In each country, an initial alliance will need to be formed to develop an implementation strategy. This is likely to be composed of like-minded organisations with some prior experience of working together and a shared commitment to campaign aims and principles. We are looking here for a balance between relations of trust within the group and diversity of membership, and between a capacity for strategic thought and a capacity for action. Getting the right allies together will be key to a successful start: a small but sound base of organisations with natural affinities, outreach and credibility, is preferable to a larger but less cohesive initial alliance. (In India, for example, the campaign started with 40 allies; within two years this had grown to 400.)

This group will aim to grow, as other organisations are mobilised or themselves seek involvement. There should be no difference of status between 'old' and 'new' members; continued effective expansion will be dependent on shared ownership. The wider alliance should ideally be as diverse a group as possible, in terms of organisational type (public, private, non-governmental) and focus, geographical coverage and target audience. As a result, previous engagement with the issue of violence against women is likely to vary widely and cannot be a selection criterion for involvement.

The critical criteria for all alliance members will be:

• Buy-in to the campaign aims and principles
• A clear rejection both of violence and of the promotion of violence
• Willingness to engage with own constituency on the issue of violence against women

The implied exclusion criterion is that membership should not be used to serve political, religious or other ends which are in conflict with the aims of the campaign.

In addition to these, more strategic considerations in involving new members in particular might include:

• What is the entity's outreach in its own constituency
• What work (if any) is it doing on violence against women?
• What might it gain by joining the campaign?
• What will it expect of other allies?
• What resources and value might it bring to the table?

Local, national and regional alliances will evolve their own ways of working, and structures may therefore vary from place to place. What is fundamental, however, is that the structure should allow each member to direct and represent campaign activities in its own geographical area or thematic field, while enabling central development of the common campaign platform.

For such a rapidly evolving, flexible, diverse and geographically widespread campaign as 'We Can', a conventional representation structure – a pyramid of delegated decision-making from local to national and regional – was found to be unsuitable. Instead, therefore, a more relevant structure is being explored.

The proposed structure is for operational support and direction at the national level to become the function of a small core group of individuals from member organisations, each with relevant distinctive competencies and acting on behalf of the campaign rather than their own organisation. This group will lead on national events and strategy, development of common communication materials, and monitoring and evaluation. It will be supported by a national secretariat. Each alliance member within its own sphere of influence – developing its own strategies for mobilising and engaging with change makers and remaining the public face of the campaign with that group, meanwhile, will exercise public leadership. Periodic discussion forums and web-based consultations will be opportunities for the wider alliance to hold the core group to account and contribute to strategy development.

The transition to this new structure is proceeding at varying speeds in the various countries, but is expected to be completed by the end of 2008. In the initial stages of the campaign, Oxfam GB has tended to provide executive leadership at the national and regional levels, while other alliance members took public leadership in their respective spheres. While this has not been a source of friction, the arrangement is clearly neither sustainable nor a sound basis for a broad social movement, and Oxfam's role is accordingly set to shift to one of shared responsibility within the core group.

The nascent civil society structures in many parts of the world often present a challenge to partner-based development interventions. 'We Can' may have an advantage in that alliance members do not necessarily have to develop their own materials or analysis, but are essentially a means of engaging the change makers and communities that are the central focus of the campaign. Nor is that engagement envisaged in terms of material, logistical or training support: responsibility for change, both individual and collective, lies at the grassroots, and the role of the campaign is solely to provide communication, information and opportunities for exchange.

Certainly the initial alliance members will need adequate strategic capacity to contribute to planning and review at the start of the campaign, and should also have both credibility with and an understanding of their target audiences.

Where the strategic capacity of the membership is weaker, the alliance can still be effective provided the core group can provide a framework of direction and support. In South Asia we have found that member organisations are largely happy to work within that framework provided they a) see the value of the common platform, b) have a large measure of autonomy in how they implement within it, and c) are recognised as the public face of the campaign within their own constituencies.

The challenge within a less developed civil society environment will be to build up the strategic competences of the core group, so that it is not dominated by one or two larger organisations. As has been the case in South Asia, this may require a period of capacity building and a progressive shift of responsibilities.

In terms of implementation the flexibility of the 'We Can' model may be an advantage in such contexts, in that alliance partners without the in-house expertise to develop their own materials can use shared campaign tools to support and develop their existing work with their respective client/target groups. 'We Can' in a context of low capacity civil society could also be an opportunity to bolster the analysis and communications capacity of local organisations in relation to VAW and rights issues generally.

The fundamental premise of the campaign is that people change through choice and opportunity. It therefore seeks to promote an environment in which violence is not tolerated, and where those who have committed violent acts see a positive opportunity and reason to change. 'We Can' focuses on engaging the perpetrator and not putting them on the defensive by 'naming and shaming'. But, by condemning the act and the attitudes behind it, 'We Can' seeks to persuade the perpetrator to take responsibility for their action and also for changing their attitudes. In this way, as the experiences of many in the campaign attest, a desire for personal change founded on personal conviction can be triggered, which is expected to be more lasting in its effect than the censure of others.

The campaign recognises that personal change alone is not sufficient to bring about structural reforms, and that political action needs to continue in parallel to reverse wider systemic trends that reinforce inequalities. What 'We Can' itself focuses on, however, is what ordinary women and men and their communities can do within their own sphere of influence to reduce violence and discrimination. It seeks to trigger a mass movement of change at this level, which will both make tangible progress in preventing violence and discrimination against women, and indirectly help to broaden the constituency for political reform. The premise is that structural inequalities are underpinned by widespread discriminatory attitudes, and it is these that 'We Can' seeks to shift.

