Civilizing the Body through Time

Pieter Spierenburg


Copyright © this collection Pieter Spierenburg 2013
Copyright © all materials Pieter Spierenburg 2013, except:

Introduction: abridged version of material copyright © Ohio State University Press 1998, and material copyright © Board of Trustees of the University of Illinois 1996;

Chapter 1: abridged version of material copyright © Board of Trustees of the University of Illinois 1996;

Chapter 2: copyright © La Decouverte, Paris 2009;

Chapter 4: copyright © Social Science History Association 2004;

Chapter 6: copyright © Cambridge University Press 2011;

Chapter 8: copyright © Stichting Leidschrift 2001.

The right of Pieter Spierenburg to be identified as Author of this Work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988.

First published in 2013 by Polity Press

Polity Press
65 Bridge Street
Cambridge CB2 1UR, UK

Polity Press
350 Main Street
Malden, MA 02148, USA

All rights reserved. Except for the quotation of short passages for the purpose of criticism and review, no part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher.

ISBN: 978-0-7456-6398-2

A catalogue record for this book is available from the British Library.

The publisher has used its best endeavors to ensure that the URLs for external websites referred to in this book are correct and active at the time of going to press. However, the publisher has no responsibility for the websites and can make no guarantee that a site will remain live or that the content is or will remain appropriate.

Every effort has been made to trace all copyright holders, but if any have been inadvertently overlooked the publisher will be pleased to include any necessary credits in any subsequent reprint or edition.

For further information on Polity, visit our website:



Introduction: Violence and Punishment within Civilizing Processes
PART ONE Violence
1 Long-Term Trends in Homicide: Amsterdam, Fifteenth-Twentieth Centuries
2 Homicide and the Law in the Dutch Republic: A Peaceful Country?
3 Violence and Culture: Bloodshed in Two or Three Worlds
PART TWO Punishment and Social Control
4 Punishment, Power and History: Foucault and Elias
5 Monkey Butt’s Mate: On Informal Social Control, Standards of Violence and Notions of Privacy
6 “The Green, Green Grass of Home:” Reflections on Capital Punishment and the Penal System in Europe and America from a Long-Term Perspective
PART THREE Civilizing the Body in Human History
7 Elites and Etiquette: Changing Standards of Personal Conduct in the Netherlands until
8 Civilizing Celebrations: An Exploration of the Festive Universe
9 The Body’s End: Death and Paradise in Human History
Epilogue: A Personal Recollection of Norbert Elias and How I Became a Crime Historian
Archival Sources


Violence and Punishment within Civilizing Processes


Violence and punishment are intimately related. For one thing, the first may result, for the perpetrator, in the second, especially in modern societies. In its turn, punishment may involve violence. This was notably the case with the public torments inflicted on serious offenders in the early modern period. More generally, as I argue in chapter 6, in all societies the criminally condemned have in common that they feel the state’s monopoly of force striking at them in one way or another. Ultimately, even the mildest form of punishment depends on a credible threat of violence. Throughout this book I will illustrate my argument with real-life examples. The reader will get to know why fights between Jews and Christians often took place on the Blue Bridge in Amsterdam; how a yellow bulldog betrayed a killer who had hoped to remain undetected; what happened to a thief whom the citizens stopped by splitting his skull with an axe; why the executioner was to activate the guillotine by cutting the rope that holds up the blade with a sword; what the punishment of Adam and Eve reveals about the work and fears of our ancestors.

Let me first delineate the two main subjects covered in this book, beginning with punishment. I am using this concept both in a broader sense and in a narrower sense than is common in everyday language. In everyday parlance we can say, for example, that a parent punishes a disobedient child with a view to raising it properly. Early modern people often viewed a disaster as God’s punishment for their sins. In a modern setting, a woman’s friends, for example, may view her extramarital affair as the punishment she gives to her husband for his earlier unfaithfulness. Such uses of the word, the examples randomly chosen, are excluded from my discussion. I am focusing on punishment that is in one way or another judicial. At the same time, however, my discussion ranges more broadly than just the act of punishing itself. The discussion extends to the cultural manifestations and the social embeddedness of punishment, as we see in the drama of the scaffold, for example. Therefore, I am additionally using the concept of social control, most notably when I am dealing with informal regulation within village or neighborhood communities.

