The unease with civilization

I The unease with civilization

In the 1968 postscript to The Civilizing Process, Elias (2000: 449–84) argues that critics have misunderstood his approach because they assume a theory of ‘civilization’ relies on some kind of automatic historical process, whether metaphysical, teleological, or evolutionary. Elias argues that such mechanistic concepts inform ‘the current pre-scientific notion of the civilizing process’, which he dismisses as ‘vague’ and ‘mystical’ (pp. 451–2). His approach, by contrast, explores the ‘factual core’ of the pre-scientific notion, through an empirical investigation of the causal mechanisms driving long-term, tandem transformations of affective and social structures (p. 451).

Elias maintains that it is possible to carry out a scientific sociological analysis of the civilizing process, by observing long-term directional movements visible within qualitative transformations of both affective and social structures. In ‘affect and control structures’, such trends take the form of a movement toward ‘the increased tightening and differentiation of controls’. In social structures, the parallel trends take the form of a movement toward ‘a higher level of social differentiation and integration’ (p. 450). Such trajectories require no mystical teleology: their causes can be studied through ‘an undogmatic, empirically-based sociological theory of social processes in general and of social development in particular’ (p. 452).

Elias recognizes that some critics read a normative judgement of different forms of society into his work. He argues that this objection arises from the erroneous belief that the concept of ‘civilization’ operates as a dichotomous variable, which a society either possesses completely, or lacks entirely. In his work, ‘civilization’ is instead understood as a continuum that, while ordinal in character, has no ‘zero point’ or fully-achieved apex (p. 135). Directional movement along that continuum can be analysed empirically, with no value judgement implied:

It can never be said with certainty that the people of a society are civilized. But on the basis of systematic investigations referring to demonstrable evidence, it can be said with a high degree of certainty that some groups of people have become more civilized, without necessarily implying that it is better or worse, has a positive or negative value, to become more civilized. (p. 456)

With these responses, Elias addresses the most common theoretical objections to his work (cf. Goody, 2002, 2006). His responses are consistent with statements made repeatedly throughout The Civilizing Process, and he thus has a basis for complaining that his critics have not attended closely to his text. Yet such responses have not dispelled the impression that his work channels Eurocentric conceptions of history, even if critics have sometimes failed to explain precisely how this is the case. In the next section, I explore the basis for this continuing unease with Elias’s work, by examining his continuum concept of civilization and the ordinal metaphors he uses to imply that long-term transformations of affective and social structure possess a directional quality.

II The continuum of civilization: An ordinal concept

In The Civilizing Process, Elias wants first to demonstrate the existence of long-term transformations in affective and social structures, and then to explore the relationship of these transformations to one another. Other theorists have analysed such transformations qualitatively (e.g. Foucault, 1995; Weber, 2002), and Elias agrees that qualitative transformations have taken place. He argues, however, that quantitative metaphors are needed to capture the directional quality of transformations over time.

As Liston and Mennell (2009: 59) note, this is the claim that, for many critics, ‘causes hackles to rise’. They argue, however, that Elias has been misunderstood on this point – that he is primarily analysing a qualitative shift: ‘The point is not that there is more self-constraint at later stages of development … but that it is somewhat different in its character and operation’ (p. 60). This interpretation, however, is not consistent with Elias’s own statements, which recognize the limitations of quantitative metaphors, but still insist on their importance for his analysis:

At the present stage of development we still lack linguistic instruments which do justice to the nature and direction of all these intertwining processes. It is an imprecise and provisional aid to understanding to say that the restraints imposed upon men and their drives became ‘greater’, integration ‘closer’, or interdependence ‘stronger’ … Each ‘increase’ in restraints and interdependencies is an expression that the ties between people, the way they depend on one another, are changing, and changing qualitatively. This is what is meant by differences in social structure. And with the dynamic network of dependencies into which human life is woven, the drives and behaviour of people take on a different form. That is what is meant by differences in personality structure and in social standards of conduct. The fact that such qualitative changes are sometimes, despite all the fluctuations within the movement, changes in one and the same direction over long periods, that is, continuous, directed processes rather than a random sequence, permits and indeed leads us to speak in comparative terms when discussing different phases … Here, as so often in history, we are concerned with structural changes that are most easily, visibly, but perhaps most superficially grasped in their quantitative aspect. (2000: 253–4)

Elias captures this quantitative aspect through metaphors best described as ordinal ones. He does not claim to precisely measure degrees of change using exact numbers or statistical data. He does, however, claim to engage in a comparative analysis of directional movements over time. For Elias, affective structures exhibit a development that moves toward greater self-control, more complex internal differentiation, and more highly organized expressions of drives. Social structures transform in a similarly directional way, moving toward greater integration and increasingly complex forms of interdependence due to rising levels of internal differentiation (Elias, 2000: 253–4, 450–1). Ordinal vocabulary allows Elias to highlight these directional movements.

