The discussion of political culture in the West, implicitly or with so many words, is haunted by the specter of the transformation from democracy into tyranny, from civilization towards barbarism. Such transitions have happened before. Can it happen again, and if so, how?

At least since the First World War two views opposed one another in this debate. On the one hand, tyranny and barbarism are seen as a reversal of progress and rationalization. On the other side, they were seen as the very culmination of rationality, and modernity.

Although these oppositions are simplistic and one-sided, they are hard to transcend. In recent years, Norbert Elias and Zygmunt Bauman have written on Nazi genocide in much more subtle and varied terms, while nevertheless each siding predominantly with one of the opposing sides. This is not the place to analyze their respective positions at length, nor to compare and evaluate them. Here, the questions raised by Bauman, and by many authors that preceded him, are taken up in a discussion of civilization theory, as proposed by Norbert Elias and his students [1].

At the very core of the civilizing process, sometimes a contrary current manifests itself: While the state continues to monopolize the exercise of violence and promotes and protects the civilized modes of behaviour and expression in society, at the same time it perpetrates, massive and organized acts of extreme violence towards specific categories of its citizens. The paradigm of such a counter current in the civilizing process is Nazi-Germany, but similar phenomena have occurred elsewhere.

Elias himself and a number of his students have presented and clarified their sociological concept of ‘civilization’ on numerous occasions, and if nevertheless it remains hard to grasp this is not just due to a lack of clarity in the argument, or to a scarcity of empirical referents, but most of all to the complexity and subtlety of the concept itself. Elias opted for a multi-dimensional and highly intricate definition that evolved over more than half a century as his publications succeeded one another (Goudsblom, 1994).

In later years Elias came to prefer the plural ‘civilizing processes’ to denote the multi-tiered development that he had observed in Western Europe. He adopted Cas Wouters’s expression ‘informalization’ to convey the idea that a civilizing process might evolve towards less rigid, that is more varied, subtle and flexible modes of interaction as he himself had already suggested in his ‘Project for a theory of the civilizing process’ [2]. He wrote at length about the precarious course of the civilizing process among the Germans, and even entitled a chapter in that book ‘Der Zusammenbruch der Zivilisation’ or ‘the breakdown of civilization’ [3].

In recent years a number of Elias’ students have taken up this thread where Elias left it and written about ‘decivilization’ and ‘decivilizing processes’, (e.g. Fletcher, 1997; Goudsblom, 1994; Mennell, 1990; Szakolczai, 1997; Van Krieken, 1999; Wacquant, 1999; Zwaan, 1996).
Both expressions, ‘decivilization’ and ‘breakdown of civilization’ refer to constellations of wide-spread and violent destruction that succeed earlier periods when civilization prevailed to a greater degree, with more restrained modes of interaction and more tempered self-constraints. The very terms suggest that something that once existed has since disappeared, that it was lost or destroyed. This sense of loss and decay is vividly evoked by such expressions as: ‘regression to barbarism’, ‘vulnerability of civilization’, ‘breakdown’, ‘decay’ (explicitly versus ‘growth’) and, ‘the open relapse of the National Socialists into barbarism’. All these terms have been taken from a single page (308) in the essay in which Elias directly confronts the extermination of the Jews in World War II [4].

In this study, Elias immediately sets out to argue that ‘civilization’ is not a permanent state but rather a precarious process, that may very well reverse itself. ‘How was it possible,’ he asks, ‘that people could plan and execute in a rational, indeed scientific way, an undertaking which appears to be a throwback to the barbarism and savagery of earlier times…’? (Elias, 1996: 302). Elias concludes that no ‘raison d’état’, no war-objectives, no goals of internal politics were served by the murder of the Jews, rather on the contrary. And although many profited from the crime, these material gains can hardly explain the enormity of the massacre. In other words, it was a deeply irrational enterprise, that can only be explained in terms of the nazi-ideology itself.