Whereas most people take the meaning of the word “punishment” for granted, the definition of “violence” often leads to controversy. In our society this word has a decidedly negative connotation, which induces many to extend the term to all kind of situations considered undesirable. Yet, in popular parlance, in English no less than in other languages, it is usually clear what is meant by violence. Here I am purposely employing a restricted definition, which is attuned to the everyday understanding of the word. I include in the category of violence all forms of intentional encroachment upon the physical integrity of the body. That formula excludes forms of encroachment for medical reasons, as well as unintentional harm such as that caused by traffic accidents. Moreover, I reject notions like structural or psychological violence, which appear to me excuses for an ideological argumentation rather than scholarly concepts.1 Readers who disagree with me on this point may insert the word “physical” each time there is talk of violence. Though restricted, my definition is not narrow. It includes minor encroachments upon bodily integrity which a court considers too trivial to prosecute. And, perhaps more importantly, it includes state violence such as armed police action, corporal punishment and warfare. This formal definition, however, should not be confused with the book’s program. In fact, I am largely ignoring the two extremes of warfare, on the one hand, and minor slaps and bruises in conflicts between citizens, on the other.

Having carefully delineated the two main subjects covered in this book, I am eager to open up the field again. To the extent that I am a criminal justice historian, I am far from practicing this sub-discipline in isolation. Key issues relevant for all historians and social scientists, such as gender, the development of the state and popular vs. elite attitudes, all figure in my discussion as explanatory elements. Moreover, I am exploring several corollary developments to the decline of violence and the transformation of punishment, among which are the refinement of manners and the rise of new forms of festivity. All this follows directly from my scholarly approach. I am not studying a phenomenon, like dueling for example, for its own sake, followed by the question whether A, B or C or a combination of them accounts for its rise and demise. As I conclude in chapter 4 with respect to punishment, social phenomena cannot be explained in terms of cause and effect. Historical explanation means clarifying what a particular custom or belief tells us about the entire “figuration” of which it forms part and the changes in that figuration. Thus, the figuration that involved duelists and their opponents and onlookers also featured the decline of feuding and the rise of elitist notions of stylized combat – and much more. It is always a question of interrelationships.

The theme of honor, central to various earlier works of mine as well as to many chapters of this book, lies at the crossroads of these complex interrelationships. For one thing, honor plays a crucial role in both punishment and violence. In punishment we encounter this theme mostly from the negative side, as dishonor. A criminal sentence often brought dishonor to the convict. Most notably, all penalties meted out by an executioner, even if not physical, made the person suffering them infamous. The infamous treatment of the body, whether male or female, extended to the whole person. Thus, the dishonor brought about by punishment was relatively independent from gender distinctions. This was quite different in the case of violence. Whereas the relationship between honor and the body was equally strong, in violence gender distinctions mattered a lot. For centuries, masculinity was linked up with a violent life-style, whereas women were expected to largely refrain from aggression. The interrelationships between gender, honor and the body, then, and the changes they underwent, form a crucial background in particular for the chapters about violence.

Gender, Honor and the Body

Anthropologists as well as historians have studied conceptions of honor, including the intimate relationship between male honor and violence.2 Honor has at least three aspects: a person’s own feeling of self-worth, this person’s assessment of his or her worth in the eyes of others and the actual opinion of others about her or him.3 The criteria of judgment depend on the socio-cultural context. The relationship between physical force and male honor has been observed in many societies, but it is not universal. Among the Suriname maroons, for example, it counts as a stain upon a man’s honor if he fights out a conflict or reacts to an insult with aggression. Only in the case of adultery, is the aggrieved husband accorded a limited right to beat up his rival; all other conflicts have to be solved through the institution of the palaver. This attitude has characterized the Suriname maroons for at least 250 years, Thoden van Velzen argues, and it is connected to the uxorilocal organization of their society.4 According to the anthropologist Frank Henderson Stewart, the concepts of honor prevalent among the Bedouins, and possibly in the Arab world as a whole, are fundamentally different from the European model. In contemporary Bedouin societies no particular connection between male honor and violence exists.5 Conversely, as chapter 3 shows, in some Asian societies the link between male honor and violence appears to be very pronounced.

For now, I restrict myself to Europe since the middle ages. For centuries, honor depended on the body or, in Blok’s words, the physical person.6 Appearance was crucial for one’s reputation. Honorable men were symbolically associated with strong, awe-inspiring animals and dishonorable men with small or weakened animals. Thus, an old English law mentioned by the legal writer Bracton says that if a nobleman is found guilty of rape, the penalty is castration . . . for his horse or his dog, coupled with cutting off its tail; or his hawk is deprived of the sharp end of its beak and its claws. Through the disabling of his noble animals, the nobleman himself is symbolically disempowered.7 We find an analogy between the disempowered human body and dishonor in the writings of a Polish convert to Islam at the end of the seventeenth century. His conversion did not prevent him from frowning upon the Ottoman custom of entrusting political functions to eunuchs: “When they are successful, the most important posts of the Empire are given to them. While they are properly speaking only half-men, this does not alter the fact that they are often at the heads of the armies and they have governed the greatest of the provinces. One sees by this that the jobs are not always given to those with the most merit, or those who are the most capable of possessing them.”8 In view of this close connection between the body and honor, it is intriguing that none of the six volumes of a recent book series entitled The Cultural History of the Human Body list “honor” in the index.9