Elias’s defenders often deploy the same ordinal language – even in the course of insisting that Elias is primarily concerned with qualitative transformations. Continuing the passage quoted above, for example, Liston and Mennell attempt to explain why Elias’s argument is fundamentally qualitative, by saying that it attempts to show how:

processes exert pressures toward increasing foresight and a change in the balance of controls; the relative weight of self-constraint over the impulsive resort to violence (among other things) increases, and becomes ‘more automatic’, ‘more even’, and ‘more all-round’. (2009: 60; emphasis added)

This is intrinsically ordinal vocabulary, implying a rankable progression. Their description of past forms of behaviour as ‘impulsive’ reinforces this impression, by implying behaviour subject, not to a different regime of control or an alternative social structuration of affect, but to lesser controls. ² That Liston and Mennell’s language undermines their expressed aim suggests the difficulty of expressing Elias’s argument in anything other than ordinal vocabulary.

Elias does not claim that the transformations he studies are strictly linear: reversals and regressions are possible (2000: 241, 253–4, 383, 438). These transformations are also not movements from a ‘zero point’ – no society possesses no level of civilization – and the level of civilization has not culminated in an intrinsic end point (pp. 52, 135). Elias claims, however, that – over the long term – a directional pattern arises. This pattern does not result from any automatic progress toward civilization, nor from intentional action. It arises instead as the unintended by-product of social interdependencies that lead social actors to behave in ways that generate unplanned, but non-random, long-term effects. These interdependencies, and the long-term trends they generate, are amenable to sociological analysis (pp. 365–8, 450–1).

Elias theorizes that this long-term, directional process is mediated by the gradual development of more centralized states out of the warrior and, then, court societies of Western Europe. In his account, more complex and extended chains of interdependence are incubated within the more pacified social environments that can be secured as courts, and then states, progressively assert a reliable territorial monopoly over both taxation and legitimate violence. Within this comparatively pacified environment, social actors become bound into longer chains of interdependence, including increasingly integrated social classes (pp. 379–82, 412–13, 421–35). Elias argues that these extended chains of interdependence complicate, multiply, and intensify the obligations of social actors to one another. In the process, they generate a heightened requirement for self-constraint (pp. 253–4, 398–9). This requirement manifests psychodynamically in the phenomenon Freud categorizes as the super-ego, whose formation is mediated by the transformation of the family into a private space of intimate relations, sharply separated from other forms of social interaction (pp. 159–60, 241, 426), in which an intensified process of childhood socialization unfolds (pp. 119–21, 158–60). In this way, Elias attempts to account for the historical genesis of psychological and family dynamics that Freud naturalizes and assumes as the starting point of his analysis (pp. 374–76, 406–10). ³

Elias mobilizes an impressive and creative array of empirical material, including extensive citations from guides to conduct, evaluations of artwork and musical forms, and summaries of the changing fortunes of specific feudal kingdoms and court societies across waves of centralization and decentralization in medieval and early modern European history (pp. 45–182, 195–362). Not all of this material, however, lends itself equally well to the ordinal metaphors Elias favours. Some of the transformations Elias describes – the development of a more complex division of labour, for example – lend themselves to ordinal comparisons without heavy interpretive work. Other transformations, however – such as which actions provoke shame or disgust – can be described using ordinal metaphors (characterized, for instance, as demonstrating a greater level of self-restraint) only if we are willing to take for granted that there is something intrinsically more difficult about conforming to the demands of certain kinds of socialization. Since The Civilizing Process associates habits that disseminate in Western Europe with those that require greater self-restraint, this dimension of Elias’s argument has attracted particular critical attention (cf. Goody, 2003). A few examples will help clarify the problem.

