Paths to social change: conventional politics, violence and nonviolence

A chapter in Nonviolent Alternatives for Social Change, edited by Ralph Summy, in Encyclopedia of Life Support Systems (EOLSS), Developed under the auspices of the UNESCO, Eolss Publishers, Oxford, UK, 2006,

Brian Martin

Go to

Brian Martin's publications on peace, war and nonviolence

Brian Martin's publications

Brian Martin's website

Keywords: Nonviolence, violence, politics, social change, participation, means, ends, suffering


1. Introduction

2. Three Approaches to Social Change
2.1 Conventional Politics
2.2 Violence
2.3 Nonviolence
2.4 Boundaries

3. Track Records
3.1 The Track Record of Conventional Politics
3.2 The Track Record of Violence
3.3 The Track Record of Nonviolence

4. Participation in Social Change
4.1 Participation in Conventional Politics
4.2 Participation in Violence
4.3 Participation in Nonviolence

5. Means and Ends
5.1 Means and Ends in Conventional Politics
5.2 Means and Ends in Violence
5.3 Means and Ends in Nonviolence

6. Suffering
6.1 Suffering and Conventional Politics
6.2 Suffering and Violence
6.3 Suffering and Nonviolence

7. Conclusion



Methods of bringing about social change can be divided into three categories: conventional politics, violence and nonviolence. Conventional politics in turn can be divided into authoritarian, representative and participatory systems. Each of these methods has strengths and weaknesses. These methods are assessed using four criteria. First is their track record, namely how well they have worked in the past. Second is the level of popular participation in the process of change. Third is compatibility between the means to create change and the desirable goal. Fourth is the level of suffering caused by the process of change as well as by the status quo. Of conventional approaches, authoritarian systems are worst in every regard. Representative systems have a better track record and have much greater capacity for self-transformation but also have shortcomings. Participatory systems seem especially good, though the evidence is limited. Of the nonconventional approaches, nonviolence is superior to violence by nearly every criterion.

1. Introduction

There is a need for social change, because society is not perfect. The list of the world's problems is a long one. For example: torture is practiced in many countries; many people live in poverty; discrimination occurs against women, ethnic minorities, people with disabilities, gays and lesbians, and many others; the environment is being seriously damaged; and chemical, biological and nuclear weapons exist, poised to be used.

So change is needed. What sort of change? People have different visions of a desirable world, so it is difficult to gain agreement about what would be a perfect world. Agreement is much easier over what is wrong. Nearly everyone is opposed to torture, exploitation and environmental destruction. Nearly everyone will agree those sorts of things should be reduced or eliminated.

The next question is how. How can torture and exploitation be reduced? How can the environment be protected? At this point there is far less agreement. Agreeing how the world should operate is far easier than agreeing on what to do about it.

People have come up with lots of different approaches. Here are some:

* Don't do anything about the problems. They will fix themselves, through a process of social evolution.

* Pray to a higher power. The higher power will fix the problems (or provide personal salvation).

* People with power will fix the problems due to a sense of responsibility.

* Purify yourself. If each individual becomes pure, the problems will no longer exist, or can be transcended.

Each of these approaches has supporters, but there is not a lot of evidence to back them up. What these approaches have in common is they operate without politics. Politics, in the broadest sense, is the collective exercise of power. The focus here is on political options, in this sense of politics. There are three main approaches.

* Conventional politics, by and through governments.

* Violence, including beatings and killings, and the threat of violence.

* Nonviolent action, including rallies, strikes, boycotts and sit-ins.

In the remainder of this chapter, I examine these three approaches through a series of lenses. First is their track record in challenging and eliminating major problems. Second is the level and type of participation in these approaches. Third is the relationship between means and ends, namely between how change is brought about and the desirable goal. Fourth is the role of suffering. But before beginning this comparison, in the next section the three approaches are outlined.

2. Three Approaches to Social Change

2.1 Conventional Politics

"Conventional politics" means the usual way of doing things in terms of the collective exercise of power. The world today is divided into countries, each under the authority of a government (though in some countries the government has broken down or has little power). The government is the political executive of the state; other components of the state include the police, military and various bureaucracies to handle diplomacy, trade, taxation and other functions.

