This article argues that violence is primarily the outcome of 'securitisation' of the fusion of the concepts of state/regime, nation and Islam. Whenever this fusion is contested by societal groups, the state/regime fears a complete break up of the hitherto 'deep' structure of the fusion of the state/regime, nation and Islam. This in turn leads to quick 'securitisation' and violence and to societal claims for knowing who kills who.

Representations of the Reasons for Violence

The European press represented until 1997 violence in Algeria as blind, irrational, unpredictable, evil, medieval. Just as great parts of the European press did as for the Bosnian war. A dichotomy between the rational, secular European individual and the wild Algerian Islamist was thus constructed. This was a complete mirror image of the Algerian regime's representation of the reasons for violence. The regime represented violence as a struggle between the rational state order and Islamic chaos and anarchy. The Islamists turned upside down this representation by ascribing the reason for anarchy to the state/regime. The discourse of the regime and of the Islamists thus excluded each other but at the same time they shared the same basic codes of order and anarchy.

In France, plenty of books about the reasons for violence in Algeria have been published since the outbreak of violence in 1992. The explanations most often run along the following lines: either the Islamists are to blame, or the state-regime, or finally both Islamists and state-regime (Mahiou, Henry, 2001: 15). The different interpretations result in different labelling of what is going on. The denominations range from state or Islamist sponsored terrorism, residual terrorism consisting of small pockets of Islamists terrorists, 'second war' with reference to the 'first' war against France, 1955-1962 (Provost, 1996), 'invisible, veiled' war with reference to a veiled war like the veiled women, hidden Islamist guerrillas and disguised security forces (Stora, 2001), anarchy because nobody knows who kills who, war against the civilians where state/regime wages large-scale war against all kind of opposition, and finally civil war where faction of society wage war against each other and the state/regime and vice versa (Martinez, 1998, 2000).

The above-mentioned semantic struggle demonstrates the competitive attempt at attaching meaning to what is going on in Algeria. Since president Bouteflika came into power 1999 some concepts/denominations have been marginalized because violence has been somewhat reduced. The marginalization goes for the concept of civil war. Instead, the concepts of crisis, violence, massacres and terrorism (Meynier, 2000; Mohsen-Finan, 2001) have become dominant. Violence and massacre are used by the Algerian newspapers and electronic medias (1) as an inherent feature of societal and state/regime behaviour. The word 'terrorism' was used from 1992 to 1994 by the state/regime. Since then it became a question of eradication of 'residual terrorism'. De Gaulle also used this term when he came into power in 1958. He used the word to convince the French that the independence war was about to be won just like the Algerian regime uses the word to convince the population that the worst has been overcome. The concept of civil war has ever been banned from official language because it signals an acknowledgement of lack of legitimacy.

The state/regime has represented the perceived threats against the survival of state/regime as Islamist terrorism. The Islamists but also the Berbers use the concept of terrorism with reference to state/regime terrorist behaviour towards society. Berbers, Islamists and state/regime thus try to 'securitise' (see below) either society or state/regime thus excluding the right of the one or the other to express political critics. All parts posit the concept of terrorism in a discourse about the need of purification either of society or of the state/regime. Hence, survival of society in the radical Islamist discourse is linked up to complete change of regime/state. Survival of the state/regime is bound - in the official discourse - to eradication of what is considered an Islamist virus in an otherwise sound societal body. Therefore, in the wake of September 11, Algeria shared with Washington a list of 350 Algerians abroad with alleged links to Osama bin Laden, and a list of alleged Islamists militants inside Algeria. (2)

Fusion of State and Regime

The independence war against France made the Algerian state. It was recognized by other states as a state, and the FLN, represented the independence as the birthday of the state-nation.But the important question is what kind of a state? Is it a successfully imported European/French state (Badie, 1992) or is it a failed state (Snow, 1996)? Donald M. Snow argues that in the post-Cold War world the result of the combination of weakening of coercive state authority and of little social cohesion in the so-called 'Third World' can be a 'failed state', unable to control violence in its territory (Snow, 1996: 58). In practice, it is neither of these, rather it is a state that is still in a process of state building i.e. 'which signifies a conscious effort at creating an apparatus of control' (Hibou, 1993: 93, Martinez, 2000: 9). Although the state monopoly of legitimate use of violence has been seriously challenged ever since 1992, the functional machinery of the state has survived the continuous violence.

Until 1988, when army and security forces killed about 500 demonstrators, it looked as if the state was strong. Internationally, the state appeared strong in the sense that it was recognised by other states as an important international player. Domestically the state appeared strong, because the state succeeded in eliminating all oppositions. But the strength of a state neither depends on, nor correlates with, power (Buzan, 1983: 66). A strong state is characterized by possessing a high level of political internal consensus centred around the idea of the state.