Changes in public policy and legislation are needed in many parts of the world to provide a legal basis for women's protection against violence and discrimination and recourse to redress. Yet the unenforced edicts on gender equality on national statute books around the world are a clear indication that law and policy are insufficient guarantees of women's rights unless they are matched by a social consensus that such rights are legitimate. While many alliance partners, Oxfam included, work towards policy and legislative change goals, therefore, those goals are not the focus of the 'We Can' campaign. 'We Can' takes one step back, in effect, and seeks first to trigger the attitudinal change that can in turn influence the practices and ultimately the policies of institutions.

We see this approach as complementary to policy-focused campaigning taking place at the national and local level, and believe that a combination of both strategies should yield greater benefits than either on its own. Where national alliances see it as relevant for taking forward campaign aims, they have accordingly taken up active support for specific policy lobbies (e.g. on the Domestic Violence Bill in India and Bangladesh).

'We Can' seeks to tackle violence against women as a social issue, and will be most effective where it brings together a very diverse section both of society and of civil society organisations to work towards the shared goal of ending it. Its focus is therefore not exclusively on women or women's organisations. The campaign's awareness raising on women's rights and gender equality across a large and diverse target group should broaden the constituency for more traditional lobbying on women's rights, however. Grassroots women's groups may also find their own understanding and argumentation around violence and gender discrimination enhanced by the campaign materials and messages.

'We Can' – as its name suggests – promotes the positive message that violence is neither acceptable nor inevitable, and that ordinary women and men have the capacity to end it. The communication strategy avoids depictions of women as victims, instead presenting them as agents of change in their own lives and those of others. By focusing on the actions ordinary people can take at their own pace and within their own sphere of influence, the campaign also avoids the ethos of militant activism which can be dis-empowering for the more reticent majority. The model of person-to-person mobilisation, meanwhile, allows the pragmatic exploration of real obstacles and alternatives and the immediate support that can help those experiencing violence to protect themselves from such violence. Far from encouraging pessimism, therefore, the campaign offers a positive image and model of the agency of ordinary people in ending violence against women.

To ensure a consistently high standard, it is essential that the fundamental communication principles of the campaign be fully understood by the creative teams developing the materials.

The communications approach and the model of change applied by 'We Can' were developed on the basis of ground-breaking work by the Uganda-based organisation Raising Voices (www.raisingvoices.org), with significant adaptation and further innovation to meet the challenges of a mass campaign. More detail on the communication principles and strategy is given in the South Asia campaign strategy, and advice and checklists are available in the resource book published on the Raising Voices website. Training support from Raising Voices has been invaluable in building up an understanding of these principles within the alliance and with a pool of artists and filmmakers over time.

The essence, very broadly, is that campaign materials should:

• be attractive and clear in their message
• present situations and characters people can identify with
• ask questions and encourage people to think for themselves, rather than telling them what to think or do
• present ordinary women and men as capable of taking action to change their circumstances
• avoid stereotypes and representations that demean women or men or present women as powerless victims

This is generally very different from the style of communication people are used to seeing on violence against women, and assimilating the new approach takes time. Some of the early materials tended towards a dramatic representation that was visually powerful but offered no possibility of change and placed women in the position of helpless victims. In the interim, the South Asia campaign has combined commissioning artists who have worked on similar communication materials elsewhere (particularly for Raising Voices in Uganda) and establishing relationships with local artists over time. We have also found that assigning communications development to a specific resource person at the national level can help to maintain consistency of approach.

Another key factor has been adaptation: the campaign in South Asia has adapted several materials originally developed by Raising Voices for use in East Africa. Similarly, communication tools created for one South Asian country have repeatedly been replicated for use in others, with the necessary adjustments of language, storyline and details of dress and setting, thereby capitalising on the considerable investment of time, funds and energy that goes into developing strong materials. Other adaptations have been made across media: poster designs used for billboards or pamphlets, for example.

Countries such as India are fortunate in having an advanced IT sector, and in various parts of the region alliances have been able to tap into the expertise of communications professionals for their materials. Some of these professionals are themselves alliance members, and other partners have benefited from tools they have developed for their own use. Where such resources are not locally available, sharing and support arrangements with other organisations, campaigns or countries will need to be developed.

Campaign communication tools to date fall into three main groups:

• Campaign identity materials (e.g. bumper stickers and rickshaw plates, t-shirts, branded change maker bags, badges, pennants and wristbands) – these serve as conversation-openers, raise campaign profile and provide a visible identity to change makers.
• Messages (e.g. billboards, murals and paintings on water tanks, songs, calendars, posters and handbills) – these are primarily designed to raise awareness and make people think.
• Interactive learning materials (e.g. graphical discussion guides, story books, games, street theatre) – these aid conscious understanding by provoking further reflection, providing additional information, making links with inequality as the cause of VAW, and suggesting that there are things people can do about it.

Common tools developed for use throughout the South Asia campaign to date have been the logo and strap line, poster and videos, the campaign website and a newsletter, and change maker books recounting stories of personal experiences. In addition, national and local alliances have developed various products in all three categories for their own use, often reflecting specific national issues and concerns (e.g. honour killing in Pakistan, trafficking in Nepal).