The economic system has important political dimensions, because it is also about the exercise of power. States normally set up regulations for handling economic affairs. Businesses, especially large corporations, have a strong influence on economic policy. Huge global corporations have an influence on whole countries.

States interact in various ways, and the strongest states - economically, militarily, diplomatically - usually have the largest influence over others, and on the global system as a whole. There are also many international organizations such as the United Nations. At the other end of the scale are local governments and other local groups.

Focusing on the government function of formal decision making, the forms of government can be divided into three main types: authoritarian, representative and participatory. This is a simplification of the actual diversity of political systems, but useful for expository purposes.

Authoritarian governments include military dictatorships, state socialist and fascist systems, and others where rule is by a single individual or group. In authoritarian systems, decisions are made by rulers without any substantive accountability to the wider public.

In representative systems, the top-level political decision makers are chosen by members of the public, typically through elections. Like authoritarian systems, decisions are made by an individual or small group, but those decision makers are formally accountable to the electorate. Note that some authoritarian governments run sham elections, in which votes are falsely counted or where only one candidate is available, to give the appearance of representative government.

In participatory systems, decisions are made by the people who are affected by them. In ancient Athens, the assembly, composed of all male citizens, made decisions for the city. A participatory mechanism today is the referendum, in which all voters choose between options on a ballot, as used in countries such as Switzerland.

There is considerable variation within conventional systems. For example, authoritarian systems can be ruthless dictatorships or have rulers who coexist with significant opposition. Representative systems can use proportional representation or single-member electorates, and may or may not have constitutional protections for human rights.

Using conventional politics to bring about change means operating through the current system. In authoritarian systems, there is often no formal way to do this. In representative systems, the formal method of change is to vote for different political leaders, who in turn will introduce new laws or policies. In participatory systems, new laws or policies can be introduced directly.

A society may have several different systems operating at the same time, in different areas. In countries with representative governments, most corporations are run using authoritarian principles: decisions are made by top executives who are not accountable through elections. Some churches are run on authoritarian principles, such as the Catholic Church in which the Pope has formal power. On the other hand, some congregations make decisions through participatory processes.

2.2 Violence

A second approach to social change is through using violence. At the international level, this includes using military force to threaten or attack another country, defeating the other country's military forces in a war and taking control of the government, corporations and so forth. Within a country, violent change can occur through a military coup, in which a segment of the military takes control of the government.

Another option is use of violence by challengers from a social movement with some degree of popular support, an approach called armed struggle. When the challengers are militarily weak, without a normal army, they typically use "unconventional" military techniques such as harassing raids, an approach called guerrilla warfare. Examples include phases of the American Revolution and the Chinese Revolution. When the challengers become stronger, they may have regular troops that fight government forces in conventional battles.

Another way to use violence is against civilians, a method often called terrorism, though the label "terrorism" is used inconsistently. The largest scale violence against civilians is by governments, for example in wars when cities are bombed. Governments can use violence against civilians as a method of social change. Genocide - the extermination of an entire group, such as the Jews under the control of Nazi Germany - is the most extreme example. Violence against civilians is also used by challengers to governments, such as by the Irish Republican Army.

2.3 Nonviolence

Nonviolent action refers to methods of action that are not violent and that are not conventional politics. Examples include rallies, vigils, ostracism, strikes, work-to-rule, boycotts, sit-ins, fasts and setting up alternative political structures. Nonviolent action can be by an individual, such as a protester who perches in the top of a tree to prevent it being logged, or by groups, such as marches. Nonviolent action can be through physical presence, such as occupation of offices, or through withdrawal, such as when voters boycott an election or workers walk off the job.

There is also a positive side to nonviolent action, including such things as developing neighborhood associations, serving the needs of the poor, promoting harmony between different groups in a community, constructing environmentally friendly buildings, setting up interactive communication systems, and fostering community participation in local decision making. These are all things that help make a community survive and thrive without violence and without domination. Gandhi called this the constructive program.