Despite this omission, the body, honor and gender are all related. Bodies, being male or female with few exceptions, form the basis of gender; gender gives rise to a dual concept of honor; honor shapes the experience of the body. The inherent circularity assures that no element of this triangle is the principal determinant. The relationship is one of interdependence: if there is a change in one element, the others are likely to change too. Nevertheless, in order to get to grips with these complex interdependencies, we may break them down into developments in three areas: the body and gender; gender and honor; honor and the body.

For the body and gender the crucial periods seem to have been the middle ages on the one hand and the late eighteenth century on the other. Important work has been done on the first period.10 Although the authors concerned criticize each other on points of detail, they agree that the medieval conception of sex differences left room for much ambiguity.11 In the view of contemporaries, to be female or male largely depended on character and habits. In accordance with this, the process of generation did not just offer two possibilities. Human specimens such as a virago, an effeminate man or a hermaphrodite could be born just as easily.12 The body’s sexual identity had fluid boundaries. Christ’s body in particular was often pictured as half female. His side wounds, a source of food, were likened to Mary’s breasts.13 Although Monica Green recently argued that medieval people did view the differences between women and men as absolute, she also cites a source which actually confirms the contemporary notion that the generation of mixed human types was possible. According to medieval physicians, the birth of an effeminate man or a masculine woman depended on the temperature at the spot in the womb where conception took place.14 Paradoxically, in the medical description of all human bodies, including women and the mixed types, one body, the male, was used as a referent. The notion of sexual ambiguity persisted into much of the early modern period.15

Thomas Laqueur offers a thesis to account for these observations.16 Although this thesis has been heavily debated, criticism concentrated on his idea of a transition from a “one-sex model” to a “two-sex model” and his supposed neglect of popular beliefs.17 Here I am focusing on another element, the relationship between sex and gender. Moreover, it appears that recent scholarship in majority still favors Laqueur’s thesis.18 According to him, gender came first in the middle ages. In medieval people’s minds, the socio-cultural experience of being male or female, or anything in between, had primacy and biological sex was made to fit it. Essentially, this conception of sex differences persisted during the early modern period. From the middle of the eighteenth century onward, however, the relationship was reversed. Sex came first now. Biology was seen as the basis of character, and biology left room for just two sexes. Sexual identity and, as a consequence, gender identity became much more strictly demarcated. This new view was especially pronounced toward the end of the nineteenth century. By then, according to Robert Nye, doctors and biologists had elaborated a standard anatomical and physiological model of masculinity, defining its features as hygienic norms with which all men should comply.19 Masculinity and femininity had become binary opposites.

At about the same time as this stricter demarcation of the sexes and gender roles was elaborated, the contrast between male and female honor weakened. Since the early modern period, notions of female and male honor have gradually converged. Of course, they remained distinct to some extent. The process of convergence had two main aspects: the active-passive contrast in gender roles became less pronounced, and men, like women before them, had to take moral standards into account. Women’s honor had always been based primarily on issues of morality. Foremost, it depended on a reputation of chastity, but in the sixteenth and seventeenth centuries a clean slate with respect to sorcery was important too. A chaste woman was a modest woman, true to the demand of passivity.20 For men, on the other hand, the domain of sex originally meant activity: the protection of one’s own womenfolk from predators, and trying to seduce others’ womenfolk. This attitude prevailed not only among elite men, but also among men of lower social status. Popular customs testify to this at least until the sixteenth century. When a cuckolded husband was subjected to the ritual of charivari, for example, he was mocked as a loser by his fellow men, rather than burdened with moral outrage.21 Attitudes slowly changed during the early modern period. Restrictive demands on men, especially from religious moralists, became stronger. Obviously, the male gender role continued to comprise a more active stance than the female, but the quest for sexual adventure was increasingly proscribed from it. By the nineteenth century, male honor, too, had become associated with sexual self-restraint, at least among the middle classes.22

Thus, a shift in the way the body and gender were perceived was accompanied by a transformation in conceptions of gender and honor. That transformation, however, appears to have come about more gradually, extending over a period roughly from the sixteenth century to the nineteenth. Moreover, whereas the shift in the perception of the body and gender implied a strict demarcation between men and women, the transformation in conceptions of gender and honor implied convergence. Indeed, the two changes were loosely related rather than connected in a cause-and-effect relationship. In this complex web of interdependent developments, the third area to be reviewed, that of honor and the body, was involved too. In that case, we are talking primarily, but not exclusively, of male cultures.