On the one hand, Elias insists that he is not interested in making normative judgements about ‘less civilized’ behaviour: ‘It is a different standard from our own – whether better or worse is not here the issue’ (pp. 54, 254). He also acknowledges that the concept of civilization has played an ideological role historically. He argues, however, that his own approach represents a means of ‘freeing the study of society from its bondage to social ideologies’ by tethering the analysis to demonstrable empirical trends (p. 465). Yet his ability to avoid normatively charged vocabulary wavers across his text. When describing the lives of medieval knights, for example, his imagery suggests that adult behaviours in past societies were less intensively socialized and closer to nature:

their affects had rather free and unfettered play in all the terrors and joys of life. Their time … was only very slightly subject to the continuous division and regulation imposed by dependence on others. The same applies to their drives. They were wild, cruel, prone to violent outbreaks and abandoned to the joy of the moment … There was little in the situations in which people found themselves to compel them to impose restraint upon themselves. Little in their conditioning forced them to develop what might be called a strict and stable super-ego, as a function of dependence and compulsions stemming from others transformed into self-restraints. (p. 241)

The ‘less civilized’ behaviour of the knight is conceptualized here in terms of a lower degree of social constraint – as opposed, for example, to a different structure of constraints, grounded in a distinctive set of social relations into which knights are fully socialized, which channels their affects in determinate ways. ⁴ Drives here are conceived, not simply as qualitatively shaped by socialization, but quantitatively dampened or suppressed: the knight does not just have a different way of expressing joy or anger – he expresses more of these affects; he does not just express affects differently – he expresses them more freely. Tacit here is a drive economy that bears more than a passing resemblance to the ideal of the free market that then comes to be fettered by intervention by the state – a comparison all the more apt given that Elias’s argument directly concerns the impacts of emerging state regulation on the drive economy. Yet such imagery also has historical roots in ideological notions of ‘primitive’ societies whose inhabitants are imagined as closer to nature and therefore less disciplined in their affect and behaviour than inhabitants of the European core.

Similar notions emerge in Elias’s discussions of children, whose ‘drives must be rapidly subjected to the strict control and specific moulding that gives our societies their stamp, and which developed slowly over many centuries’ (p. 119). In ‘less civilized’ societies, Elias suggests, adults are much more similar to children: ‘the control and restraint to which the drive life of adults was subjected was considerably less … as consequently was the difference in behaviour between adults and children’ (p. 119) – ‘because emotions were here expressed in a manner that in our own world is generally observed only in children, we call these expressions and forms of behaviour “childish”’ (p. 169). While Elias scare quotes ‘childish’ in order to distance himself from any potential normative judgement, his analysis makes clear that he accepts as a factual matter that such adults were not simply socialized differently, but socialized less strongly than adults in contemporary Western societies. They exhibit less control and a freer expression of drives that, while never wholly exempted from socialization, are nevertheless expressed in ways closer to their natural form (pp. 406–14). This imagery channels Eurocentric stereotypes that ‘primitive’ peoples are closer to nature and freer from social artifice (cf. Goody, 2002, 2003) – an impression repeated across many passages, which gradually accrue additional stereotypes relating to the cognitive and affective simplicity of the ‘less civilized’:

Something [emerges as typical behaviour], in the first place, that in comparison to later times might be called its simplicity, its naïveté. There are, as in all societies where the emotions are expressed more violently and directly, fewer psychological nuances and complexities in the general stock of ideas. There are friend and foe, desire and aversion, good and bad people. (Elias, 2000: 55)

By contrast, movement toward the ‘civilized’ end of the continuum is described using more positive language. The boundaries of repugnance, shame and embarrassment, for example, do not simply shift, or become more extreme over time: they ‘advance’ (pp. 99, 172, 418). ‘Civilized’ psychological structure is described as ‘more differentiated’ – enabling nuance, forethought, and psychological flexibility (p. 406) – while structures of drive control are ‘more firmly wrought’ – stronger and further removed from the natural flow of the drive economy (p. 409). ‘Western’ societies are placed at the most ‘advanced’ end of the ordinal continuum, while other societies are characterized as ‘simpler’ and ‘less complex’:

The social moulding of individuals in accordance with the civilizing process of what we now call ‘the West’ is particularly difficult. In order to be reasonably successful it requires, in correspondence with the structure of Western society, a particularly high differentiation, an especially intensive and stable regulation of drives and affects, of all the more elementary human impulses. It therefore generally takes up more time, particularly in the middle and upper classes, than the social moulding of individuals in less complex societies. (p. 378)

While Elias periodically highlights the potential downsides to the development of ‘civilized’ affective structures (pp. 375–6), his ordinal language repeatedly reactivates imagery historically associated with ideological notions of progress and Western superiority. Although he is critical of such ideologies, the question remains whether his theory provides sufficient critical distance to avoid importing dubious Eurocentric stereotypes into the heart of his analysis.

Beyond the normatively charged imagery, however, is the question of whether the very notion of a continuum of civilization can be sustained. To render the concept plausible, Elias must interpret an underlying body of empirical evidence that depicts essentially qualitative changes – changing mealtime customs, for example, or altered comportment around particular bodily functions – as demonstrating an ordinal directional movement. He needs to imply, for example, that greater levels of self-restraint are required, to select table utensils, blow one’s nose, etc., in the Western European fashion, than to comport oneself according to the norms of societies that follow different customs. He needs further to suggest that this is not a matter of a transformation in where self-restraint is directed, with one society requiring its members to attend closely to table manners, for example, while another might require a similar effort, but directed toward some other end. ⁵ Finally, he needs to imply that the story he is telling holds significance beyond a purely local, Western European context – that he is analysing a developmental advance that holds world historical significance.

As Liston and Mennell (2009: 54–5) note, Elias is of course aware that it is completely arbitrary which specific customs end up being endorsed as appropriate for ‘polite society’: the claim is not that any particular Western European custom is more difficult to perform, or that ‘civilized’ people would always endorse specific customs. Instead, Elias argues that what is distinctive is that an affective structure emerged that could take into account the sensitivities of extended chains of other persons. Transformations in Western manners merely provide evidence for the ordinal movement in the development of such affective structures. I will return to the plausibility of this claim for Western Europe in a moment. First, however, it is worth remarking that Elias generalizes this argument, which at best he evidences for Western European societies, into broad contrasts between rising levels of self-restraint in the West and the lower levels of self-restraint purportedly required in ‘simpler’ societies everywhere else. He makes this generalization without examining evidence for self-restraint in other societies, or indeed interrogating the assumption that societies prior to, or outside, Western Europe can unproblematically be regarded as ‘simple’. Moreover, he explicitly nominates colonialism as the institution that, in his time, is currently opening up ‘civilized’ affective and social structures in other parts of the world (2000: 383–6). While Elias likely sees this latter move as a rejection of racial essentialism – he presents it as a historical accident that Western Europe acquires ‘civilized’ comportments before other peoples, and argues that nothing intrinsically prevents members of any society from adopting the social and affective structures that originated in the West – it is hardly a surprise that critics would regard this line of argument as indicative of insufficient critical distance from ideological Eurocentric conceptions of world history.

Returning to the plausibility of the argument for the Western European case: Elias’s text periodically betrays the strain of attempting to organize his qualitative data into a form that can be directionally ordered. On the one hand, Elias acknowledges that certain data does not easily fit the trend, highlighting moments where customs seem to change to forms requiring less self-restraint (pp. 241, 380–3). He explains such moments away, however, by variously: characterizing them as short-term deviations from the longer-term trend (pp. 157–8); arguing that they represent a kind of exception that proves the rule – for example, a relaxation of self-restraint that would be possible only in the context of intense self-restraint by others (p. 157); or suggesting that they are ultimately driven by trends that still generate increased self-restraint overall (pp. 383–7, 426–7). ⁶ More strangely, Elias acknowledges that the forms of self-restraint he is analysing – which reflect changes in the behaviour of an elite class – involve restrictions that are incomparably less than those that material limitations impose on non-elite social groups. He asserts, however, that the psychodynamic effect of responding to material constraints, or even to constraints arising from vast differentials in social power – both of which one has no choice but to suffer – differs from the effect of complying voluntarily with the constraints of navigating complex webs of social interdependence. Only the latter, Elias argues, requires the development of a super-ego, and thus counts as self-restraint:

by and large the lower strata, the oppressed and poorer outsider groups at a given stage of development, tend to follow their drives and affects more directly and spontaneously … their conduct is less strictly regulated than that of the respective upper strata. The compulsions operating upon the lower strata are predominantly of a direct, physical kind, the threat of physical pain or annihilation by the sword, poverty or hunger. That type of pressure, however, does not induce a stable transformation of constraints through others, or ‘external’ constraints, into ‘self’-restraints. (p. 382)