Nonviolent action can be treated as a set of techniques of struggle, melded together into a campaign. Nonviolence can also be a way of life or, in other words, a philosophy of personal behavior and being. It means living in a way that minimizes harm to others, both avoiding any personal violent behavior and also acting positively to help others and reduce the level of domination in the world. Nonviolence as a way of life is sometimes linked to religious belief.

Many people use methods of nonviolent action because they are effective in achieving their goals. For example, workers may strike to achieve better pay and conditions. This is called the pragmatic orientation to nonviolent action. Others adopt nonviolent methods because of ethical or religious beliefs that life is sacred or violence is evil. An example is a pacifist who believes it is wrong to hurt another person. This is called the principled orientation to nonviolent action. Principled adherents refuse to use violence even when it might be more effective.

In practice, there is a lot of overlap between the pragmatic and principled approaches. In a group of activists, some may support nonviolent action for pragmatic reasons while others have a principled commitment. Furthermore, many principled adherents to nonviolence seek to find methods that are as effective as possible.

2.4 Boundaries

The boundaries between conventional politics, violence and nonviolence are not well defined, and to some extent depend on the circumstances. Consider first the boundary between violence and nonviolence. A normal distinction is to say violence involves physical harm to a person. Nonviolent action does not. At the boundary is harm to physical objects, commonly called sabotage. This includes blowing up empty buildings, smashing the nosecone of a nuclear missile, disabling equipment at a factory, breaking windows, destroying documents, and altering a website. Sometimes this is called violence against property.

In practice, nonviolent activists usually avoid actions that cause massive damage or pose any risk to humans. So setting a forest fire would usually be seen as violence. When the physical damage is low, or the damage is to something that itself is a tool of violence, then it is more likely to be treated as nonviolent action. Examples are deleting a computer file containing names of dissidents to be arrested or destroying the detonators on military explosives.

The boundary between nonviolent action and conventional politics depends on what is considered conventional. In authoritarian systems, a leaflet or petition challenging the government may be treated as subversion; people involved might be arrested or harassed. Therefore, leafleting and petitioning are definitely methods of nonviolent action in such circumstances. In representative systems, leaflets and petitions can become commonplace and accepted as routine and nonthreatening. They therefore become part of conventional politics.

In many countries, workers have the legal right to strike, but sometimes only in tightly regulated conditions. For example, strike pickets might be legally permitted to talk to other workers to encourage them to stay away from the workplace but not the right to block their entry. In such circumstances, legal strikes are conventional politics in a formal sense, but may be so unusual or disruptive they could be classified as nonviolent action. Any violation of regulations makes the strike illegal and therefore more obviously in the category of nonviolent action. For example, a wildcat strike, when workers strike without warning or the involvement of union officials, is definitely nonviolent action.

Working backwards, it is possible to use the response to methods to judge the system. If a few workers put out a leaflet critical of organizational policies, management sometimes responds by criticizing, harassing or even dismissing the workers. When this occurs, it is reasonable to say the organization is operating with authoritarian politics and that a workplace leaflet is a form of nonviolent action, even though the same sort of leaflet, used by a neighborhood group, would be conventional politics.

There is also a boundary between conventional politics and violence. When governments use military force to defend against an armed attack, this is normally treated as conventional politics, whereas aggressive war is not. But when non-government groups use violence, whether in aggression or defense, this is almost never counted as conventional politics. An example is when police mount an armed raid on a household to confiscate goods or arrest people. If members of the household, or neighbors or friends, use any means - violent or nonviolent - to resist the police, this is usually seen as well beyond conventional politics.

3. Track Records

To compare conventional politics, violence and nonviolence as means of social change, one crucial criterion is how well they work: can they actually bring about change widely regarded as beneficial? This question is deceptively simple, because people differ greatly in what they see as beneficial. Today, nearly everyone condemns slavery, but in 1750 it was widely accepted and fiercely defended. Supporters of slavery would not have seen its abolition as beneficial. Similarly, today's peace activists believe nuclear weapons should be abolished, but many people believe nuclear weapons - especially their own country's nuclear weapons - are needed to deter aggression. So to assess the track records of conventional politics, violence and nonviolence, it is easier to look at changes widely accepted today as beneficial, such as abolishing slavery, ending dictatorships and improving the situation of women.