Honor can be inwardly or outwardly oriented. Association with the body means being linked to the body’s outer appearance in particular. The outside is considered to reflect inner qualities, so appearance takes primacy. Conversely, in its spiritualized form, honor is linked primarily to inner virtues. It depends on an evaluation of a person’s moral stature or psychological condition, in which outer appearance plays a much less significant part.23 Inward and outward are two end-poles of a continuum. The conceptions of honor prevailing in a particular society are never located completely at one end or the other, but always somewhere between these extremes. In Europe over the last 300 years or so, conceptions of honor appear to have moved in the direction of spiritualization. By implication, their association with the body was strongest before this process of change set in. During most of the preindustrial period male honor depended on a reputation for violence and bravery. An honorable man commanded respect; as a patron, he protected his clients and he dealt roughly with an enemy who dared to encroach upon his property. In the streets he kept rivals at a distance, at arm’s length at least. When insulted, he was prepared to fight. Well into the seventeenth century, these attitudes were manifest in almost every European country where the subject has been investigated.24

The gradual change in the direction of spiritualization did not just mean the reduction or removal of the element of force from the prevalent conception of honor. The change also had a positive side, in the sense that something else took the place of force. Thus, by the seventeenth century, economic solidity was a major supplementary source of honor for men. A reputation of engaging in sloppy affairs would greatly diminish a man’s honor; “thief” was a common word of insult. Clearly, this implies the rise of a new ideal of masculine behavior. As Martin Wiener notes for England, the earliest attacks upon traditional manhood can be traced back to the sixteenth century.25 In the Netherlands, by the eighteenth century, the honor of male citizens was based in large part upon being a decent housefather.26 According to still other historians, the decisive moment came at the turn of the nineteenth and twentieth centuries, when a more gentle and domesticated type of man emerged; they speak of the breakthrough of a new masculinity.27 It can be argued, however, that taking pride in not being considered a thief was the earliest manifestation of a new masculinity, preceding its later counterpart by some three centuries.

This discussion of gender, honor and the body can be viewed as an example of historical explanation in terms of interdependent long-term developments, instead of causes and effects. Up to now, however, I have dealt with only one out of two types of honor: that accorded to a person by his or her peer group. This is the type most easily lost. If a man fails to live up to the expectations of honorable conduct – whatever they are in the period examined – of his peers, he loses (part of) his honor. Conversely, when a man does more than expected, his honor increases. Thus, in a society in which male honor is based on physical bravado, chasing away a particularly fierce enemy may make a man more honorable. The second, more stable type of honor derives from a person’s rank. Its primary manifestation is respect by one’s social inferiors. Usually, this type of honor can only be lost when a person is deprived of his or her rank. In another important conceptual distinction, Stewart (1994) refers to the two types as horizontal and vertical honor, respectively. This distinction helps us to extend the discussion about honor and violence, albeit a little speculatively, to the very long term of human history.

The Very Long Term

I examine the very long term in particular in the last two chapters, in which two scholars, the historian William H. McNeill and the sociologist Johan Goudsblom, figure prominently. Whereas the second bases himself explicitly on the theoretical work of Norbert Elias, the first offers many points of comparison with this work. McNeill’s Plagues and Peoples (1977), for example, can be read as a creative contribution to Elias’ theory of the triad of fundamental controls. According to Elias, all interdependent long-term processes ultimately form part of one of three overarching long-term developments: changes in the relationship between humans and the nonhuman (physical and biological) world; changes in the relations among people themselves; and changes in the relationship of all people with their individual selves. The term “control” refers to Elias’ complementary thesis that in the long run, in these three areas, the trend is in the direction of greater control. He admits, however, that this is most difficult to substantiate for the third area.28 McNeill shows how specific developments within the first and second areas involved an increase of controls over the very long term, without any of the participants realizing this. Within the second area it concerned the twin developments of societies becoming more populous and increasingly in contact with each other over longer distances. The concomitant development in the first area concerned changes in the relationships between humans and micro-parasites (bacteria and viruses). Ever larger “disease pools” emerged, in which humans acquired antibodies against more and more diseases and micro-parasites became less virulent.