This position enables Elias to dismiss a great mass of evidence that ordinary life in past societies, or among disadvantaged populations today, requires immense self-discipline simply in order to meet the ongoing challenges of survival. ⁷ By discounting such evidence, Elias ensures that the only people who could ‘advance’ in civilization are those sufficiently privileged to be comparatively well-shielded from the immediate struggle for survival.

The steady accumulation of such exceptions across the text gives the impression that Elias mobilizes a wide array of strategies to preserve an increasingly contestable central theory. It becomes unclear what sort of empirical counterexample would suffice to topple the hypothesis that transformations in affective structure can be fitted to an ordinal trend line. The theory begins to appear unfalsifiable. ⁸ This problem is to some degree obscured by Elias’s insistence that transformations of affective control structures are merely the subjective side of transformations in social structures:

The ‘civilizing’ process, seen from the aspects of standards of conduct and drive control, is the same trend which, when seen from the point of view of human relationships, appears as the process of advancing integration, increased differentiation of social functions and interdependence, and the formation of ever-larger units of integration on whose fortunes and movements the individual depends, whether he knows it or not. (p. 254)

It requires much less interpretive work, for example, to make the case that trends in the division of labour, or the length and complexity of ‘chains of interdependence’, could plausibly be described in ordinal terms over the period he analyses. The directional character of these trends, however, has no necessary bearing on the question of whether, psychodynamically, Western elites were growing more internally complex and self-restrained. Yet the ordinal character of transformations of social structure remains Elias’s strongest ‘evidence’ that transformations of affective structure also display a directional trend. By identifying the two sets of transformations, Elias begs the question that his work needs to prove.

This analytical move mirrors a major ideological rationalization for Western European superiority – one Elias himself describes:

But the idea had formed and established itself in the age of the unchallenged ascendancy of the European nations, as among all powerful and ruling groups in the world, that the power they were able to wield over other nations was the expression of an eternal mission bestowed on them by God or nature or historical destiny, the expression of a superiority over those less powerful which was founded in their very essence. (p. 464)

Implicitly, this passage suggests the possibility for Western elites to extrapolate, from the growth in their material wealth, a belief in their own moral and intellectual superiority (pp. 384–5). Such a belief could operate genuinely – but would also provide an amenable ideological rationalization for the unequal status of marginalized peoples on the periphery, as well as disenfranchised classes in the core. Elias characterizes this kind of belief as ‘ideological’, and does not regard himself as participating in it (p. 464). Yet his core argument overlaps so strongly with the ideological conception that the difference between the two becomes difficult to discern.

This overlap is troubling in its own right, but it also carries downstream ramifications. In the following section, I argue that insufficient critical distance from Eurocentric stereotypes makes it easier for Elias to overlook the ambivalent, contradictory character of the ‘civilizing process’. The consequence is that he classifies catastrophic outbursts of destructive violence that are characteristically modern, as regressions to an earlier stage, rather than expressions of the contradictory potentials of ‘civilization’.

III The ambivalent state

Both The Germans (1996) and The Civilizing Process (2000) wrestle with the relationship between ‘civilization’ and violence. In The Civilizing Process, this question is addressed most explicitly in a late discussion of the potential for interstate competition, which includes a scant few pages on colonialism. In The Germans, the question is explored through an interlocking set of studies that analyse how the Holocaust could happen in the heart of Europe. ⁹ In this section, I situate The Germans within the theoretical framework of The Civilizing Process, ¹⁰ and contrast Elias’s treatments of the Holocaust and of colonialism, in order to investigate the limits of his apparatus for theorizing state-sanctioned violence in modern societies.