As the growth of the campaign begins to reduce the funds available to spend on individual communication at the same time as it reaches out to new audiences, innovative means need to be found to keep such materials both accessible and relevant. In Bangladesh, groups of change makers are organising themselves to share one change maker kit between several people, since these are relatively expensive to produce and resources are tight. In India and elsewhere, alliances are looking at adapting some materials originally geared to a rural audience in order to make them relevant for an urban, middle-class target group. Beyond modifying details of clothing, activities and environment presented, this entails more broadly identifying forms of violence and messages about it that this group can relate to.

In seeking to trigger change rather than prescribe it, and in aiming to reach a mass audience, the campaign cannot hope to control the quality of communication at every link in the chain, and some degree of 'message drift' would seem inevitable. The critical factor in guarding against this has been the quality of the communication materials produced centrally and nationally.

To date the indications are that the materials have been sufficiently clear and sufficiently relevant to ordinary people that change makers interviewed across the region generally display an understanding of campaign messages and their relevance to their own lives, even where the organisations that first introduced them to 'We Can' do not demonstrate the same level of ownership.

Further analysis is needed into the impact of various organisations' differing mobilisation techniques on this ownership of the campaign at the individual and collective level. In the meantime, it would appear that clearly articulated communication materials developed for regional or national use and based on campaign principles offer the best safeguard.

Considerable time and thought, and the efforts of communications experts, have gone into developing clear and unambiguous messages and materials to avoid misinterpretation. Moreover, for the reasons set out above, the campaign itself is careful not to prescribe courses of action or propose ready-made answers to the problems of violence against women which might be misinterpreted. As a result, there are no known instances of problems arising from a misreading of the campaign's intent (e.g. vigilante action in the name of 'We Can', or a misrepresentation of campaign goals bringing 'We Can' into disrepute).

What will always be the case, on the other hand, is that people will relate the messages to their own situation, experience and beliefs and come to their own conclusions. Their understanding will also evolve over time. The role of 'We Can' can only be to inform, question and support.

The main point to note is that the focus is on the practice of equal relationships rather than the theory. Two years into the South Asia campaign, many change makers cannot frame the issues they are discussing with their peers in terms of women's rights, yet the kinds of personal change they report making in their own lives are a practical expression of equal relationships.

The communication materials and messaging accordingly use situations that will be familiar to the target audiences to raise awareness of various aspects of VAW, and are framed in such a way as to encourage people to question their own beliefs and behaviour. The aim is that some aspect of what is presented will trigger recognition and realisation related to their own experience. Networking between change makers to discuss and offer mutual support on the situations each face then offers an opportunity to question, challenge and reflect, helping them to move further ahead in their thinking and action both as individuals and as a group. High-visibility national and regional events and communication in turn reinforce that process.

Recognising that a real change in attitudes and beliefs is likely to be gradual, the communication strategy is phased to introduce new concepts over time (see campaign strategy for details). The first-phase materials focus on raising awareness of the violence in people's own lives and the possibility of change. The impact violence has on relations and wellbeing within the family can be a practical, uncontroversial entry point for discussing issues of rights in the language and experience of ordinary people. In contexts where resistance to notions of women's rights is high and such rights are perceived as an external agenda, interpreting them at the practical level of the family/community and how they relate to everyone (men, women and children) allows us to get beyond the reaction and stereotypes to a discussion of lived experiences.

Clearly each person will respond to these materials in their own way and at their own pace. Not everyone will take up a radical position on gender equality, but then, not everybody needs to for change to happen.

A number of alliance members have expressed a similar concern about the comic-book format of some communication materials. While avoiding representations of violence which either demean or demonise and presenting a positive message of human agency for change are principles of the 'We Can' campaign, decisions about the most appropriate style and presentation of materials will need to be made at the regional and country level, and some countries may opt not to use the comic-book format. Similarly, the materials appropriate for one audience may not 'speak' to another, and a range will therefore need to be developed.

That said, this format has had a number of advantages where it has been used:

• Cartoon depictions seem to strike the right balance in many cases between a recognisable situation and an uncomfortably lifelike picture of reality.
• Line drawings can be readily reproduced using standard graphics software.
• Cartoons can provide a useful demonstration of the effectiveness of simple graphics, as encouragement for change makers to develop their own communication materials. Workshops on simple drawing techniques have been used in India and Sri Lanka to promote confidence in what has proved an empowering and often highly effective communication medium, used particularly by students for 'We Can' college notice boards.

The messages and materials would certainly need to be adapted to ensure that they are graphically representative, culturally appropriate and relevant to the context of any new campaign location. However, there is nothing intrinsically South Asian about the 'We Can' campaign model and methodology: they are based on thinking and materials originally developed in Africa, suggesting that intercontinental exchange can be effective.

Because the physical manifestation of violence against women differs from one context to another, it is tempting to suppose that the strategies to end violence against women also need to be very different. When the campaign started in South Asia, each country was initially anxious to stress how specific its own national context was; yet they have ended up developing very similar materials and obtaining a very similar response. In focusing on the differences of context and manifestation we can lose sight of the fact that the underlying causes remain the same and thus the fundamental approach, with some adaptation, should work across different contexts. The model of communication is based on analysis of how human beings change, not how any one group changes. The response to 'We Can' in varied contexts in South Asia provides confirmation of that.

