways across types of social relations, then it can offer us lessons about the distribution of conflict in general -- lessons we would disregard if we decided that violent behavior is just something violent people engage in. If only one employee in a thousand responds to being fired or disciplined by returning with a gun, that does not mean that the other 999 employees are not also deeply hurt by the experience; it simply means that their responses attract less attention. Seeing the violent response as an index of how painful employer-employee relations can be, rather than as an indicator of the mental instability of a single person, allows us to learn much more about social conflict in general. Throughout this book I shall view violence -- along with data on violence -- as a window onto the broader phenomenon of human conflict, not as a lurid expression of deviant personalities.
The third characteristic response to the fact of violence due to apparently trifling matters is to identify a new kind of interest: something called "honor." Someone angered by an insult, rather than by some materially damaging action such as an injury, theft, or legal defeat, can be said to feel dishonored -- to feel as if his or her reputation or personal pride has been harmed, even if there is no concrete harm. If honor is important, then defending it by means of violence makes as much sense as defending other important assets. This approach amounts to expanding the range of issues considered significant. It is similar to economists' response to the discovery that people do more of something (such as gardening) than one would expect on the basis of the opportunity costs and the cost of having someone else do it. The anomaly is explained by positing the existence of previously unrecognized rewards of the activity, such as relaxation or a sense of accomplishment. In the context of violence, the corresponding answer to the fact of violence due to apparently trivial matters is to recognize that some people care not only about protecting their material interests but also about preserving honor. Urban youth frequently account for lethal attacks on others by saying that the victim "disrespected" them. To recognize honor as a valued good is to take such statements seriously as explanations.
As generally understood, then, honor is a resource every bit as important to those who have it as the somewhat more concrete resources of wealth, physical strength, formal and informal authority, social contacts, technical knowledge, and so forth. The key difference is that honor is entirely symbolic: its possession consists in the recognition by others of its possession. To have honor, to be honored, is to be thought honored by others. It makes no sense to say that someone possesses honor even though nobody realizes it, whereas it might make sense to say that someone has power or wealth without others' knowledge. Material resources such as ships, shoes, and horses can be useful to their possessors in a way that is at least partly independent of cultural representations of these resources. Honor, on the other hand, like status, sacredness, legitimacy, and other things predicated on recognition by other people, just is a cultural representation. Its value is more profoundly a product of context than is the value of a loaf of bread.
It is therefore no surprise that the study of honor, and of acts of violence driven by considerations of honor, has for decades been seen as a matter of cultural interpretation. To speak of matters of honor as involving symbolic resources is to invoke the notion of meaning systems, of gestures and utterances given significance by their location in a web of other gestures and utterances. Beginning in the 1960s, anthropologists especially -- but also historians and the occasional sociologist -- have sought to explain honorific violence in terms of the cultural setting in which it occurs. Scholars writing in this tradition insist that killing someone because of a verbal slight or disrespectful sneer, as alien as it might seem to an outsider, makes sense once one becomes more familiar with the way the people who commit such acts understand social life. Moreover, they point out that a gesture that counts as a slight in one time and place could be a compliment, an expression of gratitude, or a meaningless motion in another context. As a result, explaining acts of violence that cannot be attributed to such obvious concerns as those of property or self-defense depends on interpretation, which in turn demands detailed knowledge of cultural context.
The commitment to culturally sensitive explanation this view demands has produced a remarkably rich body of ethnographic and historical research on "honor societies," from highland New Guinea to Tokugawa Japan on the Pacific rim, from Appalachia to the Andes in the Americas, from Scotland to the Balkans in Europe, and from the Nile valley to the Mediterranean in Africa. We now have a staggering amount of information about the range of ways people think about and display respect, status, shame, integrity, modesty, courage, loyalty, deference, scorn, admiration, and any number of other traits and emotions associated with honor. We know, for instance, that squaring off face-to-face against one's adversary was crucial to being thought honorable at the peak of aristocratic dueling in America and Europe but that ambushing one's enemy -- or someone else, for that matter -- can be a perfectly honorable way to take revenge among headhunters in the Philippines, feuding families in Sicily and Corsica, and hunter-gatherers in the Amazon. We also know that in Islamic societies a family dishonored by the nonmarital sexual activities of a daughter frequently respond by killing the daughter, whereas in Christian societies it is more common for the aggrieved family to murder the offending man. Ethnographers have found that kin of Nuer murder victims may hold any male member of the killer's lineage accountable, whereas Berber customs extend responsibility only to the ten closest agnates. There is evidence that hospitality has been a central ingredient in the maintenance of honor in Crete, Andalusia, Montenegro, and the American South but was not especially important to the honor of Japan's samurai class or to that of medieval knights in continental Europe.
