When intermediaries and analysts think about armed conflict they often consider violence an obstacle to a negotiated settlement and thus that one of the keys to unlocking a peace process is the cessation of violence. A corollary of that thinking is that violence is simply a tactic and that armed groups need to be helped to recognise the benefits of a non-violent strategy. However, while the importance of ceasefires should not be under-estimated, the assumption that violence is a tactic that can be replaced by another tactic reflects a partial analysis of the nature of armed groups. Armed groups themselves will often say that they believe in peace and they are struggling for a just peace. In other cases they will say that the violence is an expression of their situation and valid even if it will not bring about change. Therefore to really understand how there can be a transition from military force to politics we need to understand the nature of armed groups. We also need to understand the communities which support them, either explicitly or implicitly, as they too are involved in the thought process discussed in this article and their attitudes are significant influences on the armed group.

Most armed groups have an analysis of their situation in which armed action seems an inevitable, if unfortunate, component. Their existence is often a challenge to the state monopoly on force or its use of terror against their community or class. Some militant groups begin with a theoretical analysis of the state that argues that change can only come through violence, but this is often not the case. Many militant groups have grown out of non-violent democratic movements for change which have been crushed by the state and in some cases the state has consciously preferred the transfer of popular protest into a terror it feels it can control more easily. Ironically, but not surprisingly, armed groups have adopted a cult of force and a power/coercion paradigm from states, imbued with the sense that nations are built through force and the attendant assumption that there is an acceptable level of violence.

The armed group and its supporters will not necessarily have a fully worked out sense of these matters but they will have grasped them intuitively and internalised them as basic aspects of their orientation. So when intermediaries explore with militant groups an end to their violence, they may treat violence as an optional element in the repertoire of options that the group has, when in fact it is an integral part of their raison d'etre and needs to be approached as such.

However, the armed group is open to tactical adjustments in its methods if it sees an opportunity and will manage its military campaign to achieve the best advantage for itself. A ceasefire could give the opportunity for regrouping militarily, presenting a more positive public image and so on. This kind of shift does not imply a deeper change of orientation. But at some stage the military option will have to be replaced by a negotiation process. Even if the military campaign is successful, a process for an orderly handover of power is preferable and this requires some capacity to negotiate the end of the war. So whether dealing with the state or armed groups we can distinguish three possible positions:

• Militancy which believes that military force is the only option
• Dual strategy which still believes in the primacy of force but will use other approaches for tactical advantage
• A conflict transformation strategy

The question for the group is when a transition to a conflict transformation paradigm is appropriate and whether it can manage that transition effectively.

While we can accept that the issue of violence is only one part of the militant mindset, which needs to be appreciated as an integrated whole, we can identify some of its component elements and consider how they may change either as a result of the passage of time or through deliberate interventions which are designed to make dialogue and negotiations attractive. We might think of a set of weighing scales on one side of which are stacked a series of elements tending towards militancy. They include:

• Lack of alternative option – council of despair
• Militant mindset
• Commitment to the campaign
• Avoidance of compromise
• Control of one's own destiny
• Avoidance of splits

There are also material factors which will encourage the continuation of the armed conflict, such as the influence of external sponsors of the conflict or the benefits that the leaders gain from the war economy, but this paper focuses on the internal political dynamics which influence the continuation or curtailment of the military campaign.

On the other side of the scales are other elements that might tend towards a conflict transformation approach, including:

• Real opportunities for change
• Framework document on possible outcomes
• Inherent weaknesses of the military option
• Legitimacy and recognition
• Capacity to minimise risks and concerns
• Guarantees and mutual dependence
• Third-party intermediaries

Can the armed group, with the balance tending towards militancy at a certain point in the conflict, change the balance of forces so that there is a tendency and then a commitment towards a peaceful resolution of the conflict? In turn, the considerations that determine a government's response to a militant challenge inhibit it from changing its own approach so as to encourage the move to the negotiation table. How can an armed group shift from its struggle-sustaining approach, and can its opponents or intermediaries facilitate that shift? A movement within civil society can influence both the government and the armed group towards an awareness that negotiations and dialogue can provide a better alternative to violence but it too needs to understand the factors that maintain the armed confrontation and not be naive about the dynamics at work. Let us look at some of the elements on the scales.