There are two issues here: procedural compliance and strategic advantage. In any country or region, there will be strategic decisions to be made about how the campaign principles can best be translated into practice in the local context. This will include negotiating space for the campaign within existing power structures, including with the formal and informal authority figures that may need to be notified or give their consent to any activities by an external organisation within their sphere of authority. Procedural or customary requirements should not extend to the private conversations and actions of friends, neighbours and colleagues, however, through which much of the campaign's outreach is achieved.

In more tactical terms, there may be an advantage to be gained by going beyond compliance and securing the buy-in of authority figures, but the risks involved should also be weighed in the balance. A powerful local leader might, in some circumstances, exercise undue influence on the campaign, orchestrate a backlash against it in the event of disagreement on content or tactics, or bring it into disrepute by association through their actions or statements. In a context of conflict between two or more factions, endorsement by one could place change makers at risk of being targeted by opposing forces. When considering seeking or accepting the endorsement of any public figure, it is worth taking the time to explore their attitudes and consider past action and pronouncements that might help to measure those risks.

At all events, any individual wishing to reject violence is free to take part, and a leader should be welcome to engage at that level as well. Their public endorsement may also be welcomed, if it is based on their own personal change and the risks to the campaign are judged to be small. What they must not be allowed to do is appropriate 'We Can' into their own sphere of authority or tie it to their own political or religious agenda.

Ultimately, the buy-in of some leaders and authority figures, while necessary, will never be sufficient, because power relations will not be changed by working only with the powerful. It is ordinary citizens, not their leaders, who should therefore be the primary targets of campaign communications.

The campaign and the communication materials are designed to prompt ordinary people to think about various situations and attitudes that they encounter in their daily lives and analyse them from an alternative perspective of equality, non-discrimination and non-violence. Materials developed for any new geographical area should aim for a similar clarity, simplicity and relevance to everyday life.

Literacy and general educational attainment vary widely across South Asia, and the campaign is working in places with zero to low women's literacy in particular. Where levels are low it has implications for the media alliance members tend to use to disseminate campaign messages. In such cases methods involving song, sketches, radio and graphics are likely to be prioritised – all well established tools of community development in other parts of the world as well. Communication materials developed for national campaigns in South Asia, such as a 'kit' of discussion guides for change makers in India and Sri Lanka to use in mobilising others, are largely graphics-based, with only a small amount of text for the change maker to read. These and other ideas could be shared for adaptation and development elsewhere.

It is worth mentioning the powerful analysis and practical conclusions, which many poorly educated people have drawn from these simple materials across South Asia. The clarity and relevance of communications to their own experience seem to be critical factors.

The campaign does not presume to effect change directly; the premise is that each individual and community can change the way they think and act – no one else can do it for them. Instead it seeks to trigger a desire for and commitment to that change, as well as awareness that it is possible.

The accounts of change makers around South Asia tell of a shift in thinking about violence and inequality in human relationships, prompted by exposure to campaign messages, and leading in turn to many micro-level changes of behaviour that demonstrate the mind shift, and significant instances of collective action against VAW. There can be no guarantees, and no two people will respond in precisely the same way, but the indication is that the model can be effective in triggering change in both beliefs and actions.

We believe this is for a number of reasons:

• 'We Can' talks about problems that many people struggle with, but often cannot talk about.
• By focusing on personal change, it taps into people's self-interest, and also puts the possibility of solutions at the level they can influence themselves.
• It engages critical thinking rather than shame or judgement, and mobilises through personal connection and demonstration.
• It offers support and encouragement through local networking and role modelling, and the wider campaign identity.

Any mass campaign presents challenges for monitoring and impact assessment, not least that of attributing outcomes to the campaign itself. The 'We Can' model adds to this the difficulties of measuring changes in private thought and behaviour, and of capturing a spontaneous and multidirectional mobilisation process.

At this stage the focus of monitoring is on changes in understanding, attitude and behaviour over time. A lean participatory monitoring framework has been developed, focusing on the process of mobilising, communicating campaign messages, and influencing the actions of change makers. It uses a combination of a survey questionnaire and a focus group discussion tool to track changes in attitude and behaviour over time among alliance staff members and existing and potential change makers.

This structured mode of enquiry will aid interpretation of the many personal testimonies of change accumulated to date, and will be used to refine methods and materials as necessary. Findings will be made available and the tool can be shared with other campaign alliances on request.

The 'ripple effect' (as it is sometimes described) is the principle of outreach from one person or organisation to others, who in turn reach out to others still, and so on in ever widening circles of influence. In the 'We Can' approach, organisations within the alliance present the campaign aims and messages to their existing constituency or target group, some of whom will become change makers. These change makers undertake to speak to at least 10 other people, some of whom will in turn become change makers and seek to influence 10 or more others. The purpose of this goes beyond numerical strength: by linking up with other people, each change maker secures support for their own change.

As the circles of influence start to overlap and the number of change makers within a given group or geographical location attains a critical mass, an environment is created in which women experiencing violence or others wishing to speak out against violence can receive support and encouragement. As the number of change makers within a particular institution grows, so their influence on the working practices and ultimately the policies of that institution will increase. More targeted mobilisation of key institutions may prove effective once the campaign has achieved a sufficient scale of grassroots support. As public awareness spreads, through regional, national and local campaigning, so more organisations seek to become involved and set off ripples of change through their own constituencies.