Both works rely on an underlying theory of the unintended consequences of a competitive process through which larger European kingdoms defeated smaller ones, ultimately leading to the rise of absolutist states. With each stage in this process, the state gains an increasingly enforceable monopoly over the means of both taxation and legitimate violence within its territory (p. 360). Elias suggests that these monopolies have their own momentum: increased control over taxation improves the ability to monopolize violence, which in turn increases capacity to levy tax (pp. 359–60). The mutually reinforcing nature of these trends helps drive the process forward at increasing scales.

Each step of this process is associated with the lengthening of chains of interdependence, which require the navigation of increasingly complex social expectations. As courts develop into absolutist and, eventually, modern nation-states, chains of interdependence extend to encompass the rising bourgeoisie and, in time, industrial working classes. Each flattening of the class structure ensures a greater transmission of elite standards of conduct down into the lower classes, along with some transmission upwards of bourgeois and working-class standards (2000: 360–2).

The development of a strong super-ego is, for Elias, the reflection of these social processes at the level of the individual psyche. In order to navigate increasingly extended and complex chains of interdependence, he argues, the moderation, suppression, and deferment of drives is essential (pp. 369–70). The state monopoly on legitimate violence compels individuals to suppress the free expression of their own violent impulses, while the monopoly over taxation encourages disciplined future planning. In particular, the capacity to exercise self-restraint automatically becomes essential when social relations become sufficiently differentiated and complex that conscious calibration of comportment would be impossible to manage (pp. 367–8). The result is the unprecedented historical development of internally pacified spheres within the boundaries of national states, moderated principally by the self-restraint of individuals, but backed up in the final instance by the coercive power of the state (pp. 365–76).

This framework gives Elias two principal tools for making sense of ongoing violence within the context of a trend toward widespread pacification of social life. First, he can understand violence in terms of the failure or breakdown of the civilizing process, sometimes characterized as ‘decivilization’. For Elias, such a breakdown implicates a failure of both state institutions and psychodynamic structures, entailing a regression to an earlier stage characterized by lower restraint of the free expression of drives. ¹¹

Second, Elias can understand violence in terms of the ongoing competition between nation-states, in which citizens can be mobilized as agents of state violence, to serve in military conflicts that have been legitimated by the state. Such conflicts can mobilize destruction on a scale as historically unprecedented as the trend toward the internal pacification of societies. In Elias’s account, however, this destruction derives in only a limited and exceptional way from failures of individual self-restraint (1996: 309). Historically, such interstate conflicts have also provided a motor for the geographic extension of the civilizing process, producing larger states whose internal territories can then be pacified (2000: 383–5). Whether this ‘silver lining’ of internal pacification will continue, given contemporary technologies for mass destruction, is unclear, and Elias expresses considerable pessimism about potential for nuclear annihilation in some later works. ¹² In principle, however, his theoretical framework implies that the greatest hope for internal pacification lies in the consolidation of the monopoly over legitimate violence within a single, global state.

In The Germans, Elias analyses the Holocaust as an example of the first option: a breakdown in civilization. Elias argues that a wide range of German institutions undermined the stable development of super-egos capable of enabling individuals to exert strong self-restraint (1996: 3–8, 319, 401–2). Individuals were inserted into starkly hierarchical social relationships that provided strong external control over individual behaviour. Capacity for self-restraint was neither needed nor encouraged (pp. 34, 44–120, 340). While outward behaviour could appear to conform with ‘civilized’ standards, the bulk of the population did not possess the psychodynamic structures to sustain this behaviour in the absence of a very specific constellation of external constraints. Given the right circumstances – which Elias believes were provided by the Nazi regime – a dramatic decivilizing movement was poised to occur, in the midst of an otherwise modern state with vast capacities for industrial and bureaucratic organization (pp. 15, 140–2, 303–8, 319, 386–90). While the Holocaust required a scale and efficiency of destruction that only a modern state could mobilize, Elias argues that it must be understood as a breakdown of the civilizing process (cf. Burkitt, 1996: 140–2).

While Elias tends not to cite the work of other theorists, he is clearly aware of competing theories that view the Holocaust as an expression of contradictory potentials within fully modern societies, rather than as a regression to an earlier stage of development (e.g. Adorno, 1967, 1968; Arendt, 1964; Horkheimer and Adorno, 1972). Arendt (1964) in particular provides compelling empirical examples of what, in Elias’s terms, would appear to be highly controlled, ‘civilized’ forms of behaviour, applied to the systematic annihilation of the European Jews. Elias opens his analysis of Nazi Germany with a discussion of the Eichmann trial (pp. 302–4), suggesting that he may have Arendt’s work in mind. Regardless, Elias was not unfamiliar with the kind of evidence Arendt presents, and produces similar material of his own, even as he continues to interpret the Holocaust as a regression (1996: 302–3, 307–8).