Since 1988, the domestic support for the idea of the state as a distributor of economic welfare and as a social entrepreneur broke down in the wake of the oil crisis in 1983-84 and because of the killings of demonstrators in 1988, who protested against the economic policy of the state and the lack of political transparency. The increasing impoverishment of the population resulted in a profound distrust of the state, the functional state machinery and the regime. The FLN-state (3) was severely shaken because the economic and social crisis laid bare the de facto fusion of state and the particularistic interests of the regime. Thus conceptually, the statet rather has to be characterized as a neo-patrimonial state being privatized by various political groups who consider the state as their instrument with regard to economic accumulation and political power. The state is thus seen as the property of the regime. Thereby state and regime is fused. You cannot refer to the former without referring to the latter (see below).

At the theoretical level, it has been argued that any appropriation of a neutral impersonal state apparatus by particularistic interests can encourage the quick resort to violence, with the use of force becoming a central rather than a peripheral feature of domestic political life (Buzan, opus.cit: 60). In such instances, the raison de régime prevails over the raison d'État, and the praetorization of the army is its natural consequence (Salamé, 1994: 23). Whenever society protests against the policies of the regime, the latter is prone to perceive the protest or critics as a threat to the survival, not only of the regime, but also the state. Any political change at the regime/governmental level is perceived as entailing the risk of the complete breakdown of the state. The regime therefore tends to cling to the armed status quo that in turn results in further erosion of legitimacy, bringing about further protests. The outcome of such a process is a weak state/regime that permanently considers itself besieged by opposition of any kind, which at any cost has to be broken down. This obsession with state-regime security has been clear in Algeria but is by no means unique to it. All third world nation-states are concerned with state-regime security (Ayoob, 1995: 191). However, in contrast to many African states, the Algerian state/regime has succeeded in containing violence to an extent that does not break up the state-institutions. But the state-regime has not regained the monopoly of legitimized use of violence. Instead, a continuous privatization of violence is going on.

If one follows the argumentation of the sociologist Luiz Martinez (Martinez, 1998, 2000), the main reason for this privatization is due to the fact that social progress through violence is the only way in Algeria to bypass the economic and political power of the neo-patrimonial state/regime. Still according to Martinez, the social advancement through violence is linked to what he calls 'a war-oriented' imaginaire which he links up to the permanent character of the corsair, the cadi and the colonel whose wealth and prestige were derived directly from their participation in the wars of conquest and liberation. Therefore he suggests, that the actual 'emirs' (note, lamchichi), the Islamic warlords or bandits, far from constituting a breach with the past or even a revolution in Algeria, form part of the image of war since the Ottoman Empire as the specific way of getting access to wealth. Martinez is well aware of being accused for being a culturalist by underlining the notion of war-oriented imaginaire(4), and therefore he underlines the necessity of understanding culture as processes that give rise to more or less strong, more or less widely shared and more or less stable figures (Martinez, 2000: 9). His partly essentialist way of analyzing is thus balanced by a more constructivist stance which opens for interaction between structure (war imaginare) and agents (emirs) which explains the dynamic of the emergence of another, alternative system of profit and power.

It is not the goal of this article to investigate the social field in order to analyze violence, as Martinez has done, but to use the theory of 'securitisation' in order to analyze the process of violence linked to the 'securitisation' of the fusion of state-regime, nation and Islam.

Securitisation of the Fusion of State/Regime

The violence that Algeria experiences, is to a large extent triggered by the fusion of the state and regime with its attendant 'securitisation' of this fusion. One cannot speak of the regime without speaking about the state. The two concepts are inherently linked to each other. If the concept of state is disconnected from the concept of regime, the regime will try to 'securitise' both state and regime in order to survive as incarnation of the state. The regime has ever since the cancellation of the elections January 1992 feared the disintegration of the state and consequently its proper loss of power. Any opposition to the regime has been stated in terms of threats to the state and thereby to the regime. The means of the regime to counter the perceived threat has been imposition of a state of emergency ever since 1992. This state of emergency is what characterizes a full-scale 'securitisation' of state/regime.

The theory of 'securitisation' - elaborated by the so-called Copenhagen School - operates with 'security' being understood as a speech act through which a condition of insecurity is identified, threats are pointed out, and an object of security is constructed. It is thus only from the moment when somebody (a securitising actor) claims that something or somebody (the referent object) which has an inherent right to survive is existentially threatened that an issue becomes a question of security concern (Wæver, 1995: 54). In order to cope with the exigencies of the situation 'securitisation' requires extreme measures that bypass the 'normal' rules of the political process. However, an instance of 'securitisation' is not met simply by situations where the breaking of the normal political rules occurs, or by the simple identification of existential threats. It also requires acceptance by a sufficiently large 'audience' that will tolerate violations of rules that would otherwise have to be obeyed (Buzan, Wæver, Wilde, 1998: 6, 24).