The model of change which the campaign seeks to support at each level is one of awareness > reflection > networking > action. This is not a linear process: it starts with small actions, generating further awareness and leading to further actions. The progression of awareness, understanding and readiness for action will take time, repeated exposure to ideas, demonstration, peer challenging and support, and a series of triggers to reflection that will be highly individual and may occur at any time.

The process of individual and collective change is supported both centrally and by alliance members in a variety of ways. Centrally, communication materials are produced to spark discussion and reflection and help change makers to form their own arguments. The quality of these materials is critical to avoid any distortion of campaign messages. Training can also be provided to help groups articulate their own views and experiences through local materials, sketches and songs.

At the level of the individual alliance partners, the diversity of structure, mandate and constituency is such that forms of support vary widely. In any new campaign location, individual allies would similarly need to work out what works best for their constituencies. What is important is that each works with their existing target group and gives them opportunities to come together, whether at village, district or national level, to network, exchange experiences and ideas, and draw enhanced confidence and motivation from the wider group. In this way individual change makers, communities and alliance partners can learn from what works and share innovations.

Some partners in South Asia have developed specific campaign activities. A student's union, for instance, might start with an event on campus targeting existing activists, youth leaders, and others who are already thinking about the issues. This will involve talking about the campaign, violence and inequality in a variety of ways, generally including song or theatre, etc., with a strong focus on recognising violence and the power and responsibility of individuals to do something about it. This initial event will then be followed by opportunities to sign up as a change maker, provision of materials to help discuss the issues with others, and some form of ongoing contact with the union or with other change makers.

Others have opted to integrate the campaign into their normal activities, with the initial introduction to 'We Can' and subsequent discussions and sign-up opportunities incorporated in regular programme meetings. This implies that programme staff with other specialisations, such as agricultural extension workers, becomes the first point of contact. Programmatic events, such as agricultural fairs, will then include 'We Can' stalls and activities, and most campaign visibility will be through opportunistic 'piggy-backing' on other programme work, although specific change maker meetings may later be held for recognition and mutual support.

As the campaign grows, the sheer numbers of change makers will make it a challenge to communicate with each individual. Their diversity, too, is an added complication, especially with their differing contexts, skills and experience, and levels of commitment, the support they look to the campaign for will vary widely. Person-to-person mobilisation and the iterative nature of personal change will also ensure that change makers are at different stages in the process of reflection, networking and action at any one time, with correspondingly differing needs and questions.

The South Asia campaign will need to keep learning and developing new ways of addressing this challenge as it grows in scale year on year, but a number of factors can already be identified.

Key to the communication strategy is mobilisation through alliance members: organisations of various kinds with outreach to a particular audience with whom they can broadly maintain contact, making it possible for phased messages to be communicated over time. While a proportion of change makers will become involved through participation in open-forum events, therefore, the core of the campaign is likely to comprise groups targeted by the alliance partners, with whom they have an established relationship and communication mechanisms. As new change makers are mobilised directly by their peers, alliance members will need to be setting up contact opportunities for the new group, at the same time as the campaign explicitly encourages networking and support among change makers at the local level (a specific focus of phase two communications).

The means of follow-up may vary between countries and organisations. In Sri Lanka, where campaign activities have been integrated into the core programmes of diverse organisations, it is expected that ongoing programme meetings and events will continue to provide a channel for communication and support over time. In Bangladesh, the alliance has instituted a series of review meetings with change makers, led by member organisations, as a space to explore individual and collective problems, concerns and plans and a means of supporting the progression of thinking over time.

At a very practical level, the database of change makers developed for monitoring purposes could also aid follow-up, particularly where individuals are mobilised through public meetings and do not have a prior link to a member organisation. At the same time, regional events and communication are intended to provide a continuity of profile while disseminating a carefully designed sequence of messages to help maintain the momentum of awareness raising and reflection in established change makers and their communities.

Many in the first circle of change makers are already engaged in some form of action against violence and discrimination. However, since the aim is to shift social norms, the main target is the silent majority who together set the boundaries of acceptability. The 'We Can' approach is therefore designed at two levels to extend beyond this first group to others with no prior engagement:

• At the macro level, a diverse alliance – from self-help groups to government offices and from colleges to trade unions – increases the chances of outreach to audiences with less prior exposure to concepts of women's rights.
• At the micro level, the principle of interpersonal mobilisation likewise broadens the potential range of influence, since personal ties can cut across other lines of social categorisation – and prove far more powerful than a campaign slogan from a stranger.

As the reach of the campaign extends, so that a whole community or institution is aware of it and a critical mass are directly involved, it is hoped that the environment created will be sufficiently supportive of equality that violence will become harder either to commit or to ignore. In practice, this means that, as communities begin to articulate and demonstrate alternatives, perpetrators will come under increasing social pressure to take responsibility for their attitudes and practices, and change.

It should also be noted that resistance could be encountered where it is least expected. For example, the staff of NGOs, often suffering from 'message fatigue' due to an accumulation of workshops, strategy meetings and theoretical frameworks, can prove less receptive to the campaign than others with less previous exposure to the ideas. Some alliance staff have been involved in organising campaign activities for a considerable time before some incident, image or conversation triggers a realisation that this is directly relevant to their own lives and problems.