Processes involving pacification and the relative monopolization of force belong to the second area, that of inter-human relations, while changes in people’s propensity for aggression belong to the third area comprising the relationship of all people with their individual selves. The interdependence between these two can be demonstrated most clearly for early modern Europe, with the rise of nation-states and the concomitant decline of homicide. It is worthwhile, however, to briefly examine the very long term. Goudsblom (1998) does so, but primarily for one side of the coin, that of the relative monopolization of force, and hardly for the propensity for aggression. This monopolization is relative because every individual act of violence implies a breach of the monopoly. “Force” essentially stands for the organized, “professional” exercise of violence. Goudsblom hypothesizes a sequence of three world-historical phases in the monopolization of force. During a first phase, all adult males monopolized force, excluding women and children from its exercise. This phase is hypothetical rather than empirically demonstrable; its beginnings probably coincided with the differentiation between hunting as a male activity and gathering as a female one. The means involved were most likely of an ideological and psychological nature: men successfully persuaded women and children that force constituted a male domain. Goudsblom mentions rites of initiation and “taboos” in this respect. During a second phase, an elite of warriors monopolized force, excluding peasants, artisans and priests from its exercise. In this case, the means were primarily of a material nature, consisting of a complementary monopoly on bearing arms. Goudsblom calls the social figurations associated with this phase “military-agrarian” societies.

The third phase is the one that many scholars are most familiar with. Relatively autonomous warrior elites increasingly had to yield to larger organizations. Force came to be monopolized within the framework of the institutions that henceforth were called “states.” All specialists in violence were either incorporated into the state or eliminated, and practices like feuding, exercising one’s own private justice and keeping armed retainers were increasingly banned. There is no need to elaborate on these well-known processes. I am, instead, concerned with the first two phases of the monopolization of force, aiming to take Goudsblom’s analysis one step further. He does not venture to speculate about the question of what these two monopolizations meant for the propensity for aggression among the groups excluded from the “official” exercise of force. Indeed, we will never dispose of, say, accurate homicide rates for the millennia concerned. Yet, it is possible to hypothesize a little about the related themes of male honor and the propensity for aggression.

In this exercise, another theoretical contribution from Norbert Elias forms our point of departure. I already introduced the term “figuration.” In the course of interdependent long-term processes the figuration that all people constitute together over generations gradually changes, until it finally becomes another figuration. However, Elias adds, elements from an earlier figuration often remain present, in a slightly transformed manner, in the subsequent figuration or even in still later ones. This theoretical contribution may help to shed light on two riddles in the history of violence: its tenacious link with honor and the surprisingly constant fact that violence, in particular homicide, is so disproportionately committed by males. I hypothesize that both may be considered as remnants from earlier figurations. First, serious violence as a male preserve, observed in many societies until today, is a remnant from the figuration that resulted from the first phase in the monopolization of force. When force was monopolized even further, the propensity for “ordinary” aggression largely remained with men. My second hypothesis locates the origins of the link between male honor and violence in the second phase of the monopolization of force. The traditional, body-related concept of honor originated as the vertical honor of the warrior caste: they took pride in what they were doing and got respect for it. Later on, but still within military–agrarian societies, peasants and artisans adopted elements of this honor code and transformed them into a vehicle of primarily horizontal honor.

Civilizing Processes and the Study of Homicide

From the very long term I return to the more familiar theme of civilizing processes in European history since the late middle ages. Civilizing processes are reflected in both the decline of violence and the transformation of punishment. Here we must inquire into the directions for the study of homicide following from Elias’ theory of civilization. Once more, this follows from my conviction that historical explanation means elucidating the interdependence between various long-term processes. Theorizing and gathering new evidence should always go together, in a two-way process. As new empirical data necessitate us to revise our theories, these very theories suggest what kind of data to collect, and how to categorize them and group them together. Thus, the theory of civilizing processes entails recommendations on how to study homicide – one of this book’s subjects.

One of the advantages of Elias’ approach is that he wishes to keep his analysis as free as possible from the intrusion of moral standards or judgments. Several participants in the earliest homicide debates, notably that between Lawrence Stone and James Sharpe in the first half of the 1980s, implicitly assume that a high level of violence in past communities automatically translates into a low quality of life in them. In particular, they tend to equate violence and lack of affection in personal relationships. While Sharpe appears to think that a low level of tensions in early modern English villages would contradict the thesis of their being relatively violent, Stone tries to bolster up his argument with the statement that life was “not very pleasant” in these villages.29 Such an equation is anachronistic, however. Elias, drawing on the Freudian notion of a link between love and aggression, claims that impulses for both affection and aggression became subject to constraints as a result of the same overall process. Consequently, we should not be led astray by the current assumptions of our time, according to which violent behavior is always destructive, “dysfunctional” and devoid of meaning. Such an unrealistic view of violence can only distort our historical judgment of aggressive behavior in the past. In our analysis of peopl e’s propensity for kill ing, then, we should concentrate on the mode in which aggression was expressed and the extent to which different modes were socially accepted or rejected.