The theory underlines the importance of distinguishing between perceived threats to society (nation) (5)and state because the ultimate criterion for societal security is identity (see: Securitisation of the concept of nation), whereas state security has sovereignty as its ultimate criterion. Both usages imply survival. A state that loses its sovereignty does not survive as a state. A society (nation) that loses its identity fears that it will no longer be able to live as itself (Wæver, 1995: 67).

The regime has ever since the independence in 1962 felt besieged by what it considers traitors to the unified and centralised state/regime. Algeria's experience of politics under the 'normal rules' of the game has been limited to the period 1989-1992. A new constitution (1989) established a pluralist democracy that allowed for participation of multiple parties. In 1992, however, the Constitution was annulled. FIS was banned and extreme measures were taken in order to securitise the state/regime: a twelve-month a state of emergency was imposed. In 1993 the state of emergency was extended indefinitely and in October 1992 an 'anti-terrorist' decree was passed. In 1995 Algeria witnessed a horrifying escalation of murders and bombings. At the judicial level, the regime responded by incorporating the anti-terrorist decree into permanent legislation.

The most recent constitution of 1996 permits a certain political pluralism. All parties that do not refer to a religious, regional, or ethnic affiliation in the party name have the right to present themselves at all state and local elections. In reality, religious parties considered moderate by the regime and the Berber parties have participated in the various government since 1996. As long as they remain loyal to the regime and thereby to the army, the regime connives at the cultural and religious affiliation of the parties. In addition, they are considered a kind of bulwark against radical Berbers and Islamists. But because the regime instrumentalizes especially the moderate Islamist parties these lost ground at the Parliamentary election May 2002 and at the local election fall 2002.the most recent local elections October 2002. The conservative Islamic party, El Islah (reform party) won over another Islamic party, MSP (Mouvement de la société en paix) (6) which had been in government since 1997 and therefore discredited. But whenever the regime perceive of a threat to the concept of a unified territory it can ban the party that openly goes for autonomy of a region, like the Kabyle. Fear of fragmentation of the territory thus results in a process of securitisation of the unity of the territory which is an inherent part of the concept of the fusion of regime and state..

The theory of 'securitisation' implies that the actor has sufficient legitimacy for receiving support to the enunciation on something being threatened. But in Algeria it is difficult to 'measure' an actor's legitimacy because opposition is more or less suppressed. It is therefore not easy to analyze how successful 'securitisation' is. It is difficult, for example, to 'measure' the public support to 'securitisation' of the survival of the state/regime because elections have never been free or fair - with the possible exception of the election of President Zeroual in 1995. But an indicator of an actor's lack of legitimacy is for example riots, demonstrations and lack of participation in elections. Still more riots and protests are going on in present Algeria. The revolts are triggered off by the huge demonstrations in Kabyle spring 2001, which resulted in the killing of about 100 demonstrators and about 2000 were severely wounded by the security forces. Bouteflika tried to downgrade the importance of this so-called 'black spring', by transforming the goal of the demonstrations to a question of specific Berber culture that had nothing to do with the Arab Algeria. But this divide and rule policy did not succeed. Socially and politically inspired riots have spread to the whole territory thus destabilizing the regime. The result of this wide spread riots has been a further opening-up for societal demands. The Berbers are still more contesting the regime's suppression of specific Berber culture. This makes a researcher writing about the possibility of a future Berber claim for either autonomy, or a federal state (Kaki: 2003). These two models are a complete heresy to the hitherto construction of the centralized and unified Algerian nation-state (see below). Any such demands will certainly be met by enforcement of the state of emergency and thereby by further 'securitisation'.

The theory implies also that the actor utters that something is threatened. If somebody with sufficient authority declares that something is threatened, it is threatened. The problem in the case of Algeria is that the most important actor is the army. It is the real power container. But is very seldom that the army utters anything at the political scene. It remains the hidden puppeteer pulling the political strings. Hence, it is very difficult to judge whether it is for example the president or the army who speaks, when the president utters that the state is threatened. Therefore the air is always thick with rumours of possible coups d'état and of internal fights in the army and the security forces. These rumours 'pacify' to some extent the population because of the implicit references to the danger of collapse of the state. Thus these rumours ease the process of securitisation.

However, during the last two years still more officers criticize in newspapers the way the army functions and its perceived role as a hidden political actor. These critics have been obvious since the publication of a very controversial book La sale guerre, written by a former Algerian officer (Souaidia, 2001). He describes the absolutely cruel methods of the army and security forces with regard to suspects. This book caused of course an outcry of anger amongst Algerian politicians and high-ranked officers. But there was nothing to do. The book was there and the discussion continues. This does not mean that the army has lost its hidden political power, but it does mean that great parts of the population still more puts into question the legitimacy of the army and thereby also the regime.