There is an additional complexity: when change occurs, it can be difficult to determine exactly why, for example when conventional politics, violence and nonviolence were all used. So the examples here are not definitive.

3.1 The Track Record of Conventional Politics

It is useful to look separately at the track records of authoritarian, representative and participatory systems. Authoritarian systems, at first glance, might seem to be a problem in themselves rather than a source of beneficial change.   But if a dictator decides on a policy, it is straightforward to implement it. In the Soviet Union - a highly repressive political system - the government greatly improved the situation of women. As well, health care and education were made widely available. Economic inequality was reduced, and all members of the population were protected from poverty.

But authoritarian governments have a terrible record on human freedom. The Soviet government was responsible for the deaths of tens of millions of people through killings and policies that caused mass starvation. Authoritarian regimes around the world deny free speech and free assembly, and use arrest, imprisonment, torture and murder to attack opponents.

Sociologist Deena Weinstein argues that bureaucracy - a form of organization based on hierarchy and the division of labor, and characteristic of most corporations, government departments, churches and other large organizations - is analogous to an authoritarian state, except violence is seldom used. Bureaucracies can accomplish many valuable tasks, usually by implementing the policy of their top managers or external controllers. For example, companies can sell food or produce medical equipment; government departments can run schools or national parks. But bureaucracies can also do great harm, such as sell cigarettes or produce land mines. The extermination programs in Nazi Germany were run by smoothly operating bureaucracies. So it seems reasonable to say that bureaucracy - in other words, authoritarian organization - is not a reliable path to a better world.

Next consider systems of representative government. They have a built-in mechanism for bringing about change: elections. If the mood of voters changes, then candidates may seek election by responding to public opinion. Entire parties may adopt platforms endorsing radical change, and be elected on that basis. Change is also possible through the deliberations of elected politicians, who sometimes change their views as a result of debate over issues.

Social movements, like the peace movement, can affect the climate of opinion, leading to representatives changing their views or feeling forced to shift their policies. Given that social movements are not part of the formal apparatus of representative government, this sort of change could be attributed to participatory dimensions of systems that are formally representative.

Some of the best evidence in support of representative systems is the relative absence of really bad actions. Most of the genocides during the past century occurred under systems of authoritarian rule, including Turkey, Nazi Germany and Rwanda. Representative governments are less commonly responsible for genocide, with the main example being extermination of indigenous peoples in Africa, Australia and the Americas. (Furthermore, representative governments frequently let genocide happen elsewhere without intervening).

People sometimes die of hunger in representative systems, but the most deadly famines in the past century have all occurred in countries with authoritarian rule. Representative systems have a public sphere with open discussion of social issues, usually with independent media. This enables information about impending famine to be freely circulated. Elected officials feel compelled to act, because the consequences of famine would be disastrous electorally. The combination of available information, open debate and electoral accountability seems to be sufficient to ensure that political action against famine always occurs. However, representative governments have often done nothing when famine is occurring in another country.

Representative systems also have successes in bringing about beneficial change. In the 1800s and early 1900s, workers were seriously exploited in many industrializing countries. Socialist and labor parties were formed to champion the cause of the workers. When elected, they introduced worker-friendly legislation. When out of power, they applied pressure to improve conditions for workers: non-labor parties, if they were too blatantly anti-labor, risked losing a future election.

However, labor has been the only challenger constituency strong enough to create political parties - labor or socialist - that have won office nationally. Since the 1960s, green parties have formed to advance the causes of the environment, peace and grassroots democracy. But no green party has ever held power nationally on its own. Instead, peace and environmental goals have been promoted largely through extra-parliamentary action, including both conventional political methods (lobbying, petitions, education) and nonconventional action (rallies, occupations, blockades). Change peace and environmental issues thus has resulted from direct action plus conventional politics, so representative government cannot take all the credit. The same conclusion applies to the issues of women's equality.