But in the same context, Elias (1996: 307) expressly mentions another aspect: ‘the killings in the gas chambers.’ And he comments: ‘Compared with pogroms, and with military procedures, this new form of extermination meant an advance of rationalization and bureaucratization.’ And no doubt, many of the preceding stages in the extermination of the Jews, their registration, concentration, deportation, exploitation, proceeded in a thoroughly planned, systematic, bureaucratic manner.

At the heart of Elias’ thinking is the twofold movement of rationalization and bureaucratization on the one hand, and regression, breakdown, increasing barbarism on the other. Most accounts of the nazi-genocide and other episodes of mass extermination, proceed in terms of either the one or the other perspective, either rationality, bureaucracy, and modernity or barbarism, regression, breakdown. But the main momentum of Elias’ theoretical work veers towards an interpretation of the extermination of the Jews in terms of a ‘breakdown of civilization’. Thus, Elias has stressed that the German state was a weak state that failed at the task of pacifying and civilizing the Germans and therefore allowed a reversal to barbarism to occur [5].

 A clear example of the contrary approach, one that considers the extermination of the Jews, and genocide in general, as the very core of modernity is provided by the writings of Zygmunt Bauman, in his Modernity and the Holocaust, and even more explicitly in his Postmodern ethics: ‘The modern era has been founded on genocide, and has proceeded through more genocide’.

One must grant Elias and Bauman in their better moments, that they discern both aspects concurrently in National Socialism: order and barbarism, design and impulse, organization and wildness. Framed in these opposing terms, the discussion dates back to the aftermath of the First World War when after a century of relative peace and widespread faith in progress the mutual mass destruction of trench warfare had to be accounted for somehow [6].

In a brief and lucid account, Arpád Szakolczai (1997) addresses precisely this issue. At the first level of explanation, he argues, impulsive behaviour can be understood as a relaxation, an escape, one might say, a Ventilsitte, among civilized persons in a civilized society. On the second level there are historical ‘inflection points’ where impulses and tendencies that before had to be warded off now become acceptable and are even cultivated: e.g. the profit motive (Weber) or sexuality (Foucault), or - Szakolczai’s own example - bellicosity during the crusades. And Szakolczai (1997; italics added) continues:

" There is, however, an even more important third level of explanation. This concerns the conditions under which the civilising process can turn against itself, where the question is no longer simply a paradoxical compromise between the civilising process and its opposite, the impulses set loose by a previous dissolution of order, but where the fundamental mechanisms of the civilising process are effectively, purposefully and explicitly undermined. It is at that level that the totalitarian movements of the twentieth century can be located, with the important caveat that they are very closely related to the previously mentioned inflections of the civilising process, therefore they cannot be fully externalised and exorcised, restricted to the cases of Nazism and Bolshevism"". 

It appears that Szakolczai is on the verge of transcending the opposition between modernity and barbarism, and is ready to identify the dialectics between them. In fact, the civilizing process may indeed be ‘undermined’, or ‘inflected’. The assumption in Elias’ theory of civilization is that state formation, i.e. the monopolization of violence (and taxation) will lead to more civilized modes of intercourse and expression, i.e. a lessening of all forms of violent behaviour, state violence included. And implicitly it is assumed that the state will treat all law abiding citizens more or less equally, i.e. that there will be some measure of equality before the law. But this need not occur.

The monopolization of violence by the state may result in the overall civilization of society and yet, in certain cases, these civilized canons may nevertheless exclude certain categories of citizens from protection who will then be exposed to all the violent resources of the state monopoly. The regime may mobilize the entire machinery of the state to persecute and annihilate this target group, and this more thoroughly than could have been achieved in societies where the state apparatus has not succeeded in monopolizing the means of violence so effectively. In the process of bringing about this destruction, the intended victims must first be identified, they must be registered, they must be isolated and made the object of a persistent campaign of vilification and dehumanization; hatred and loathing must be evoked against them among the population at large. This is what I have called elsewhere the social work of ‘disidentification’, which goes together with a campaign to strengthen positive identifications among the rest of the population. (de Swaan, 1997). In the next phase special units must be recruited and trained to round up, isolate, and destroy the target population, and for this task specific locations must be screened off from the uninitiated so that the torturing and killing may proceed unnoticed (but not unbeknownst to them) in reservations of destruction. Thus, both in a psychological, a social and a spatial sense, this process occurs as one of compartmentalization.