Discussions of the all kind of issues have to a certain extent been allowed in Algerian newspapers since 1996 (Mahiou, 2001: 27). But on May 16, 2001, the lower house of parliament - the National People's Assembly - approved new amendments to the Algerian Penal Code that prescribe imprisonment and heavy fines for individuals found guilty of defaming the president of the Republic or state institutions such as the army or judiciary. The amendments were legitimized as necessary to 'preserve the dignity of the state and to protect individual and collective freedoms' (7). In the context of still increasing revolts the word ' dignity of the state' is identical to 'security of the state'. It is not bluntly declared that the state is perceived as threatened by the riots and critics. But the state is represented as a person who will sink into a state of shame if the 'children' do not respect their 'father'. Obedience is what matter in a family and the father might use further extraordinary means if the children do not obey.

When Bouteflika came into power in 1999 he promised judicial reforms. He appointed a commission whose task was to clarify the function of the whole judicial system and the lack of implementation of individual rights. The report of the commission was published in 2000. It recommended the government to put the security forces and the army under political control in order to ensure strict observance of the individual rights. The conclusions of the report did not result in any laws that could have limited the power of the army and the security forces. In the wake of the riots in Kabyle in spring 2001, yet another commission was appointed whose purpose was to investigate into the reasons of the escalation of violence. The report was published in July 2001. It singled out the gendarmerie, the army and the security forces as the instances that killed the demonstrators. Even specific persons were mentioned in the report. (8) But once more, a report had no political or judicial consequences because this would have been a heavy accusation against the army that considers itself as the guardian of justice. Hence the definition of who is a criminal or a terrorist continues to rest with the army and the security forces.

Although Bouteflika tried in the beginning of his electoral period to 'desecuritise' the state and the regime by allowing for political critics, he has not been able to de-link the fusion of the two concepts because of the power of the army and the security forces. This in turn had made 'desecuritisation' extremely difficult. Furthermore the securitisation of state/regime with the attendant exclusion of alternative voices has led to societal violence.

Securitisation of the Concept of Nation

"A society (nation) that loses its identity fears that it will no longer be able to live as itself " (Wæver, 1995: 67). An important problem of this statement is that society/nation is regarded as a whole, as constituting one identity. The Copenhagen School has therefore been criticized for fixation of the notion of 'identity' and 'society/nation' (McSweeny, 1996: 85, Albert, 1998: 24-25). In line with Mathias Albert, I will argue that one has to respect the need for boundaries and categories, yet recognize "that any entity can be situated in more than one mental context" (Albert, 1998: 24). This interaction between fluidity and fixation opens up for an understanding of the various processes of securitization that take place in society/nation which are constituted by a multiplicity of identities. In a state like Algeria, the political elite's construction of the idea of the nation is highly contested by different societal factions. In the beginning of the 1980s the Berbers revolted against the suppression of their culture and the Islamists started the armed attack against the state/regime in order mobilize for the foundation of a Sharia-state (Meynier, 2000: 208) (see below).

The 'violisation' of identities (Neuman, 1998: 19) (9) became so intense because the speech act on security of both parts was stated in mutually excluding conceptualizations of the construction of the idea of the fusion of state/regime and nation. As Jabri puts it: "Articulations of identity in circumstances of conflict draw upon deeply embedded identitional continuities mobilised in the construction of bounded political grouping. The second element deriving from a conceptualisation of conflict as constructed discourse is that violent conflict is constituted around the construction of a discourse of exclusion" (Jabri, 130).

State and regime have fused. But the concept of nation has also fused with the state/regime. (10) The role of FLN in this has been vital. Since its foundation in 1954, until its dismissal as the state/regime party in 1989, the FLN constructed the narrative of national identity around the fusion of the state/regime and the nation. FLN represented itself as the incarnation of the unified nation whose survival had to be guaranteed by an all-compassing state represented by the FLN-regime (Lamchichi: 1991). All kinds of ethnic, religious and political opposition were excluded from the political arena. Only a representation of an undifferentiated and harmonious territorial nation-state was allowed at the political arena. Opposition was represented as treasonous and the work of enemies. The concept of an all-compassing unified state-regime-nation thus resulted in clear-cut borders between inclusion and exclusion thereby creating the 'radical Other' (Leveau). This in turn lead to violisation, i.e. violence on a massive scale.

Logically and as a result of this fusion, whenever discourses on de-linking one of the elements of this fusion become manifest in riots in the streets or when they try to gain audience in the political arena, the state/regime securitises the state/regime and the nation either by means of violence or by violating existing laws.