Anthropologists and historians have also documented wide variation in the kinds of gesture or remark that may give offense. In many honor cultures in the West, one person may gravely insult another with the epithet "coward" or by insinuating something about the sexual habits of the latter's mother. An elite white man in the antebellum southern United States might have insulted a rival, inviting a challenge to a duel, by publicly calling him a liar or by pulling his nose. But for the Tauade in the mountains of Papua New Guinea, it is more common to insult someone with a comment about his (or her) age or white hair or with an invitation to drink the urine of the speaker's spouse. Bedouins can wound each other's pride -- to the point of provoking lawsuits -- by saying things like "you came here barefoot and on a donkey." If I said that to a neighbor, it would surely provoke more puzzlement than outrage.
In short, recognition that people can arouse anger in others by entirely symbolic means, means that, moreover, vary enormously across cultural settings, has pushed many social scientists to focus on the reasons for, or at any rate the occurrence of, such variation. The challenge is to explain why a man's nose is symbolic of honor in the American South and in Kabylia but in many other places is not, why hospitality is essential to honor in Andalusia but not in Crete, or why female chastity is a key ingredient of family honor in Mediterranean societies but typically has no such implications among indigenous peoples of Oceania. Or rather, since most cultural analyses are only implicitly comparative, the project for each scholar has been to explain why the set of symbolic resources is what it is in his or her chosen setting. Few students of cultural systems expect to construct a general theory of how particular symbols become linked to or detached from honor across a range of times and places. In fact, the very richness of the observations made of honor and its symbolic contents has led inevitably to a sense that the general concept is useless. With so many variants of honor, many writers have become suspicious of the notion that there is any reason to keep thinking of them as instances of the same thing. The goal, then, is to represent in as meaningful a way as possible the experience of those whom the researcher is studying, allowing for metaphorical connections and perhaps an acknowledgment of family resemblances across contexts but not for reduction to a common structure. The interpretive tradition, in a sense, does for groups of people what the notion of free will does for individuals: it exempts each culturally distinct community from the expectation that it will obey strong regularities by endowing its members with the collective power to live in a world of their own (collective) making.
It is indeed easy to be enthralled by the spectacular range of ways human groups build and understand their social world, and more particularly by the creativity with which they produce and react to challenges, insults, and insinuations, on one hand, and gestures of submission, deference, and admiration, on the other. But in concentrating on variability, one runs the risk of becoming blind to just how much similarity there is across time and place in the kinds of contests that lead some people to inflict grievous harm on others. Just as the "deviance" approach to violence obscures the similarities between harm less verbal exchanges and those that eventuate in injury or worse, emphasizing the variation in forms of honor obscures the very tangible constants that persist amid such variation. Even as we recognize that moustaches or facial scars may or may not carry significance as indicators of honor, depending on the time and place, we would do well to acknowledge the overwhelming consistency with which certain features are absent from, and others are present in, honor codes.
There are, for instance, no examples (none that I have encountered, at any rate) of social groups that express greater admiration for people who choose to ambush their enemies rather than confront them face-to-face, even though the former is often considered the more prudent approach. Although there may be examples of societies in which it is not insulting to spit in someone's face (though, again, I know of none), there are hundreds in which it is. It also appears that the tendency to see vertical placement -- high versus low, tall versus short, standing versus kneeling -- as a metaphorical representation of status or honor differences is nearly universal.
As observers and interpreters, human beings are typically contrast-seekers: we tend to be better at noticing differences than we are at recognizing uniformity. (If you are an urbanite, think of the last time you tuned out the sound of a car alarm, only to notice it again when it shut off; if you live in a rural setting, think of crickets.) When we look at a range of social contexts, consequently, we often place the ways they seem different in the foreground and relegate their common features to the background -- either defining them as uninteresting or forgetting them entirely. Yet it would be a grave mistake not to take seriously the frequency with which certain patterns recur in the context of rivalrous or conflictual social interaction -- even as one acknowledges that human groups are also full of surprises. In the first place, the value of broad comparisons is enhanced if one is searching for general patterns (not to say "laws"). If the purpose of examining multiple contexts is primarily to look for differences, the aggregate outcome of thousands of studies may turn out to be little more than a catalogue of human variation -- a kind of naturalist's guide to the flora and fauna of human lifeways.
If the purpose is to identify regularities, on the other hand, then comparison helps us distinguish what is incidental to a pattern from what is basic. For example, with regard to the matter of honor, the full range of historical and ethnographic research teaches us that, even though the particular manifestations vary -- likening someone to a certain species of animal, remarking on physical flaws, calling someone a liar or a coward -- verbal insult of some kind is a universal feature of interpersonal conflict. The fact is so obvious that our attention is drawn to the variations in content: Why a pig here and a dog there? Why the chin here, the nose there, and posture somewhere else? But if one were to stand back and look at the whole range, one might decide that it is equally important to understand why it is everywhere the case that people can make each other angry enough to kill just by saying things.