There are also many cases in which representative systems have served as brakes on beneficial change. On some issues, the major political parties take the same stand, despite contrary public opinion. In the 1980s in the United States, most of the population favored a freeze on nuclear weapons development, but the government carried on with its nuclear programs. Both major political parties have supported the maintenance and development of nuclear arsenals and have not been responsive to contrary popular pressure.

On such issues, representative systems operate in an authoritarian manner: governments control information through secrecy justified by national secrecy; they suppress critics, especially internal ones; they use propaganda and disinformation to manipulate public opinion; and they use funding to bribe constituencies. This is most obvious during wartime, when civil liberties are curtailed and governments operate in an openly authoritarian way. Furthermore, even when there is not an active fighting war, an enemy may be used as a pretext for emergency powers, most notably during the cold war (1945-1989) and the war on terror, beginning in 2001.

Representative governments are most resistant to change that challenges power systems central to the society. This includes challenges to the state's systems of exercising violence (police and military), challenges to private property (in countries with capitalist economies), challenges to male domination in key sectors in society, and challenges to inequality in decision-making power within organizations.

Representative systems only seldom change their own decision-making features. For example, seldom does a country using preferential voting (in which voters number the candidates in order of preference, with second and following preferences allocated in specified circumstances) change to first-past-the-post voting (in which voters vote for only a single candidate, with the candidate with the highest number of votes winning the election), or vice versa. Although there are many different ways to organize voting, representative bodies, relationships with executive decision makers, and other features of representative government, most systems remain rigidly locked into whatever processes were arrived at decades or centuries ago, often when representative institutions were first introduced.

Participatory systems are uncommon at a national level, so it is hard to evaluate their track record. Of contemporary societies, Switzerland stands out for its participatory political mechanisms, including local autonomy for the regional political units called cantons and use of referenda for decision making. Unlike most other industrialized societies, Switzerland has a citizen-based defense system with very few professional soldiers. It has remained neutral during major wars in the past century. In a referendum, one third of Swiss citizens voted to abolish the army, not enough to pass. But such a referendum is almost inconceivable in other countries, and suggests the potential for change in a more participatory system. However, Switzerland is not noted for progressive stands on labor, environmental or feminist issues.

Participatory systems are far more common at the level of organizations and networks and have a good track record for bringing about beneficial change. A few businesses - especially those called cooperatives - involve workers in making key decisions about how to do the work and sometimes in making wider decisions about wages, products, marketing, expansion and investment. These businesses are usually the most responsive to concerns about worker safety, environmental impact and producing goods that benefit society.

Quite a few environmental, feminist and other activist groups use consensus decision making (sometimes in a highly formal process, sometimes on a more casual basis). These groups are dedicated to bringing about social change; participatory processes seem most common in activist groups.

Participatory groups are constrained by the systems in which they operate. A group of soldiers might make decisions on a consensus basis - in contrast to the usual command system - but still operate within the parameters established by higher-level commanders. Participatory businesses usually have to compete in the marketplace with traditional authoritarian businesses. But there is another alternative: establishing systems based on different principles. For example, the use of a local currency - an alternative to money authorized by the national government - fosters a more participatory local economic system. Free software is produced by volunteers in a process far more participatory than for-profit software produced in companies.

To summarize, the track record of conventional politics is quite mixed. Authoritarian systems are responsible for most of the worst abuses and do not contain internal mechanisms for ensuring beneficial outcomes or for self-transformation. Representative systems are much more responsive to emerging concerns, but still have significant rigidities, especially when major political parties adopt similar policies and when it comes to reforming the representative system itself. Participatory systems are the most responsive to social concerns; indeed, activists who are campaigning for change often prefer to use participatory processes.

Looking at track record alone, it could be said authoritarian systems are usually a cause of social problems, not the solution. Representative systems have many capacities for responding to social problems, but also many rigidities. Participatory systems are the most responsive.

3.2 The Track Record of Violence

Violence has brought about change through various routes, including war, military coups, people's armed struggle, and terrorism. Each of these will be considered here.