Elias could, however, have found warrant to treat the Holocaust as a catastrophic example of the second type of violence available within his framework: the deliberate state mobilization of its citizenry against an ‘external’ enemy – in this case taking the form of state-sanctioned violence against an oppressed and marginalized group within the nominally ‘pacified’ boundaries of the state. Elias seems to reject this possibility because the Jews were never a ‘rational’ threat to the security of the German state (pp. 310–16), but this begs the question: if ‘irrational’ forms of state violence – those not justified by the state’s material interest in preserving its territorial monopoly over the means of violence and taxation – are labelled ‘uncivilized’ by definition, it becomes impossible to explore whether the civilizing process may generate its own distinctive forms of violence, in and through the exercise of ‘legitimate’ state power, which may include the manufacture of ‘irrational’ ends.

Here I want to suggest that Elias’s inability to consider this possibility is closely related to his problematic concept of the continuum of civilization, and the way this concept shapes his analysis of colonialism. When Elias considers colonialism, belatedly and briefly in The Civilizing Process, he seems to view it as an example of the second type of state violence – not as a breakdown of civilization or as a decivilizing outburst, but as a mechanism for the expansion of the state. He thus considers colonialism solely in relation to its function as the unwitting tool of the civilizing process, which unfolds over an ever greater geographic area, as more of the world falls under the control of strong state monopolies that can secure internal pacification within their boundaries. Here I will quote Elias at length, to enable the reader to appreciate the contrast between his immediate characterization of the Holocaust as a breakdown in civilization, and his account of colonization, which never directly discusses the forms of state-legitimated violence.

First, Elias equates the civilizing process with Western institutions and standards of conduct, and explicitly includes science and technology as examples of kinds of self-restraint:

What is happening under our eyes, what we generally call the ‘spread of civilization’ in the narrower sense, that is, the spread of our institutions and standards of conduct beyond the West, constitutes, as we have said, the last wave so far within a movement that first took place for several centuries within the West, and whose trend and characteristic patterns, including science, technology, and other manifestations of a specific type of self-restraint, established themselves here long before the concept of ‘civilization’ existed. (2000: 383–4)

He argues that colonialism is the latest wave of the process of spreading civilized forms of conduct ‘downwards’ through society, by linking the members of colonial possessions into the longer chains of interdependency that connect them to the home states of white colonizers:

From Western society – as a kind of upper class – Western ‘civilized’ patterns of conduct are today sweeping over wide areas outside the West, whether through the settlement of Occidentals or through the assimilation of the upper strata of other nations, just as models of conduct earlier spread within the West itself from this or that upper stratum, from certain courtly or commercial centres … The spread of the same patterns from the ‘white mother-countries or fatherlands’ follows the incorporation of other areas into the network of political and economic interdependencies, into the sphere of elimination struggles between and within the nations of the West. (p. 384)

This process spreads Western institutions and, as their psychodynamic reflection, Western forms of foresight and self-control:

In the areas into which the West has expanded, the social functions with which the individual must comply are increasingly changing in such a way as to induce the same constant foresight and affect-control as in the West itself. (p. 384)

While members of Western nations strive to maintain their marks of distinction, the very process of asserting their superiority over colonial populations also extends their institutions and forms of conduct through the world:

But while on the one hand these upper classes – and in some respects, as noted above, the Western nations as a whole have an upper-class function – are thus driven to maintain at all costs their special conduct and drive-control as marks of distinction, on the other their situation, together with the structure of the general movement which is carrying them along, forces them in the long run more and more to reduce these differences in standards of behaviour. The expansion of Western civilization shows this double tendency clearly enough … On the one hand they are building, through institutions and strict regulation of their own behaviour, a wall between themselves and the groups that they are colonizing and whom – by ‘right of the stronger’ – they consider their inferiors. On the other, with their social forms, they also spread their own style of conduct and institutions in these places. (p. 385)