'We Can' does not offer training, for a number of reasons. One is the sheer scale of a mass movement, making direct training unfeasible. Another is that training has often been found to have little impact on practices and ideas that are rooted in social norms and reinforced by the immediate environment.

The campaign posits that people do not need to become experts on VAW in order to tackle it in their own lives; they need only to see it as something they can and want to change. An effective change maker is not one familiar with the concepts and jargon of gender analysis, but one with the ability to understand and discuss violence and discrimination in their own lives, and talk to their peers about the sensitive and challenging issues it raises. Rather than gender training, therefore, the best way to support them is by providing exposure to the issues and materials that can help them to initiate and sustain a process of understanding and change in themselves and in those around them. 'Conversation-openers' to date have included badges in Bangladesh, pennants in Sri Lanka, wristbands in Nepal, calendars in Pakistan and t-shirts in India.

The campaign seeks to spark awareness in ordinary women and men that VAW is an issue for them personally and one they have the capacity to do something about. The idea is not that the campaign alliance recruits volunteers to work to promote its aims, but that change makers identify goals of their own in response to 'We Can' messages and discussions with peers. The campaign can provide the stimulus and encouragement to people to live up to their own moral aspirations. Once someone embarks on that course, it is then in their interest to raise awareness in others in order to garner support for their own change process – and so on in ever-widening circles of change. Alliance partners and change makers in Nepal remark that requests for material incentives are common when individuals and communities first come into contact with the campaign, but that these tend to die away once people start to make it their own and see that they can make changes in their immediate sphere, to their own benefit and without necessarily incurring personal costs.

At the same time, there are many reasons why someone might be attracted to the campaign, beyond the immediate benefits of reducing violence in their own life. Some change makers see the appeal in improved self-esteem, social recognition and respect – and these are benefits the campaign can foster through public acknowledgement of the role of change makers. For others it is a chance to be part of an international campaign, or to support a local organisation they admire. Whatever the 'pull', to date a combination of more peripheral factors such as these and personal identification with the central aims of the campaign has proved sufficient to spur over half a million change makers into voluntary action.

Organisations, like individuals, have to come to the campaign by choice, not under pressure. Many will have some aspect of women's rights, violence prevention or human rights more generally as their main focus, and will see 'We Can' as a medium for advancing awareness and action on those issues. Others will have their mandate in a very different area, but see gender equality or ending violence against women as a critical factor in achieving their other aims. Some may have no organisational link to women's rights but become involved through the personal commitment of their staff or leaders.

In each case, there is 'something in it for them': an inherent incentive to get involved. (Secondary factors such as the opportunity to raise the organisation's own profile can also be an attraction.) As with individuals, there are reports of NGOs initially seeing the campaign as a funding opportunity above all, then becoming convinced that it is something they want to take part in for its own sake. Without this sense of what might be termed enlightened self-interest, the campaign will not maintain the momentum to achieve the breadth or duration of coverage required to make a significant impact. For that reason, no incentives as such are provided, although significant central funding has been allocated for campaign activities where alliance members are unable to meet the costs themselves, particularly in the early stages. (See below for more on campaign financing.) How the alliance in each country organises to support campaign activities will be largely a matter for national decision-making. However, if the campaign is to achieve the scale necessary to mobilise a social movement, central funding for the activities of each member organisation will quickly become unfeasible.

One way of addressing this is for alliance members to source their own funding separately, but that will not always be possible. An alternative approach taken by the Sri Lanka alliance has been to integrate campaign mobilisation into members' existing activities: to include 'We Can' communication materials and orientation sessions in the agendas of business skills training or public health promotion events, for instance. This both reduces the campaign-specific costs and broadens the potential outreach beyond groups already targeted by other gender equality initiatives.

If it is to become a social movement in a new region, it will need to be owned by a new alliance in the countries concerned: they will need to see it as relevant and necessary in combating VAW in their own context. In South Asia, some considerable time was given over to developing the strategy for the region within a group of interested organisations, and it would seem advisable to adopt the same approach elsewhere.

This initial alliance should small enough for reaching a consensus to be a feasible proposition, and broad enough to have a significant constituency and potential influence at the community level; 20 organisations may be taken as a rough guide. They need not be all women's groups but should have some natural affinity to ease communication and keep them together. (In South Asia, for instance, the initial group was composed of Oxfam partners; some funded by Oxfam and others co-implementers on other activities, which agreed to the need for a large-scale, community-focused campaign.)

The key tasks of this initial alliance will be:

• To crystallise and agree on the common campaign strategy
• To develop appropriate communication materials for the first phase (awareness raising)

Other considerations that might be taken into account at the planning stage are:

• Whether a phased introduction is preferable (starting in the countries where the alliance members are most enthusiastic, means of implementation are clearest or resistance is likely to be lower)
• The timeframe to which the alliance will need to commit in order to have a lasting impact; an initial commitment of five years would be the minimum
• Realistic but ambitious target figures for the numbers of ordinary people to be mobilised as change makers, based on this timeframe and the existing outreach of the initial alliance members
• Development and management of a change maker database from the outset, to enable numbers to be tracked and contact maintained with change makers

In South Asia, for the first two years Oxfam GB has financed the initial core administration and development costs (campaign identity, communication materials, coordination and regional and national support, sharing and learning) and covered some expenses of other alliance members for specific activities, largely from its own funds (including specific supporter donations) and those of other Oxfam International affiliates. It has also tapped into other donor funding through the Oxfam programmes into which the campaign has been incorporated.