All the while, the rest of society maintains its pacified ways, and the vast majority of citizens continues to be protected by law, custom and etiquette. Just as it would not occur to the butcher to use his knife outside his shop or on anything but animal flesh, the guards and henchmen would not dream of attacking anyone beyond the designated category, or brutalizing their victims outside the spaces marked off for the purpose. Obviously, what occurs under these conditions is the bureaucratization of barbarism. The most barbarous acts are perpetrated, sometimes in a calculated and detached manner, sometimes wildly, with passion, lust and abandon. What matters is that the barbarism occurs in demarcated spaces, in delineated episodes, well separated from the rest of society, from the everyday existence of the other citizens. The barbarity is compartmentalized. This compartmentalization refers at once to the categorization of a target population, the physical isolation of the sites of destruction, the institutional identification of the authorized agents, the censoring of all information and opinion on the subject, the social demarcation of brutalization from other forms of interaction, and for the perpetrators the psychological separation of their psychic experiences from all other mental processes or social encounters. Mark Danner (1997:59) quotes observers of the ‘Bosnia Genocide’: 

"Western and his colleagues were struck not only by the cruelty of these abuses but by their systematic nature; they very rapidly came to understand that though the Serb soldiers and, especially, the “paramilitary” troops responsible for “mopping up” were committing wildly sadistic acts of brutality under the influence of alcohol, their officers were making rational, systematic use of terror as a method of war. Rather than being a regrettable but unavoidable concomitant of combat, rapes and mass executions and mutilations here served as an essential part of it…" 

Here, the wildness and brutality are let loose, or maybe even instilled, and at the same time instrumentalized, for specific purposes, within demarcated spaces at an appointed time: an archipelago of enclaves where cruelty reigns while being reined in all the while.

The term ‘compartmentalization’ refers to a ‘ defence mechanism’, in this case one that operates through the strict isolation (A. Freud, 1966) of certain, especially problematic emotions and impressions. But the notion (like e.g. ‘repression’] immediately evokes social correlates, at every level of social life. Both at the personal and the group level this compartmentalization proceeds through disidentification from the designated victim population, the withdrawal of identificatory affect, the denial that the target population might be similar to oneself and the repression of emotions that result from identification, such as sympathy, pity, concern, jealousy etc. (De Swaan, 1997).

Under these conditions of state monopolized violence, a high level of civilization is maintained in almost all respects and for the vast majority of the population; however, the regime creates and maintains compartments of destruction and barbarism, in meticulous isolation, almost invisible and well-nigh unmentionable. It is as if the civilizing process continues with the same means, but with a different turn: in one word, it has become a dyscivilizing process.

'Within the confines of these compartments the civilizing process has been suspended; under carefully controlled conditions decivilization is allowed to proceed, barbarism is expressly provoked and unleashed against the target population that has been exempt from all state protection. If decivilisation may be described at the psychological and social level as ‘regression’(into a prior, a more primitive, more disorganized stage) then this dyscivilization may be described in terms of ‘regression in the service of the state’ [7]. 

Civilization has not broken down, the social order has not fallen to pieces, barbarism has not spread all over, decivilization has occurred only in well defined episodes and spaces. What takes place is dyscivilization: the totalitarian state continues to function in a bureaucratic, planned, ‘modern’ and even ‘rational’ manner. The ruling elites have mobilized barbarism for their own purposes and carefully encapsulated it into special compartments of local decivilization, where even wild destructiveness has been made instrumental, functional in the regime's campaign against its designated enemies.