Wars can be the source of profound change, causing massive death, changes in systems of government, and sweeping social change. Offensive or aggressive war is most commonly launched by authoritarian governments and, if successful, leads to an expansion of authoritarian rule, such as the conquest of most of Europe by Nazi Germany during World War II. This was disastrous rather than beneficial.

Sometimes an authoritarian government launches a war, loses and is consequently undermined. The military government of Argentina tried to conquer the Falkland Islands (also called the Malvinas) in 1982 but was defeated by Britain; the Argentine government was replaced the next year with an elected one. But this is hardly a prescription for beneficial change, but rather a beneficial outcome from a miscalculation by rulers.

Sometimes representative governments launch attacks on other countries. The US government fought in Vietnam for over a decade but was eventually defeated in 1975. The US government has launched attacks on many countries, for example Grenada in 1983, Panama in 1989 and Iraq in 2003, and helped instigate coups in many others, such as Iran in 1953 and Chile in 1973. The record from these and many other such wars and coups is not inspiring. In every case, predictably, the attacking government - the US or some other aggressor - justifies its actions as defending against a greater evil or promoting beneficial change. But an examination of these countries attacked, before and after, does not reveal any shining examples of liberation from oppression. If such examples existed, they would be well advertised. The best examples of beneficial change through war are Japan, Germany and Italy after defeat in World War II, but this was a war initiated far more by them than by their opponents.

Military coups bring about changes in rulers but not necessarily profound social changes: often they replace one authoritarian ruler with another or, worse, replace an elected leader with a military dictator. But occasionally a coup brings about beneficial change, such as the 1974 coup that overthrew the fascist regime in Portugal.

Violence has been used frequently to bring about change from the grassroots. The most famous cases are revolutions, including the American, French, Russian and Chinese revolutions. In each case, an oppressive regime was overthrown through people's action, including organized violence. The French and Russian revolutions were quick, without lengthy armed struggles, and the American revolutionary war was preceded by a decade of powerful nonviolent action. How beneficial these revolutions were is another question. For now, the key observation here is that violence played a critical role in some important revolutions, especially those with protracted armed struggles as in China.

Successful revolutions are rare: there are many failed attempts at armed liberation. Furthermore, some of these attempts make society much worse. For example, the Tupamaros challenged the elected government in Uruguay in the 1960s, including through urban guerrilla actions. This triggered a repressive response, leading to a transition from representative to authoritarian government.

The record of armed struggle in industrialized countries is one of repeated failure, without a single clear success. Urban guerrillas - the equivalent of rebel soldiers, in an urban setting - are normally labeled terrorists, whether they target soldiers, police or civilians. Examples include the Red Brigades in Italy, the Baader-Meinhof Group in West Germany and the Weatherpeople in the US, each prominent in the 1960s and 1970s. Whatever the label, this approach has never generated significant support but instead has stimulated an increase in the repressive powers of the state.

For social movements operating in countries with representative governments, any use of violence is likely to be counterproductive, providing a pretext for the government to use force against the movement. This is the reason police sometimes send undercover agents to join movements and encourage the use of violence, thereby discrediting the movement. Furthermore, governments regularly say movements are violent, even when they are not. This is strong evidence that violence is not an effective means of social change in representative systems.

In many campaigns in representative systems, violence is used to a limited degree or not at all: it is hard to find an issue in which violence was a crucial tool in bringing about change. For example, in protests against corporate globalization, such as in Seattle, USA in 1999, most participants have been nonviolent, but a few have smashed shop windows, thrown rocks at police and caused other physical damage, though never yet killed any police or civilians.   The value of this limited violence is debatable; there is no evidence it has provided a significant boost for the campaigns.

Some animal liberation activists have damaged goods such as fur coats, destroyed laboratories using animals for testing, and "liberated" animals from them. Some environmental activists have put spikes in trees (to damage logging equipment), damaged tractors, defaced billboards and destroyed off-road vehicles. Opponents of genetic engineering have destroyed plots of experimental crops. There is considerable debate within these movements as to whether these tactics are worthwhile or counterproductive. In any case, they are at the boundary between nonviolent and violent action. In nearly all cases, they are violence against physical objects or against non-human life (genetically engineered crops), not against humans, and can be classified in the nonviolent category of sabotage.