Other alliance members have likewise integrated campaign activities – change maker mobilisation, campaign visibility, messaging on ending violence against women, etc. – into their ongoing programmes, and in many cases the staff of alliance organisations also take it forward in their day-to-day work. In this way the administrative and mobilisation costs are shared across the alliance, albeit in varying proportions.

However, as 'We Can' in South Asia moves into the next phase, with larger numbers of men and women coming on board as change makers, more and more funds will need to be put into the communications that will form the backbone of the campaign. It will increasingly become difficult for a single organisation to raise the funds required for a campaign of this scale. Hence, alliance members will be encouraged to raise funds directly for specific components of the campaign.

Attracting funding for gender equality is always a challenge, and an innovative approach such as this does present a degree of risk for a donor. This makes it likely that, when launching 'We Can' elsewhere, larger NGOs will need to invest a share of their own funds in the early stages at least, until sufficient momentum has been demonstrated to attract either institutional or private funding. That being said, as the campaign becomes more widely known and evidence of impact becomes available with time, it is hoped that the case for funding will become easier to make.

The first point to note is the precedence of quality over quantity: 'signing up' a large number of people who do not see it as a real commitment is evidently not the aim, nor is instrumentalising individual change makers as mere conduits for outreach to others. The value of the campaign lies in the genuine understanding and commitment to personal change that it can trigger in ordinary women and men; the aim is to spark such authentic change in the widest possible number. For the purposes of achieving a critical mass, motivating campaigners and measuring the scale of impact, a quantitative target does need to be determined, however. Setting numerical targets for change maker mobilisation will not be an exact science, although more precise modelling may become possible as trends emerge over time. Total population, previous mobilisation successes and the existing outreach of initial and potential alliance members will be factors in that calculation. What is important, however, is that the numbers should be sufficient to constitute a mass movement in the population in question, but not so unrealistically high that it fails to motivate. The target should also have some symbolic value of relevance to violence against women.

In South Asia, we have referred to the statistic of 50 million 'missing' women in setting the target at 5 million change makers over six years: if each one speaks to 10 other people, we reach the symbolic figure of 50 million people influenced. The annual mobilisation forecasts for the region are as follows:

The South Asia alliance makes no claim of intellectual property rights on the 'We Can' model, and anyone is free to learn from it and apply what they feel is relevant in other contexts.

At the same time, the potential impact of the 'We Can' campaign is dependent on two factors: scale and ownership. Selecting change makers, directing their mobilisation activities or disseminating communication materials on their own or as part of a training exercise is therefore unlikely to achieve those results, as is any operation of short duration. Rather than being incorporated into a programme, it may be more effective to run the campaign in parallel.

A separate document sets out what are felt to be the core principles of the 'We Can' approach, which may provide a useful reference when considering the 'fit' between the campaign and planned or existing programmes.

The first people to join the campaign are often those associated very closely with alliance members and/or people who are already taking a stand on issues of rights and equality. Their personal experience and demonstration of the benefits of change tend to provide the initial impetus to change maker mobilisation by helping people to see the possibility of change as a reality.

Some degree of resistance to campaign messages is nonetheless to be expected. Peer support from other change makers is often the primary recourse: individual change makers facing resistance in their families or communities may seek support or advice from others experiencing similar problems. Youth groups or women's groups within the alliance are frequently fora for mutual support. Networking within a given community or institution, which is a specific focus of the second phase of communication, can broaden the scope for this.

The campaign can also assist by providing materials that help specific groups facing particular resistance – such as young people, who often have little voice in the family – to make the case for change in their own circumstances. In addition, by engaging with people of all ages and social groups, 'We Can' generates a further potential basis of support: young change makers may find allies in teachers who have joined the campaign, for example, or in other adult change makers in their community.

In some contexts the whole issue of women's rights and equality is so taboo that an unconsidered dissemination of campaign messages could place change makers and women facing violence at particular risk, and ultimately constrict the space for discussion of the issues. How the campaign operates in such circumstances is a strategic matter, which can only be based on in-depth understanding of the local context. The elements of risk mitigation outlined above will be of particular importance here. In such cases the aims of the campaign will remain unchanged, but it is likely that the communication strategy will need to be adapted to avoid generating unmanageable levels of resistance at the early stages.

Culture, ideology and religion can present formidable bases of resistance because they are tied up with people's sense of identity, and as such tend to invoke strongly protective instincts. To attempt to justify the campaign as consistent with the prevailing cultural, ideological or religious belief system is consequently to engage in an unwinnable and possibly interminable debate that distracts from the fundamental issues. The campaign has therefore sought to avoid engaging in sterile theoretical debate, and instead has focused on people's lived experiences of violence. Communication materials aim to spark recognition of situations people have themselves experienced or witnessed, and reflection on the implications.

Since some forms of violence and discrimination are very closely bound up with cultural, ideological or religious identity, in such contexts it is advisable for the early messaging, graphics and content of communication to steer clear of these more sensitive areas. The initial materials are likely to focus on those aspects that are less controversial, such as the values of respect and family harmony, with the aim of building up sufficient consensus around those before moving on to the implications for equality and non-violent relationships.

While such contexts do present challenges, it is worth noting that the campaign is successfully operating in various difficult situations across South Asia, indicating that it is not impossible to raise awareness of violence and discrimination effectively even in an unfavourable environment.