It would be an exercise in purging moral judgments from our scholarly view if we asked ourselves whether it is possible to commit a “civilized” murder. The question follows directly from a primary concern of Elias’ theory – that is, what kind of constraints did people impose on themselves and on others? If increasing affect control, the taming of spontaneous drives and impulses, is indeed the dominant socio-psychological trend over the last seven centuries or so of European history, a high incidence of deliberate killings today would not be incompatible with it. Such a killing requires a high amount of rational planning and rest raint of momentary impulses. Elias’ theory about affect control, then, would not necessarily predict that we find a declining trend in homicide rates. Instead, the proportion of “killings in affect,” as a result of sudden rage, may have declined, while that of care fully premeditated murders may have remained stable or even risen.

Two alternative reactions to this proposition are possible. The first, which I will reject, consists of a separate study of murder and manslaughter. This might be necessary if we want to distinguish “killings in affect” from the “civil ized” ones. However, there are no less than four reasons why it is better after all to combine the figures for murder and manslaughter into one homicide rate. The first is that precisely the affect-control component of Elias’ theory is highly contested, since it is so difficult to substantiate it empirically (as already said, he admits this himself). Did medieval people really have fewer self-restraints, or did they simply control their behavior in a manner qualitatively different from ours?30 Second, the empirical evidence on related developments, such as the changing attitudes to slavery, punishment and animal sports, definitely indicates that the dominant trend within the civilizing process moved in the direction of non-acceptance of violence, physical subjugation and the deliberate infliction of suffering generally. If we fail to take this evidence into account, we would go too far toward the extreme of relying only on deduction.

The third and fourth objections against separating murder cases from manslaughter cases are of a methodological nature. The definitions of these two categories may vary, historically as well as in individual instances. Finally, we cannot be sure in individual cases: if we calculate separate rates for murder and mansl aughter, we are in fact counting the outcomes of judicial trials. To conclude, our calcul ations have little validity, unless we combine all cases of killing into one homicide rate. My rejection of the idea of calculating separate murder and manslaughter rates leaves us with the second device for distinguishing between different sorts of killings: to supplement the raw data with an analysis of the entire context in which each single homicide or assault was committed. Of course the state of the evidence has to allow for such an analysis. To get to grips with the context, I am introducing two “axes of violence.”

Homicides as well as assaults can be characterized according to their position on two related but distinct axes: impulsive violence vs. planned or “rational” violence, on the one hand and ritual or expressive violence vs. instrumental violence, on the other.31 Note that these four categories are extreme poles of a continuum. The archetype of a highly impulsive killing is the tavern brawl in which a knife is drawn and one of the fighters is stabbed to death. At the other end of the specfrum we find deliberate acts of violence. A carefully planned murder out of jealousy or revenge, for example, may be called rational, even if the perpetrator is caught. The impulsive–rational axis has to do mainly with the psychological state of the person who engages in violence. The social meaning of the act, on the other hand, is the determining factor in the ritual–instrumental axis. Highly ritual violence belongs to a social context in which honor and physical bravery are valued and linked. Whether homicidal or not, highly ritual violence is guided by implicit cultural codes and often its primary aim is to degrade the victim. It is violence for its own sake. On this axis, the opposite pole refers to violence used in order to get something, as with mugging, rape or loan-sharking. It should be added that the violence used in rape often serves both to subdue and to degrade the victim. Such cases, obviously, should be coded somewhere in the middle of the ritual–instrumental axis.

In principle, the two axes are independent from each other. Consequently, the position of a single act of violence on the first axis (near one end, near the other or around the middle) tells us nothing about its position on the second axis. A highly instrumental stabbing, in the course of a robbery for example, can be done either with a high degree of planning or in a moment of sudden greed. Likewise, an act of violence may have a score near the ritual pole on the axis in question and near the impulsive pole on the other. This is something which several historians refuse to accept; they believe that ritual always implies planning and hence self-restraint. The codes of ritual violence, however, constitute a fixed pattern that is ready in people’s heads.32 A knife fighter knows that he degrades his opponent with a long cut to his face and is perfectly capable of inflicting one upon a sudden fit of rage. Nevertheless, my hypothesis is that any long-term trend would be from impulsive to rational and from ritual to instrumental violence.