Violence against people, as a tool within or against representative systems, cannot claim any major successes, even as an important contributor to success. The 1995 bombing of an office building in Oklahoma City did nothing to help the cause of the militia movement opposed to abuses by the US government. The 11 September 2001 attacks on the Trade Towers and the Pentagon had the effect of greatly increasing popular support for the US government. There is not a single unambiguous case where violence - beyond sabotage that carefully avoids danger to humans - has been a key factor in bringing about beneficial change within representative systems.

3.3 The Track Record of Nonviolence

Nonviolent action has been widely used to oppose repressive regimes, with both successes and failures. The most obvious successes are when these regimes are toppled, with rulers resigning. Consider East Germany in 1989. After the government of Hungary opened its border with West Germany, allowing East Germans a route for emigration, large numbers soon left for West Germany. At the same time, some anti-government rallies began, before long swelling in size to hundreds of thousands. East German leaders had a well-developed repression apparatus of spies, police and troops, but in the face of emigration and escalating protest, they decided to resign rather than use force. This scenario was repeated throughout most of Eastern Europe in what are called the velvet revolutions.

For these changes to occur, the political conditions had to be right. President Gorbachev had signaled that the Soviet government would no longer intervene militarily to maintain the rule of Communist parties in Eastern Europe. This wider geopolitical realignment was crucial to the 1989 revolutions. Nevertheless, it is fair to say nonviolent action was a crucial factor. In most Eastern European countries, there was no armed struggle against the regimes - insurgent violence was not the cause of the change. Nor was there any military action or serious threat from the west.

Was the process done through conventional politics? In a sense, yes, because the East German ruling elite decided voluntarily to vacate government. But if there had been no emigration or protests, the ruling elite almost certainly would have remained in their positions. Consider, for example, the Communist governments in Cuba since 1960 and Vietnam since 1975: they have remained in power despite not having a reliable protector like the Soviet government in Eastern Europe from 1945 to 1989.

For any change to occur, the conditions have to be favorable. This applies to conventional politics, violence and nonviolence, though the most favorable conditions for these three options may be different. The point of looking at track records is to assess successes and failures. If a method is repeatedly unsuccessful - like revolutionary violence in industrialized societies - then it makes sense to say the method is flawed or the conditions are never right, which lead to the same outcome.

There are many other successes of nonviolent action against authoritarian governments. Examples include:

* the decades-long struggle in India for independence from British colonial rule, achieving success in 1947;

* the toppling of the Marcos dictatorship in the Philippines in 1986 through massive rallies, a dramatic process called "people power";

* the ending of the system of white rule in South Africa called apartheid in the early 1990s;

* the forced resignation of Indonesian President Suharto in 1998, leading to the introduction of representative government;

* the ending of the rule of President Slobodan Milosevic of Serbia in 2000 after months of protests mobilizing ever greater numbers of people;

* the overthrow, through nonviolent insurrection, of dozens of authoritarian regimes in Africa and Latin America from the 1940s onwards.

Nonviolent action has also been used against coups. Prominent successes include:

* popular resistance to the Kapp Putsch in Germany in 1920, which featured strikes and noncooperation by civil servants to the new rulers, leading the coup to collapse in four days;

* opposition to the 1961 Generals' Revolt in Algeria, with protests in France and noncooperation by French troops in Algeria, many of whom stayed in their barracks or flew airplanes out of the country, causing the coup to collapse within days;

* popular resistance to a coup in the Soviet Union in 1991, leading to its failure and the break-up of the country into a confederation.

There are also failures of nonviolent action against authoritarian regimes. The most famous is the crushing of the Chinese pro-democracy movement in 1989. Another prominent example is the lengthy popular struggle in Burma against its repressive government, a struggle that so far has been unsuccessful. But it should be noted that no other method has been successful in China or Burma: conventional politics is not responsive and armed insurgency in Burma has had no success.