As noted above, armed groups often feel that violence seems to be an inevitable aspect of their struggle. It could be described as a council of despair – used both by states and armed groups. It seems that there is no option when faced with the militancy and obduracy of their opponents. This could be a political stance to justify violence but more often it is a genuine expression of the belief that other options such as democracy, dialogue and negotiations will achieve nothing and conversely undermine the struggle.

So a militant mindset emerges, often with justification from the actions of their opponents. It incorporates a belief that the state and the establishment have no interest in sharing power – but maintains itself using this power and capacity to oppress. The state may recognise democratic processes but only for certain privileged sections of society, and it responds only to force and coercion. In turn the armed group relies on military force and is suspicious of the power of reason, moral imperatives or conflict resolution processes in the absence of some equality of military power. It also rejects the value placed on stability, which is the point at which it differs in its orientation from the state. It recognises that stability is often accepting the status quo and is rather impervious to the arguments of well-meaning intermediaries that they should think about the suffering of their people and that it is only through ending the violence that a new society can be created. They are well aware of the cost of the conflict but see no hope that stability will bring justice and find those who argue this naïve or even disingenuous.

Beyond that, ideology is not necessarily a defining characteristic of militant movements, nor can it be used as a determinant of the commitment of the group to armed action. As shown by the case studies, a group can contain within itself a wide range of positions – from the right to the left, from a nationalist analysis to a class analysis – if all are against the status quo.

Certain characteristics develop during the course of the campaign, which in themselves help to sustain commitment to the campaign. It is seen as crucial that there are no compromises, no retreat, no turning back and no doubt. Any of these can undermine the campaign. Banishing doubt in turn means maintaining discipline so that there is no challenge to the leadership and no questioning or rethinking. Armed groups will often hold fast to maximalist goals including the break-up of the state even though observers might think these impossible for their opponents to concede. This may in part be a sign of political naivety but it also demonstrates the perceived need to be uncompromising and not to weaken one's demands.

A common explanation for how the struggle will bring victory in the face of superior strength is based on the determination of the armed group and the people who support them. The rhetoric that captures this sentiment is, 'he will prevail who can bear the greatest suffering, not he who can inflict the greatest suffering.' This logic allows the campaign to continue even when it appears hopeless. It is ironic that sometimes both sides, in a mirror image of each other, try to influence their opponent by inflicting suffering but are aware that their own response to more suffering is continued resistance.

A powerful motivator of continued struggle is the distrust of both friends and enemies. The group will argue that they can only rely on themselves and they can only do that while the struggle continues. Once they enter into negotiations they are at the mercy of both their opponents – who clearly do not share their goals and want to undermine them – and their friends – who do not share their commitment and will be willing to compromise on core issues, not realising their fundamental importance. This distrust underpins an avoidance of compromise and supports the belief that one is in control of one's own destiny. Even if the struggle fails, they will know that they persisted and believe that nothing more could have been done.

The cumulative effect of these concerns is the armed group's acute awareness of the risks involved in moving into negotiations. It is also believed that a change of strategy or the making of compromises exposes the group to the danger of internal splits. The folklore of armed groups is that they must remain united to win. While the group maintains consistent intransigence there is little space for alternative factions to emerge, but introducing new ideas can disrupt that unity. If they do start to contemplate new possibilities, the leadership will need to be very sure of their authority if they want to lead their group in a new direction that almost inevitably challenges some of the old certainties that have been central tenants of the movement. A militant once said that you become aware of the effectiveness of your propaganda when you try to promote a new direction and you find your group and supporters wants to hold on to the old arguments and now use them against you.