The identification of two axes of violence suggests still another hypothesis. Both the highly ritual and highly impulsive violence of past centuries often had a distinct community-character. The first derived its meaning from being understood by all participants, and the second was closely associated with daily sociability. Killer and victim often were residents of the same local community. In a populous place they might be strangers to each other, but they usually belonged to the settled population. Homicides were public events, at the center of community life. To a large extent, this is no longer the case today. Serious violence has retreated partly to the margins of society. A large number of contemporary homicides are connected with a property crime or with illegal economic activity such as the drug trade. This applies to instrumental as well as to rational violence; in the latter case, we can think of the liquidation of competit ors. Thus, qualitatively speaking, marginal ization was one of the major long-term developments in homicide. The trend was from violence at the center of local communities to violence practiced by groups with a professional interest in crime.

The increasing proportion of killings within the biological family, first identified by researchers in the 1980s, forms another conspicuous development, important for a contextual analys is. If family homicide maintains a relatively stable level even today, this would be compatible with the theory of an increase in affect regulation. As affects and emotions were the subject of increasing constraints in the wider society, the nuclear family came to serve as an island where emotions were cultivated. Historians such as Mitterauer (Mitterauer and Sieder 1977) argue that, because of this development, families have become more crisis-prone. James Cockburn is the only historian skeptical about the thesis of an increasing share of killings within the family. In this connection, he makes two claims. First, he thinks that homicides on spouses in the early modern period were seriously underreported and those on lovers often unidentifiable. Second, according to a somewhat elusive argument, Cockburn posits that infanticide should be included in the category of family homicide. That operation would raise the level of family homicide in early modern England to over 30 percent.33 I disagree with both claims.

The difficulty of identifying lovers as victims of a homicide is probably peculiar to the uninformative English indictments. For the rest, the biological family and legal spouses have received too much attention from historians studying homicide. Understandably, they have looked for a factor that could be easily quantified. From a theoretical point of view, however, the crucial question is not whether killer and victim were related by blood or marriage, but whether they had an intimate relationship. It is only in the latter case that we suspect the killing to have been the outcome of tensions within this relationship. In the categorization of killer–victim relationships, then, intimates should be kept separate from acquaintances and total strangers as well as from non-intimate relatives. It follows that the question whether or not to rank infanticide with the family homicides is largely irrelevant, because, in any case, it should not be included in the category of killing an intimate person. The children in question never were granted the time to become intimates and their premature death was not the outcome of a protracted conflict between the perpetrator and the victim.

Thus, infanticide rates tell us little about people’s propensity for aggression and much more about shame and desperation. When a mother kills her infant child, there is neither a fight nor a robbery. Moreover, the perception of the act by the killer and those prosecuting her may be radically divergent: from its being “something many women might be so unfortunate as to have to do one day, because no other option is available” to “an inexcusable assault on Christian morality.” The tremendous differences in the social contexts of infanticide and homicide make the former into a distinct category. Again, an objection is possible: this line of reasoning exceeds the bounds of a historian’s neutrality, since infanticide obviously involves the killing of a human being. However, so does the death of a soldier at the hands of an enemy. Historians are constantly classifying, including some acts in a specific category and excluding others, according to their evaluation of the social context. This is fine, as long as the process of categorizing is made visible. Hence, we should always construct murder rates with and without the killing of infants.

The Chapters of This Book

The chapters that follow have been chosen for their mutual coherence and their broad scope. They all apply the principle that historical explanation means elucidating the interdependence of at least two long-term processes. The theoretical work of Norbert Elias is a major source of inspiration (and I write “inspiration” on purpose, because I want to creatively expand on it), but I also discuss other theorists. In each case I have thoroughly revised the essay and updated it where necessary. The chapters constitute a mix of a few of my regularly cited articles and others which were unavailable in English up until now. Three were originally published in Dutch, one in French and one in Chinese. All speak to broad questions and several contain international, even inter-continental, comparison. Also, the three chapters focusing on the Netherlands alone do so from the perspective of issues having a much wider relevance.

Chapter 1 presents my work on homicide in Amsterdam, updated for recent years. The chapter insists that, next to constructing homicide rates, it is of crucial importance to gather evidence indicating trends in the character and context of violence. As yet, few scholars of other countries have taken up this program. Chapter 2 equally examines developments in the Netherlands with an eye on European-wide issues. It includes a discussion of homicide in self-defense – a subject that some scholars discuss with reference to legal rules but which few have studied from texts of jurisprudence. It also examines public attitudes concerning the death penalty for manslaughter, thereby extending my analysis of the criminalization of homicide – a process that occurred throughout Europe but which few scholars have tackled head-on. The third chapter subjects the link between historical violence and male honor to inter-continental comparison. It maintains that the relatively high homicide rates of the United States have to do with peculiarities of the American process of state formation and it ventures into comparison with a few Asian countries such as Taiwan.