The famous writings of Frantz Fanon are a telling illustration of the glorification and need of use of force against the traitors in this case the French soldiers and collaborators during the independence war. Fanon argued that the major weapon of the colonizers were the imposition of their image of the colonizers on the subjugated people. In order to be free, therefore, the subjugated population had first to purge their minds of the French culture. Therefore Fanon recommended violence as the way to this freedom, matching the original violence of the alien imposition (Crenshaw, 1978). This use of violence is not at all new in the history of Algeria. As written above (p? ) some researchers even writes about a long history of 'culture de guerre' or ' imaginaire de guerre' (Martinez, 2000; Carlier, 1996). There has for many centuries been struggles as for 'purifying' different territories for what was considered 'intruders'. This friend/foe - outside/inside - dichotomy has framed the representation of the national history writing.

The securitisation of the idea of the unitarian state-regime and nation thus excludes alternative representations of the construction of the history of the Algeria. A narrative around war has constructed the heroes that have acquired symbolic significance in the reproduction of national identity. The reference to the heroes or the martyrs (shouhada) (11)- as the dead freedom fighters are called in Algeria- is a constitutive part of the story telling about the construction of the nation-state. The preambles of the various constitutions, including the latest one from 1996, have inscribed that Algeria was born in 1962 upon the sacrifices of its sons and that 'November 1 1954 was the height of its destiny. The current struggles are therefore firmly anchored in the glorious past of the nation'. (12) Furthermore, article 8 of the 1996 Constitution stipulates that it is forbidden to betray the ideals of the November Revolution. A consequence of the securitization of the glorification of the Revolution is that the constitution of 1996 also requires that the president is able to prove his participation in the independence war if he was born before 1942. If born after, he has to prove that his parents did not collaborate with the French. Hence, the purpose of preserving presidential revolutionary purity is to guarantee that no traitor to the revolutionary ideals penetrates the national body. The memory of the war is thus considered the glue of the fusion of the state-regime and nation. Purification of the national territory from traitors, the glorification of martyrs, the story about one million dead in the war, the unification of all Algerians in the struggle against France, and military victory over the French thus constitute the building blocks of the representation of the national history. Contenders for power therefore come from among those who have participated or are descendants of freedom fighters, the Mujahidin. They are the watchdog of the continuation of the revolutionary ideals with other stories being silenced. Notably, there exists no official history writing dealing with the internal killings and massacres of Algerians by FLN. Neither are the harkis who were Algerians that joined the French dealt with in history even though about 40.000 were killed just after the war by FLN (Stora, 1991). Nor is any history writing about the discussions inside FLN with regard to the construction of national identity and of the military's role allowed. Similarly, no history about the Berbers has been officially written, even though the future of the Berbers in a new nation-state was discussed in 1940s-1950s (Stora, 1995). The state-regime possesses a monopoly on history writing and due to the fact that it is obsessed with the break-down of the fusion of state/regime with the nation, it has consequently securitised history as an identity construct.

This securitisation of the representation of a unified state/regime and nation is under pressure. Thus Bouteflika made a concession to the Berbers in 2002 by promising the recognition of their language, tamazight, as a national language having thus the same status as Arab. But no measures have been taken to materialize this promise. At any rate, this process can be stopped at random if claims for cultural equality are considered too dangerous for the unity.

Securitisation of Islam

The state/regime has ever since the independence constructed a chain of equivalence between modernity, rationalized Islam and unity versus a chain of difference of anti-modernity, irrationality and fragmentation. This way of constructing radical 'otherness' has resulted in 'securitisation' of FIS which is represented as a threat to the modern state/regime and nation.

Islam has been and still is represented as the most important national identity marker. (13) But like the concept of the nation, the concept of Islam is held in check by the state/regime. In all of Algeria's constitutions (1963, 1976, 1989, 1996), it has been stated that 'Islam est la religion de l'État'(14) or 'Islam est religion d'État' (15) (Sanson, 1983: 18). But the two statements are not identical because state and Islam are positioned in a different relationship. Islam as state religion signifies that the state administers the religion, whereas 'Islam as religion of the state' indicates that Islam is either subordinated to the state or the state is subordinated to Islam (Babadji, 2001: 56). In the former case, Islam puts its resources at the disposal of the legitimisation of a political project. In the latter, the state acts in conformity with the religious dogma.

The state/regime is not subordinated to Islam. It is Islam that is subordinated because state/regime administers, rationalizes and instrumentalizes Islam for political purpose. Islam is represented as a social, moral and political public service system (Leca & Vatin, 1975: 58), whose purpose has been the legitimisation of state-sponsored modernisation. Thus, until the 1989 constitution when references to socialism were deleted, Islam was linked to the socialist modernisation project in the FLN discourse. In particular, the egalitarian principle of Islam was represented as an underpinning of those of socialism. In Jean Leca's and Jean-Claude Vatin's formulation: "C'est bel et bien en tant qu'idéologie, reflet d'une infrastructure sociale et économique que l'Islam est considéré" (Leca & Vatin, 1975: 260). The state/regime thus monopolized the interpretation of Islam. Alternative interpretations of Islam have been more or less silenced. This especially goes for the marabou tradition and the Sufi-brotherhoods, which were considered anti-modern in the sense, that they were seen as irrational and associated with small communities that the state/regime represented as a hindrance to the concept of the modern and unified state/regime and nation.