All of these considerations will often make it almost impossible for a group to change its approach in spite of good arguments that show how the strategy is not working or the costs are too high. The group may well feel the force of these arguments and know that their campaign is proving counterproductive but can do little to change. The more the group has suffered and the greater the sacrifices made, the harder it is to explain why all that effort is now to be put to one side and a new approach adopted. They will also be faced with the problem of the law of diminishing returns: the longer the conflict continues the less impact they have unless they can find some way to escalate the struggle and that may be very difficult to do. So continuing as before in effect leads to a loss of momentum.

In order to change direction, the group has to be satisfied that there are likely tangible benefits and real opportunities for change. They need to be able to show that more could be gained from negotiations and dialogue than from the continuation of the struggle. They need signs that their analysis of their opponent's intransigence may be limited and that there is space for alternative ways of relating to their opponents. The signs may be small but must be enough to allow an internal debate to develop. For example the Irish Republican movement was helped by the statement by the United Kingdom government that they had no selfish or strategic interest in Northern Ireland and would leave if the people of Northern Ireland asked them to leave. The Naga leadership in northeast India were helped by the Indian government's statement that it recognised their special history.

It is not surprising that in preparing for talks the parties may be unduly optimistic in looking for opportunities for change. For example, the armed group will often want some agreement for the interim government of the areas they control and they will want to use these arrangements to try to advance their legitimacy, while their opponents will resist conceding such points as they will want to bargain over such recognition and safeguard their own interests. It seems attractive to decide on interim arrangements before protracted negotiations start but experience suggests that parties can only enter into negotiations when they have some idea of the parameters of a settlement. A framework document outlining these parameters has often been an effective element in bringing about a ceasefire and peace process.

If such indications are forthcoming they may then be able to evaluate the impact of their present military campaign and it may then be possible to explore some of its inherent weaknesses. Armed groups are normally very secretive and reluctant to interact with people outside their movement, but at this point they often reach out to potentially sympathetic observers or individuals from their community. In such dialogues it has been possible to help the group to reflect on alternative analyses and strategies, letting the group see how violence may often alienate constituencies whose support is necessary, or how opponents' and the general public's resolve is strengthened. This is an argument that the new Palestinian leadership is trying to develop in relation to the impact of suicide bombings. They can argue that as a result of the bombings the Israeli people are more afraid and therefore support more oppressive policies. The end of the bombings may also make the international community more amenable.

Not only is there the possibility of more support but also there is the possibility of more legitimacy and recognition. A ceasefire is often demanded as a precondition for allowing militant groups to enter into talks. In Sri Lanka the ceasefire in the conflict between the Liberation Tigers of Tamil Eelam (LTTE) and the Sri Lankan government allowed direct talks to begin between the parties and over time a number of states were willing to deal with the LTTE. An alternative scenario can also be demonstrated from the Sri Lankan experience. One of the reasons the talks broke down (though the ceasefire remains more or less intact) was the US government's refusal to allow the LTTE to attend a donor meeting because it is still a proscribed organisation in the US. The negotiations are still stalled, the scepticism of the LTTE has increased. It continues to demand a clear framework for renewing negotiations and it is careful not to be seen to respond to threats and inducements.

The issue of recognition and leadership is not straightforward. Recognition of the validity of its claims – if not its legitimacy – is important for armed groups giving powerful impetus to the shift from violence to negotiations. But on the other hand it means that the militant group is recognising, if only implicitly, the legitimacy of its interlocutors to represent the state. It is interesting that in the Karen case study we see that at certain points the Karen were more willing to engage with the State Law and Order Restoration Council than were other pro-democratic forces within Burmese society. This was presumably because their struggle was secessionist and they were more ready to deal with the power brokers within Burma, while the other groups in their struggle were questioning the very legitimacy of the government. The government tried to use that difference in orientation to split the ethnic movements away from the Democratic Alliance of Burma.