The section on punishment and social control opens with an assessment of the theories of Norbert Elias and Michel Foucault, viewed in relation to the history of punishment in particular. In fact, chapter 4 implies a re-assessment, compared to the more outright dismissal of Foucault in some of my earlier work. I argue that Elias and Foucault converge to the extent that they find historical study indispensable for an adequate understanding of present-day society and that they both maintain that power is everywhere. They diverge, nevertheless, when it comes to the pace of penal change. Chapter 5 takes the discussion to the history of informal social control. I argue that, while informal regulation within communities and self-help by citizens were common in the early modern period, internal cohesion and communal controls continued to characterize many workers’ neighborhoods until the mid twentieth century. When notions of privacy began to change, informal social controls declined. The essay also discusses the changing appreciation of violence by citizens who attempt to stop thieves and other lawbreakers. The sixth chapter essentially represents a renewal of my vision of the long-term history of the penal system, in Europe as well as America. Next to civilizing processes, it introduces Elias’ theory of the long-term diminution of power differentials between social groups as an explanatory factor. It ranges from the sacralization of executions around 1500 to the modern resurgence of punitiveness on both continents, engaging along the way with James Q. Whitman’s thesis.

The last section discusses several corollary developments to the decline of violence and the transformation of punishment. Chapter 7 focuses on the refinement of manners in the Dutch Republic. Its broader interest lies in the counter-model to the uses of etiquette and the balancing of power in courtly society as analyzed by Elias. It turns out that, despite an aversion to dueling, the Dutch elites of the seventeenth and eighteenth centuries were eager to adopt French aristocratic manners. Chapter 8 discusses long-term changes in festive behavior. Assessing the theoretical contributions of Victor Turner and Emile Durkheim, it goes back several millennia, but it ends with a discussion of the civilizing of festivals during the last two centuries. Chapter 9 covers the longest period of time. It draws the themes of religion and death into the discussion, imagining what it meant for humans when they first began to realize that they were mortal. The book concludes with a brief personal retrospective, commenting on Norbert Elias’ influence on my becoming a crime historian.




Long-Term Trends in Homicide

Amsterdam, Fifteenth-Twentieth Centuries


Homicide is haunting historians. Some welcome its possibil ities for a study of the long term, calling it “the one crime for which the evidence is most reliable.”1 Others, more skeptical, are trying to prove that any visible trend is only a chimera. This clash of arguments and approaches particularly concerns the period before the availability of national statistics. Evidence from that period is indispensable, if we want to make meaningful statements about the long term. Throughout this chapter, I will usually speak of the homicide rate, meaning the annual average, over a specified period, per 100,000 population, in a specified area.

The study of homicide rates has progressed much since the first ten years (1981–91) when England was the main focus of research and the long-term decline of violence since the middle ages was first established. Yet some of the problems emerging from the debates of that period have not been satisfactorily solved. In the most recent synthesis of the quantitative evidence about murder by Manuel Eisner (2001, 2003), for example, it is not specified which of the data are based on coroners’ records and which on prosecutions alone. He also conflates the categories of family homicide and homicide on intimates. On the other hand, recent syntheses pay much more attention to the trend reversal since about 1970, which was still hardly visible when the pioneers wrote their studies.

The Inadequacy of Trial Figures

Examining the Dutch evidence available when I started my research, we can definitely conclude that frequency rates based on prosecuted homicide alone are unreliable. Combining the work of several historians, we might construct a graph of murder and manslaughter cases tried by the city’s court from the beginning of the sixteenth century to the beginning of the nineteenth.2 This enterprise leads to a surprising conclusion: the average annual level of prosecuted homicide in Amsterdam tended to remain constant, at about 1, over the entire period from 1524 to 1811 (although in the short run there were peaks and quiet decades). The city’s population, however, rose dramatically from 15,000 in 1500 to 100,000 in the early decades of the seventeenth century, and to well over 200,000 at the end of that century. Thereafter there were minor fluctuations, but the population always remained over 200,000. Thus, a graph of homicide trials per 100,000 inhabitants would simply be the inverse of the population curve. The figure would be less than 0.5 per 100,000 throughout the eighteenth century. It is highly unlikely, although not impossible, that this graph would bear any resemblance to the real incidence of killing in Amsterdam.

The evidence on default procedures, discussed by Sjoerd Faber and Marijke Gijswijt-Hofstra, is one indication that the real incidence of killing in Amsterdam was higher than the prosecution records suggest.3 Their data from two archival series allow me to calculate that the rate of homicides with identified killers (i.e. default procedures and prosecuted cases combined) lay between 1.5 and 2.1 during the period 1680–1790 and that it was 0.7 during the period 1751–1801. These figures indicate a declining trend, but the total number of homicides committed was higher than the rate of homicides with identified killers.