The harbous (religious properties) which constituted religiously independent properties, thereby putting into question the power of the state as organizer of the whole territory, quickly became nationalized after independence. Furthermore, the construction of mosques was controlled by the state, which did not allow for 'free' mosques because they could be politically uncontrollable. Finally, a decree from 1969 stipulated that the Imams' status had to be identical to that of ordinary state-employees. The Imams thus became politically and administratively subordinated to the state administration, which very often dictated the content of the Friday prayer. (Lamchichi, 1988). The political control of the Imams is still functioning and has even been strengthened. An amendment of June 2001 to the constitution curbed political speech in mosques by lengthening to five years the maximum sentence for delivering sermons that are capable of harming social cohesion.

This permanent control of the mosques and thereby also of the practice of faith, is due to the inherent tension between administration and subordination of Islam to the state, and the state as subordinated to Islam, that is, between 'Islam as the religion of the state' and Islam as state religion. This has been increasingly evident since the weakening of references to Socialism from the mid-1980s. One might argue that both the disappearance of the reference to socialism in the 1989 Constitution and the fragile balance between 'Islam as religion of state' 'Islam as state religion' has opened up for discursive space for the political Islamism represented by the FIS (the Islamic Salvation Front). The disappearance of the concept of socialism has left references to the political egalitarian principle orphaned. Here, political Islam offers itself as a substitute by referring to an egalitarian concept of umma (the community of believers). This offer was for example very evident during the earthquake in Algeria in May 2003. As usually, the Islamic NGO's were the first on the spot with regard to the organisation of help. They were so efficient that the regime prohibited these NGO's from helping. Only the state-sponsored NGO's were allowed to help although the regime was not able to organise quick help.

The dominant state/regime discourse and that of the banned FIS share the concept of the fusion of the state/regime-nation and Islam, which does not allow for cultural and political pluralism. But they differ on one essential point: The FIS has turned upside down the order of the concepts of the state/regime, nation and Islam. The FIS discourse posits Islam as the transcendental sacral referent object: Allah. That is to say that Allah is the sovereign ultimate referent object that justifies and legitimizes earthly politics. Legitimization of the state/regime and nation is thus derived from Islam. Thereby FIS - but of course also other Islamic parties - subscribes to the concept of 'Islam as the religion of the state'. These parties thus took seriously one of the constitutional terms of the state/regime itself. Therefore, one might say that FIS is the son of FLN (FIS est le fils du FLN).

The state/regime discourse posits, by contrast, the sovereignty with the people, which in turn is incarnated in the immanent state/regime. Thus the immanent state/regime is the ultimate referent object because it incarnates the people.

State/regime as the people and state/regime as expression of Allah's will are in both discourses represented as untouchable. Nobody is allowed to enter the 'sacralized' state space unless they prove that they are the true representatives either of the people or of Allah. The result is that both parts securitise their representation of the 'just' the state/regime. This securitisation leads to killing either in the name of Allah being incarnated in the Sharia-state or in the name of the state of the people.

In the radical Islamist discourse (16), it is not only the Sharia-state that is securitised. It goes also for the faith itself. The way the radical Islamists interpret the expressions of faith is securitised. Thus, all who do not adhere to their interpretation of faith are considered to be living in a state of 'impiety', and therefore in a state of djahilliyya (ignorance) (Labat, 1995; Roy, 1995). The infidels are considered as vira that contaminate a former healthy societal body. Purification of the present impure society is therefore considered as a necessity if life shall be meaningful. License for killing is therefore easily given. However, the random killing undermines the legitimacy of the discourse of the radical Islamists. The same goes for the legitimacy of the state/regime discourse that securitises its version of Islam and therefore does not open up for alternative interpretations of Islam.

Religion is an easy target for 'securitisation' because faith is about being. If somebody is not allowed to practise his way of conceiving faith, life is worth nothing. But although faith can be the referent object of securitisation, it does not automatically results in actions of violence. In the words of Mark Juergensmeyer: "Religion is not innocent. But it does not ordinarily lead to violence. That happens only with the coalescence of a peculiar set of circumstances - political, social and ideological - when religion becomes fused with violent expressions of social aspirations, personal pride, and movements for political change" (Juergensmeier, 2000: 10). All these above mentioned social and political elements are present in Algeria. Together with the securitisation of state/regime, nation and a specific way of instrumentalizing Islam makes the societal situation extremely explosive.

The Future: Which kind of Future?