Even when there are potential benefits of a new negotiation-based strategy, such as relief from superior military force or recognition of the concerns of the militant group, on the other side of the balance will be the risks and concerns that follow from a ceasefire and involvement in talks. Can splits be avoided? Can there be a return to war if necessary? How will the morale of the cadres or volunteers be maintained during the ceasefire and are there ways to integrate them into a political movement? The group will have to consider if the risks can be minimised or neutralised. This will probably require a sustained period of consultation and political education within the movement and its supporters. Many such changes originate among prisoners who are removed from direct involvement in the struggle but are living it out in their captivity and therefore have the interest, credibility and opportunity to begin a process of rethinking. Often the group's opponents will not help this process, trying to delay serious negotiations because they believe that the longer a ceasefire holds the more difficult it will be for the armed group to return to war and maintain that threat. But the opposite reaction may well occur. The armed group will be acutely aware of the pressure to return to war, and if there are no obvious benefits from or progress in the negotiation process they may end their ceasefire to show that the threat of violence needs to be treated seriously.

The armed group may ask for guarantees and their opponents or third parties may offer these. However, as the rationale of armed groups is partly based on lack of trust and the need to control their own destiny, it is very difficult for the group and its members to believe that these guarantees will be meaningful. On the other hand, guarantees may be useful in justifying a change of position after it has been decided upon. Developing an analysis that can demonstrate that opponents are likely to meet their commitments because it is in their own interests to do so is more helpful. In this way one may reach a stage of mutual dependency and that may be a more comfortable prospect for the armed group in that it equalises the balance of power to some degree. There are other examples where networks and alliances have been forged which can provide mutual protection and security and avoid the influence of wider global politics, such as the Esquipulas process in Latin America. This can create space in which deals can be made.

A third-party intermediary may also provide guarantees in some situations, but this may not be their most useful role in facilitating the shift from military action to political engagement. Intermediaries can provide a framework for negotiations and invite the parties to take part. The conflicting parties can then come together under the sponsorship of the intermediary rather than one or other party being seen to be taking the lead. This is particularly important when one party is an armed group without the same status as a state party. Problems of protocol and status can then be avoided if a third party is willing to play an intermediary role. The Norwegian government has played this role in relations between Israel and Palestine, in Sri Lanka and elsewhere. In the Sri Lankan case they also tried to play the role of a guarantor and 'referee' and it has been argued that in trying to fulfil the latter roles they made their task more difficult.

This discussion has concentrated on the substance of the concerns that have to be dealt with in order for an armed group to enter into a negotiation process. The case studies provide insights as to how those issues are dealt with within the organisation, where decisions are taken, by whom and through what process. While these shifts vary considerably depending on the nature of the organisation, they generally do not follow from explicit reasoned analysis or expression of underlying principles, nor do they necessarily originate in one element of the group's command structure. Of course there will often be leadership figures who have enormous personal authority such as Veluppillai Prabakharan in the LTTE or, in a very different way, Joseph Kony in the Lord's Resistance Army. In some cases it appears that they can take independent decisions while in others they have to be sensitive to potential reactions within the ranks. But in either case the considerations mentioned above will be at the heart of the thinking behind a change unless they are totally nihilistic leaders.

The process of internal debate is often presented as an argument between 'hawks' and 'doves' with each tendency trying to win over the other side. The case studies bear out a different dynamic. Shifts in movements seldom come from a change of attitude on the part of individuals but rather because the balance of arguments at any one time favours the analysis of one or other group within the movement. Perhaps the most significant influence on the credibility of any set of arguments is the inferences drawn from external events. When those circumstances favour maintaining the war then the doves will be silent but when circumstances favour negotiations then the hawks will be silent. For example, the El Salvador case study shows that the reforms before the peace agreement were partial and imperfect but they "gave credibility to the idea that working politically in the context of peace was more beneficial than continuing war." Opponents often try to expose differences in opinion within the armed group but this may have the effect of closing ranks. But ultimately, if the state or intermediary, or indeed the armed group itself, wants to promote negotiations, it is necessary to strengthen the trends that give a peace process credibility.