Compartmentalization is the social arrangement and the psychic defence mechanism par excellence in a dyscivilizing society. To maintain it requires both rigid separations and carefully staged passages between the different emotional and interactional domains. As a consequence, the transition to a more flexible, more varied repertoire of relational and emotional modes, as Elias observed it in the contemporary civilizing process, cannot occur under conditions of dyscivilization. What Cas Wouters (1986) has described a process of 'informalization' that is much akin to what I have characterized elsewhere (De Swaan, 1990) as a 'shift from relational and emotional management through command to a management through negotiation.' Such a transition is incompatible with the defence mechanism of compartmentalization. Dyscivilizing societies will develop quite strong, but also quite rigid types of social control and self-control. Very elaborate codes of conduct and expression will be maintained to the smallest detail, until the moment that one steps over the threshold and into the compartment of barbarity, where all cruelty and wildness are permitted, until one leaves this reservation again and resumes one’s controlled demeanour, as if nothing had ever happened: that is dyscivilized behaviour.

The student of civilizing and dyscivilizing processes will be especially interested in these transitions, these recurrent ‘rites de passage’ from ‘civil’ to ‘dyscivil’ conduct and experience: How after a day’s work the guard gets ready to leave and go home (washes up, changes clothes, forgets all about it, remains silent about it all at home, denies everything, lies about it, or recounts the day’s events in vivid, lurid detail). Is there a precise schedule and calendar or do personnel simply slip in and out of their roles in haphazard, irregular fashion? Are the venues hidden, inaccessible, isolated by deserts, woods, screened off by walls and fences, or rather visible to passers by, who may even enter and watch at will? How do the guards, the torturers, the militiamen think about themselves? We almost always get to know them in a defensive stance, forced to speak in front of their judges, little do we know about them when they were in full action, on the offensive and may have had to prove precisely the opposite: their zeal and zest and gusto, their loyalty and commitment to the task. But again, how during one phase do they think of themselves in the other phase: are they ‘a different person’, ‘do they turn off all emotions’, ‘try not to think’ or are they proud and pleased with themselves in their other capacity? These are all questions about the nature of personal and social compartmentalization. 

The modus operandi of compartmentalization need not be so extreme, it may occur under comparatively innocuous conditions. Thus, in contemporary consumer societies, butchery is equally relegated to special compartments: not only abattoirs, but even pig and chicken farms are hidden from the public’s view and once out of sight are effectively out of mind. Somehow, when enjoying their meat, consumers manage to forget that they are actually eating an animal and to ignore the way it was raised and killed, even though they know these facts very well.

In most societies prostitution is effectively shielded off from the rest of social life: there are spatial enclosures, ‘zones of tolerance’, ‘red light districts’, ‘closed houses’, there are temporal separations (‘darkness’, ‘girls of the night’), and both the prostitutes and their clients usually succeed in slipping in and out of these prostitutional reserves without being noticed. Similar observations can be made about prisons, insanity wards and Michel Foucault’s other favorite haunts [8].

The spatial isolation and social exclusion of a designated category of people was taken a momentous step further in the ‘ghettoisation’ of American inner cities, as Loïc Wacquant (1993, 1999) has described it. What adds much interest to his detailed account is Wacquant’s explicit analysis in terms of ‘decivilisation’: as the state withdraws from the inner city areas, chains of interdependence break down, self-restraints disintegrate, ‘depacification’ proceeds as violence proliferates without the police intervening anymore, social differentiation is reversed as only informal economic activities remain, and so on. Islands of ‘decivilization’ have emerged in the very midst of a relatively civilized society without very much affecting it in its entirety. Again, it is effective compartmentalization that maintains this precarious separation of ‘civilized’ and ‘decivilized’ spheres. Wacquant stresses the necessary disidentification that keeps the ‘underclass’ as a separate category outside the bounds of normal citizenship. Outside these ‘ghettos’ life proceeds ‘as usual’ [9]. Wacquant is especially interested in the decivilizing process that occurs within the inner city ghetto’s. What matters here is how these pockets of decivilisation are effectively shielded off from surrounding society, warded off from consciousness, exempted from affective or moral identification. No doubt, the onset of a dyscivilizing process already exists. From almost lethal neglect to actual extermination would howeever require many momentous further steps [10]. 