FIS did not come into power in 1995. Instead, the military and the security forces got increased power. The state/regime has succeeded to a certain extent in containing violence especially in the big towns. But the tension between securitisation of state-regime, nation and Islam and the increasing societal protests has resulted in the political paralysis of the army and thereby of Bouteflika. It becomes still more difficult to introduce political and economic reforms because of violence that continues to been perceived of as a threat to the state/regime although the state/regime has won the battle of the survival of the state as institution because it has succeeded in splintering the different Islamic groups. Furthermore, the FIS program for a Sharia-state has lost much of the support as a result of the violence they have carried out in the name of Allah. This, though, does not mean that the question of the position of Islam in relation to the state/regime and the nation has disappeared. There is still a discursive struggle going on concerning this issue where the conservative, traditionalist Islamic parties have gained electoral ground.

Societal demands for justice and transparency regarding the position of the army have increasingly come to the political fore. But the state of emergency is still functioning and FIS is still banned because of its message about a Sharia-state. So, although the state has survived as an institution, the regime, and thereby the army also, still has not given in to a de-linking of the regime from the state. The army still remains the only actor that decides when and how the relationship between the basis codes might change. As long as the state of emergence lasts it is the army who decides the 'stop and go' policies of securitisation and timid attempt at desecuritisation.

It is possible to put forward three future scenarios: the first one is total chaos with its attendant full-scale securitisation of state/regime, nation and Islam. The second one is a controlled process of democratization which has been on track since 1999 up to 2002 but which is slowly broken down. Finally there is status quo. The latter one is highly improbable because of the overt societal claims for reforms. The first one is possible but not evident. This is due to the fact that the reforms have started and the army is not keen on intervening overtly in the political process as it did in 1992. A military coup would damage severely the image of Algeria at the international scene. But it would also make especially Europe but also the USA think about the possible necessity of intervention that they never have did before even during the worst period of violence 1992-1996. But time has changed and intervention is at the American agenda. Therefore the state/regime certainly will try to cultivate a 'civilized' image. Thus, the second scenario is the most likely for the future. But that demands that Bouteflika will be able during the presidential election 2004 to convince the population that the regime will de-link its survival from that of the state.

REFERENCES
Ayoob, Mohammed, 1995. Third World Security Predicament. State Making, Regional Conflict and the International System. Boulder CO: Lynne Rienner.
Babadji, Ramdane, 2001. 'De la religion comme instrument à l'identité comme sanctuaire: quelques remarques sur la constitution algérienne du 28 novembre 1996', in Ahmed Mahiou, Jean-Robert Henry, eds., Où va l'Algérie? Paris, Aix-en-Provence: Karthala - IREMAM. (53-71).
Badie, Bertrand, 1992. L'État importé. Paris: Fayard.
Bayart, Jean-Francois, 1989. L'État en Afrique. La politique du ventre. Paris: Fayard.
Buzan, Barry, Wæver, Ole, Jaap de Wilde, 1998. Security: A New Framework for Analysis. Boulder, London: Lynne Rienner Publishers.
Carlier, Olivier, 1995. Entre Nation et Jihad. Histoire sociale des radicalismes algériens. Paris: Presses de Sciences PO.
Hibou, Béatrice, 1996. L'Afrique est-elle protectionniste? Paris: Karthala
Hutchinson, Martha, Crenshaw, 1978. Revolutionary Terrorism: The FLN in Algeria, 1954-62. Stanford, Calif: Hoover Institute Press.
Jabri, Vivienne, 1996. Discourses on Violence. Conflict Analysis reconsidered. Manchester and New York: Manchester University Press.
Juergensmeyer, Mark, 2000. Terror in the Mind of God: The Global Rise of Religious Violence. Berkeley, Los Angeles, London: University of California Press.
Kaki, Maxime, Ait, 2003. 'le Maghreb face aux revendications berbères', Politique étrangère 1, Printemps: 103-119.
Lamchichi, Adderrahim, 1991. L'Algérie en crise. Paris: L'Harmattan.
Leca, Jean,Vatin, Jean-Claude, 1975. l'Algérie politique. Institutions et régime. Paris: cahiers de la Fondation national des sciences politiques. No. 197.Presses de la fondation nationale des sciences politiques.
Leveau, Rémy, ed.,1995. L'Algérie en guerre. Bruxelles: Éditions complexe
Mahiou, Ahmed, Henry, 2001. 'Les contraintes et incertitudes du système politique', in Ahmed Mahiou & Jean-Henry-Robert, eds., Où va l'Algérie? Paris, Aix-en-Provence: Karthala-IREMAM.
Martinez, Luiz, 1998. La guerre civile. Paris: Karthala. ´(The Algerian Civil War. 1990-1998). London: Hurst & Company.
McSweeny, B, 1996. 'Identity and Security: Buzan and the Copenhagen School', Review of International Studies, 22/l: 81-93.
Meynier, Gilbert, ed., 2000. L'Algérie contemporaine. Bilan et solutions pour sortir de la crise. Paris: l'Harmattan.
Mohsen-Finan, Khadija, ed., 2002. 'L'Algérie: une improbable sortie de crise?' in Les notes de l'IFRI, No 37 (2002), Paris.
Neumann, B. Iver, 1998. ‚Identity and the Outbreak of War', International Journal of Peace Studies 3(1): 7-23.
Provost, Lucile, 1996. La seconde guerre d'Algérie. Le quiproquo franco-algérien. Paris: Flammarion
Roy, Olivier, 1995. La généalogie de l'islamisme. Paris: Éditions EPRIT.
Salamé, Ghassan, ed., 1993. Démocraties sans démocrates. Paris: Fayard
Salamé, Ghassan, 1994. 'The Middle East: Elusive Security, Indefinable Region', Security Dialogue 25(1): 17-37.
Sanson, Henri, 1983. Laïcité islamique en Algérie. Paris: Editions du Centre National de la Recherche Scientifique.
Snow, M. Donald, 1996. Uncivil Wars. International Security and the New Internal Conflicts. Boulder, London: Lynne Rienner Publishers.
Stora, Benjamin, 1991. La gangrène et l'oubli. Paris: la Découverte.
Stora, Benjamin, 1995. L'Algérie en 1995. La guerre, l'histoire, la politique. Paris: Éditions Michalon.
Stora, Benjamin, 2001. La guerre invisible. Algérie, années 90. Paris: Presses de Sciences PO.
Tilly, Charles, 1984. 'War Making and State Making as Organized Crime', in Peter.B. Evans, Dietrich Rueschmeyer, Theda. Skocpol, eds., Bringing the State Back, Cambridge: Cambridge University Press (169-191).
Wæver, Ole, 1995. 'Securitization and Desecuritization', in Ronnie D. Lipschutz.ed., On Security. New York: Columbia University Press (4686).