In this way, the inequalities of social power and conduct that divide the colonizers from the colonized will gradually be overcome through the dissemination of Western institutions and comportments throughout the world:

Largely without deliberate intent, they work in a direction which sooner or later leads to a reduction in the differences of social power and of conduct between colonists and colonized … The incipient transformation of Oriental or African people in the direction of Western standards represents the last wave of the continuing civilizing movement that we are able to observe. (pp. 385–6)

Elias does not, of course, argue that colonial conquest was justified: he aims instead to assess colonialism from what he regards as a factual, neutral perspective. Yet his analysis is intrinsically normative, because he must evaluate the nature of state-sanctioned violence against colonial populations and, when doing so, he evaluates it as a carrier of the civilizing process, rather than as a regression. This move, I suggest, seems plausible to Elias precisely because the continuum notion of civilization leads him to assume that the conquered peoples lack, not only material resources or hard power, but also psychological and cognitive resources Elias ascribes to civilization. Elias thus becomes implicated in a new flavour of ‘white man’s burden’ narrative that casts conquest and occupation as being, at least in the long run, for the colonized population’s own good.

In the process, Elias also overlooks the ideological choice to treat the colonies as existing in a liminal location – residing both within and outside the boundaries of the occupying state. Elias therefore does not process state-sanctioned violence against colonial populations as occurring inside the ‘pacified’ internal territory of the monopolistic state, as he does with Nazi Germany’s treatment of the Jews. Elias thus immediately recognizes and calls out the ideological character of Nazi anti-Semitism, and harshly criticizes other theorists who attempt to find a ‘rational core’ of interests that could underlie the ideology (1996: 303–10). Interestingly, he even points out that Nazi Germany, unlike other European powers, had plans to build a colonial empire within the European core:

Nearly all European nations were drawn sooner or later into the competition for an empire. But their energies were directed, at least in recent times, towards an empire overseas. The Germans were practically the only ones who had as part of their tradition of beliefs and behaviour the image of an empire in Europe. They were bound to this tradition all the more strongly by the fact that, in the struggle for an overseas empire, they lagged so far behind on internal and external grounds. (p. 345)

Yet this observation does not lead Elias to reflect further on the nature of colonial violence. Instead, he seems to imply that the German state is particularly backward in mobilizing its desire for an empire in a way that undermines its ability to maintain internally pacified social relations within the bounds of the territorial state. This analysis leaves open the question of whether it would have been equally backward to adopt a similar ambition toward territories overseas.

Elias’s relationship to colonialism as applied to the periphery thus remains strangely quixotic: he readily calls it out as ideology and, as mentioned above, rejects any sort of racial essentialism; but he does not so much criticize this ideology as attempt to ground its major claims on a more adequate theoretical foundation. He searches, in other words, for its rational core – and, in the process, finds a way to rationalize the perception that the greater material resources and hard power possessed by the colonial powers, is associated with their cognitive or psychological advantage over the colonized.

Critics’ unease with Elias’s theory of the civilizing process is therefore justified – with consequences reaching deeply and pervasively into his theoretical apparatus. The two sides of the Janus-faced modern state – its capacity to secure a particular kind of pacification within its borders, and its potential to mobilize its population in the service of state-sanctioned, but still barbaric, forms of violence – are severed in Elias’s analysis and explained asymmetrically. The ‘desirable’ property of the state – the use of its monopoly power to prohibit illegitimate force and thus secure internal peace within its borders – is evaluated positively, and explained through the theory of the civilizing process. The ‘undesirable’ property of that same state – the use of its monopoly power to legitimize force directed to barbaric ends – is evaluated negatively, and treated as a regression to an earlier stage. Elias’s failure to thematize the processes of state-sanctioned violence against the populations of colonial possessions, I suggest, undermines his ability to integrate these two aspects of state power into his theory of ‘civilization’. It leads him into an unproductive and problematic attempt to grasp some forms state-sanctioned violence against vulnerable populations as a regression to an earlier stage or a breakdown of civilization, when these would be better seen as a fully modern expression of ‘civilization’. It prevents him from understanding the trends he groups into his overarching concept of ‘civilization’ as parts of an ambivalent and contradictory process whose violent character is often – although not always – more exposed and visible on the periphery than in the core.