At the core of Elias’ ideas on the civilizing process is an implicit assumption of minimal equality, of some measure of equal treatment and equal esteem. Such a modicum of equality means that people identify with all others in their society as beings that are more or less the same as they themselves are [11]. It implies, moreover, a degree of equality before the law, and even some equalization in living standards. When one category of people is completely excluded from this minimal equality, the civilizing process may take a different turn and proceed along a different track. It takes a radical and annihilationist regime to complete the shift in the direction of a dyscivilizing process.

Subsequent stages of compartmentalization are increasingly incompatible with a free press, or with legal guarantees such as freedom of movement or freedom of speech - all of which by their very nature tend to transgress, to transcend the very borders that are essential to maintain compartments. Unless, of course, the target population is exempted from these rights and there is a consensus among all others to ignore whatever is done to it (somewhat like the situation that prevailed in the slaveholding society of the antebellum South of the US or in the early twentieth century on the plantations of the Netherlands Indies under Dutch colonial rule).

Even a rather generous welfare state could exist in a dyscivilizing society, pried loose from its universal- egalitarian foundations, if only the targeted victims are excluded from its benefits (somewhat like the welfare policies that prevailed in Nazi-Germany). 

In terms of Elias’s theory of civilization, what are considered the twin processes of the monopolization of the means of violence and the overall civilization of society have in the present argument been pried apart: even when monopolization of violence prevails, the overall civilization of society may or may not occur. In the first case we have the ‘normal’ civilizing process in the sense of Elias’s theory. But in the second case, the state has achieved the monopolization of the resources of violent coercion, and yet civilized relations do not prevail in society in its entirety. Certain spaces, certain groups are excluded and have become the target of the full destructive apparatus of the state: this second trajectory I have termed the ‘dyscivilizing’ process.

Once this separation of the monopolization of violence on the one hand and the civilization of society on the other has been conceptualized, two more possibilities follow from the theoretical construct. Both refer to conditions of an incomplete or receding monopolization of violence by the state. First, as concerns the incomplete monopolization of the means of violence: since no monopoly of violence has yet been brought about, less civilized behaviour is to be expected in terms of the theory., When an effective monopoly of violence has been established at some point, but has begun to disintegrate since then, the theory suggests that human relations and modes of expression will regress and that a process of ‘decivilization’ will occur throughout society.

Civilizing theory as it stands excludes another possibility: that the monopolization of violence has not yet been accomplished or that it has broken down again and that nevertheless civilized modes of behaviour and expression are prevalent throughout society. In terms of civilization theory, this would be an ‘abnormal’ state of affairs. And yet, this is clearly the situation that prevails among a Maroon community of Surinam, according to Bonno Thoden van Velzen (1982) who ascribes a high degree of self-control to the Djuka, without a monopoly of violence having been achieved in their society.

A more rigorous empirical and historical discussion of these matters would have to establish the degree of monopolization of violence in different societies [12] on the one hand and on the other hand assess the degree and distrubution of civilized modes of behaviour and experience within these societies. Apparently, relatively civilized modes of conduct and expression may exist in the absence of effective monopolization of violence, and conversely when a strong monopoly of violence has been achieved, more civilized forms of interaction and experience may spread, or society may follow a different trajectory: while overall a degree of civilization prevails, the full violence of the state is unleashed against specific categories in well-demarcated local, temporal, social and mental compartments - the trajectory of dyscivilization.

 Abram de Swaan, Theory, Culture and Society 18.2-3, April-June 2001.