In the second place, and perhaps counterintuitively, the value of detailed, individual case studies is also enhanced if they are seen as tools for generalization rather than items for a catalogue. The way the academic world currently works, authors of case studies (communitylevel ethnographies, local histories, analyses of a single industry or profession) face a no-win situation. A new study may report the same things earlier studies have observed and thus confirm existing generalizations. In this case, the value of the new study is merely incremental and diminishing with the number of prior studies reporting the same pattern. Or the new study may diverge in some major or minor respect from previously documented patterns, thereby offering a novel finding. But since the divergence is ordinarily seen as refuting or weakening extant generalizations, the impact is in an important sense the same: each new study serves only to reinforce the idea that things differ, that the world is complex, and that generalizations are nearly always wrong. Eventually, there will be no generalizations left to attack. This kind of particularism is self-defeating: once we accept it, we have no good reason to examine a new context or revisit an old one unless we have some specific reason for being interested in that particular time and place.
What I am proposing is that additional cases be seen as constraints on generalization, not as adversaries to it. The more we learn about more times and places, the more difficult it becomes for existing general propositions to go unfalsified. One reaction to this fact is to lose faith in the project of generalization, as many social scientists have. But another reaction is to ascribe these recurrent failures to the propositions themselves -- not to the practice of formulating and testing them. It is quite possible that the principal flaw of general statements is not that they are general but on the contrary that they are not general enough. To the degree that general theories are extrapolations from what is already known, they are likely to be tainted, as it were, by features particular to the settings on which they are based. When a theory fails to fit new data, the mismatches point to precisely those features.
To shift ground for a moment, take the following example from economic anthropology. Once upon a time, many social scientists believed that markets, surplus production, and division of labor went together: societies in which people produced goods for exchange were thought to be characterized by high levels of occupational specialization and a cash economy, whereas societies in which people produced goods for themselves or for expropriation by social superiors were characterized by low levels of specialization. A few economic anthropologists suggested that the whole range of systems could be accounted for by a general theory presupposing rational, utility-maximizing actors. A variety of findings such as the discovery of large-scale societies with high levels of differentiation but no markets (most of them defunct, like the Maya) contributed to a sharp debate about whether a single model -- in this case, formal neoclassical economics -- could explain economic behavior across time and place. The issue, of course, was not (or ought not to have been) the semantic one of whether the concept "economy" could be used to refer simultaneously to market capitalism, nonmarket exchange systems, and subsistence production, or whether it was properly restricted to the former. The real issue was the theoretical one of whether the same principles could be adduced to explain behavioral patterns in all of these types of economy, despite their superficial differences. "Formalists" argued that, even if explicit monetary exchange did not occur, it was still possible to infer the operation of such principles as cost and profit in the decisions made by producers -- and thus to use these principles to explain their behavior. "Substantivists" insisted that inhabitants of different societies produced their means of livelihood according to fundamentally different principles. They therefore derided the effort to construct general models of economic behavior as misguided universalism or -- worse yet -- as cultural imperialism.
Both sides had a point. The market model exaggerated the degree to which production decisions were determined jointly by price and the desire to maximize returns. In many peasant societies, for instance, cultivators produce "enough" for their families, not "enough" to produce a surplus for reinvestment and expanded production. In many horticultural societies, cultivators do produce a surplus -- of yams or shellfish, for instance, or of sumptuary goods -- but it is donated to village leaders who redistribute it to gain status and attract political support. Neither pattern looks like the kind of entrepreneurial investment in productive growth that occurs in Western capitalist societies.
But some substantivist critics of economists went too far in the other direction. The fact that data from a new study fail to confirm a specific prediction stemming from a specific economic theory is no reason to conclude that all economic theories must be wrong. Making such an inference is like concluding that it is impossible to build a flying machine because the first flying machines didn't fly. (Many intelligent people, however, concluded just that.) The more reasonable inference is that certain specific assumptions, such as that of profit maximization, fail to generalize beyond capitalist society, whereas others, such as the importance of travel distance in determining whether a transaction is worthwhile, generalize pretty well. Comparative study helped distinguish what was general in economic behavior from what was culturally specific. Ironically, those who have concluded on the basis of specific failures of general models that models will always fail are guilty of the same charge of overgeneralization they have leveled at the formalists: they have generalized from a particular general theory to all general theories.
I would like to suggest that, as in the case of cross-cultural studies of production and exchange, documenting cultural variation in forms of interpersonal conflict can be immensely useful, but not as a weapon against the project of generalization. There is a difference between showing that a particular generalization fails to match the range of observed behavior and showing that generalization as such is not a worthwhile pursuit. The value of a deep investigation into a specific society is maximized not by invoking the society's unique traits to justify particularism but by obliging general theories to be better -- that is, to reformulate their core concepts so as to accommodate a greater range of observed behavior. There is always a danger that general propositions will thereby become so abstract as to be vacuous, but that is in the nature of the enterprise: a good theory consists of propositions abstract enough to hold across a broad array of settings but concrete enough to still say something meaningful.
Excerpted from Collision of Wills: How Ambiguity About Social Rank Breeds Conflict by Roger V. Gould Copyright © 2003 by Roger V. Gould. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
Table of Contents
Foreword by Peter Bearman
1. Conflict, Honor, and Hierarchy
2. Dominance Relations
3. Strife out of Symmetry
4. Solidarity and Group Conflict
5. Conflict and Social Structure
6. Honor and the Individual