* Dr. Ulla Holm, IIS (Institute for International Studies), Copenhagen

NOTES
1) El Watan, La Tribune,, le Quotidien d'Oran, L'Algerie-Interface.
2) El Watan, November 11, 2001
3) FLN is the acronym for Front de Libération nationale. FLN launched the Independence War in 1954. FLN was the only legal party since the independence 1962 up to 1989.
4) L. Martinez is very much inspired by the research of O. Carlier (1995) who analyses the reasons for violence on the basis of the concept of 'culture de guerre'. Another source of inspiration is the works of Bayart on 'plunder economy' in African states (Bayart, 1989)
5) I do not distinguish between society as a social construct and nation as an imagined community. The two concepts are more or less fused in the theory of the Copenhagen School. (McSweeney, 1996; Albert, 1998).
6) Movement for a Peaceful Society.
7) Http://www.algeria-watch.de/farticle/presse/diffamation.htp. Le Matin, April 17, 2001.
8) http://www. Algérie-interface.com/new/article.php?article_id=217. Alger, May 8, 2003.
9) Neumann slots the action of violence, which he formulates as 'violisation', into the three categories of the Copenhagen School: non-politicised - politicised - securitised - violised. He does that in order to take into account the difference between securitisation as a speech act and violisation as the act on a certain scale, which might be the possible consequence of securitisation (Neumann, 1998: 18). But even if this is done, the question remains why in certain cases a speech act on security is backed up by violisation. This is a case for empirical research, which Neumann also writes.
10) The irony of history is, that the Algerian freedom fighters wanted to construct a clear line of demarcation to French identity. They certainly did, but at the same time they copied the fusion of the French state-nation. But in opposition to the French idea of a political nation, the Algerian nation is defined in religious, and ethnic terms.
11) After the extremely violent clashes between security forces and Berbers in spring 2001, the Berber coordination group published a list of demands to be fulfilled by the government. One of the demands concerned the status as martyr given to Berbers killed during the confrontation (El Watan, June 15, 2001). This claim contested the hitherto right of the state/regime to define a martyr as one who died during the independence war against the French. Now, the killed is a Berber who contested the state/regime, represented as an actor in line with the former French colonisers.
12) http:///www.mission-algerie.ch/anglais/index2.htm. Preamble to the 1996 Constitution.

13) Classic Arab became the other building block of the representation of the unified nation-state. Only few Algerians were able to read and speak classical Arab. But social and linguistic realities did not count. The choice of classic Arab was made because Islam and the Koran language represented the state vision of a unified Algeria in a unified Arab-Muslim world.
14) Islam is the religion of the state.
15) Islam is state religion
16) The banned FIS is not a part of the radical Islamist discourse. It has rather to be considered a conservative Islamist party. But it is difficult to define the content of its discourse exactly because it is banned and has been split up. Some parts of FIS are voting for other conservative legal parties. Others have joined various Islamist guerrilla groups.