Promoting recognition that the low-intensity abuses commonly tolerated in people's homes and communities are in fact forms of violence is a key stage in the campaign's communication strategy. This is the point at which people are encouraged to acknowledge violence against women as a problem that relates to them, not something that only affects other families or other sections of society. Without that first critical realisation, personal ownership of campaign goals and corresponding personal change cannot be expected.

Yet the experience of armed conflict in various countries in South Asia has certainly caused the 'minor' violations of slapping, verbal abuse or sexual harassment to seem comparatively insignificant to many. Change makers in Nepal, for example, report that this can make it harder to get the message across that this too is violence, and therefore unacceptable.

In such cases the alliance will need to consider whether the language of the messaging needs to be modified to make sense to people in a given context: for instance, it was decided that the South Asia campaign's message about the 50 million 'missing' women should not be used in Sri Lanka, where it might have led to a sterile debate about the relative importance of this figure and the number of men 'missing' as a result of war. The particular violence that women face in wartime also demands recognition and attention in such cases, and campaign communication will reflect this.

At the same time, it should be noted that the broader definition of violence, encompassing psychological as well as physical and sexual abuse, frequently meets with initial resistance even in places where armed conflict is not a recent phenomenon. Across South Asia, people typically associate violence against women with dowry deaths and honour killings, not with more commonplace violations. As a result, much violence goes unacknowledged and unaddressed. To overcome this resistance, the campaign needs to be speaking about the abuses that women and girls are experiencing in the local context, and thus relating directly to people's lives.

'We Can' focuses on the critical issue of the 50 million women missing from South Asia's population due to discrimination and violence in South Asia. It thus acknowledges that the attitudes and practices that have led to such a crisis are deeply entrenched in every institution of society. Challenging such attitudes and practices clearly entails risks. Recognising and minimising such risks is a crucial part of the campaign strategy.

One aspect of this is the principle of ownership, which implies that people themselves determine the pace of change and the forms of prevention or resistance to violence adopted. No one is under pressure to take action they are not comfortable with. This is inherent in the communication strategy and the role of the change maker – both encouraging reflection, not prescribing solutions.

Since isolated challenges to entrenched attitudes are risky, another aspect is networking at all levels to create a supportive environment. At the individual level, the model of interpersonal mobilisation ensures direct communication and affirmation for each change maker. Locally, change makers are encouraged to come together for mutual support and collective action. And the campaign alliance itself is broad-based, which reduces vulnerability by bringing together a range of forces for social change. Specific efforts to support community leaders and opinion formers in making the case for violence prevention can be an effective reinforcement.

Support services can also help to keep women safe from violence, so disseminating information on service availability, promoting referral, coordination and community linkages, and building capacity in counselling and support can be other ways of managing the risks.

There have been no reports to date of anyone seeking to gain unfair advantage over others by presenting themselves as change makers or otherwise associated with the campaign. While certain aspects of the campaign, aside from its stated aims, may appeal to some people – such as involvement with an international movement, or the social status that may accrue to change makers as a result of their support to others – these are not seen as contrary to 'We Can' aims.

Nor does declaring oneself a change maker confer particular power or authority to act in the name of the campaign or a given alliance member. Moreover, interpersonal mobilisation between neighbours, colleagues and friends and the encouragement to network local people and organisations for mutual support should limit the risk of any masquerading of this kind.

What has happened is that others have challenged the right of change makers to intercede in cases of domestic violence particularly, leading to calls in Nepal and elsewhere for some form of visible identification with the campaign. There seems to be broad acceptance for the idea of some form of ID card bearing the campaign logo and the change maker's pledge, signed by the individual as an undertaking rather than by an alliance organisation taking responsibility for their actions. ID signed by a campaign representative has deliberately not been issued, to avoid misuse. In Bangladesh a cardboard change makers' badge has been adopted; however, this has served more as an entry point for a discussion of the issues than as ID as such.

The campaign refers to the widely accepted UN definition of violence against women: 'Any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm done towards women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life.' Violence against women, being chiefly committed within the context of close personal relationships, presents a specific set of issues; as a subset of gender-based violence it therefore requires a specific response.

Others may prefer to make reference to 'GBV' when addressing the same issues. What the campaign in South Asia has found is that the term 'violence against women' is more commonly understood, as being one that translates into language most ordinary people can readily relate to and recognise in the societies in question.

The term 'change maker' is simple by design: it was chosen as a readily understandable but intentionally broad expression. It seeks to describe, but not prescribe, the role of individuals engaging in the campaign, which is to address the violence in their own attitudes, behaviour, family and community. We did not want to use either social studies jargon ('change agent') or the language of political campaigning ('activist'): while there is a place for both, either might be off-putting for the majority who have not previously seen gender inequality as their problem.

Campaigns in other parts of the world might find another term that better meets these needs in their own context, and we would be interested to learn of them. Certainly a direct translation will not always be the most engaging option, and finding a gender-neutral expression that covers both male and female change makers will doubtless be a challenge in many cases.

While the goal of ending violence against women may seem Utopian to some, it is important to note that 'We Can' offers no prescription for change – no blueprint. Individuals are strongly encouraged to think through, question and interpret the messages of the campaign for themselves, and make the changes that make sense to them. A change maker is thus not an initiate in some exclusive order, but someone making their own choices and encouraging others to do the same.