In a number of cases, both violence and nonviolence have been used against authoritarian rule. The Palestine Liberation Organization used terrorism to challenge Israeli rule in the occupied territories, gaining international attention but alienating many potential supporters. Then in 1987, the first intifada - uprising - spontaneously erupted. It was mostly nonviolent, with protests, strikes, boycotts and other forms of noncooperation, but with some minor violence such as throwing of stones, and is best described as "unarmed." The intifada generated far more support for the Palestinian cause, in Palestine, Israel and around the world, than terrorism ever had. However, the second intifada, beginning in 2000, has used violence, most prominently in the form of suicide bombings, and has had far less success in attracting international support.

East Timor was invaded and occupied by Indonesian military forces in 1975. There was an armed guerrilla resistance, but it could not dislodge the Indonesian occupation. In the late 1980s, the resistance changed tactics, emphasizing nonviolent protest in the cities, especially during visits by foreign dignitaries. This approach had far more success in building support within the East Timorese population. After the 1991 massacre of East Timorese civilians during a funeral procession in the capital, Dili, international support for the cause of East Timor's independence greatly increased.

In South Africa, nonviolence was used from 1906 to 1914, under the leadership of Gandhi, with some success against racial discrimination but with much still to do. Later, in the 1950s, nonviolence was used against the white government, but without a lot of progress. In its apartheid system, South Africa had a representative government, but only whites were allowed to vote, while the majority black population was treated harshly and, in essence, lived under authoritarian rule. Some black opponents then adopted armed struggle from the 1960s, but this was easily crushed by the powerful South African military and police forces. In the 1980s, the black resistance switched to primarily nonviolent means, now aided by international diplomatic and economic pressure. This was eventually successful in bringing about a peaceful end to apartheid.

In the southern states of the United States, slavery was abolished in 1865 but serious racial discrimination continued in a system called segregation. Although the federal government passed laws against discrimination, they were widely flouted by southern state governments and by white segregationists. Segregation was backed up by physical intimidation, including lynchings which of course were illegal - they were murder - but almost never penalized. Blacks were discouraged or prevented from voting, so electoral politics did not offer a ready solution to segregation. In 1954, the US Supreme Court ruled against segregated schools, but this ruling was not implemented. So conventional politics, via laws and elections, was ineffective against segregation.

Blacks seldom used violence to further their cause, and when they did it was almost always counterproductive. In East St Louis in 1917, limited violence by a few black men, in defense against attack, triggered a white riot that destroyed hundreds of black homes and killed dozens of black people. Armed struggle was not seen or supported as a viable strategy against segregation.

Nonviolence, in contrast, was highly effective. In 1955, a black boycott of buses - in which seating was segregated - was launched in Montgomery, Alabama. This nonviolent rebellion against segregation, called the civil rights movement, attracted enormous support from black communities and from some whites in following years with numerous rallies, boycotts of segregated services, sit-ins at segregated restaurants, and "freedom rides" (bus tours by integrated groups of blacks and whites).

The movement's most prominent leader was Martin Luther King, Jr., who was assassinated in 1968, triggering riots in black communities in many cities. In the 1960s, the civil rights movement was challenged by an emerging black power movement, whose members took a more overtly aggressive stance, some of them carrying arms. But black power was soon crushed following surveillance, disruption, arrests and killings by police forces. In so far as black power even advocated violence - in practice, violence was seldom used - it was ineffective in improving the position of the black minority.

Conventional politics had been largely ineffectual against segregation for decades, but in 1965 the Civil Right Act was passed by Congress, institutionalizing major changes. This act gained support because the civil rights movement, through nonviolent action, had generated support among voters and politicians. So it can be said conventional politics had been unresponsive or ineffectual against segregation but was prodded into action by the civil rights movement.

There are many other examples in which nonviolent action has been used, often effectively, in societies with representative systems, including campaigns by feminists, environmentalists, gay and lesbian activists, students, ethnic minorities, homeless people, and people with disabilities. Typically, nonviolent action is used when conventional politics is not responsive. Indeed, many members of social movements prefer to use official channels when they are available, for example by lobbying politicians, participating in formal inquiries, and campaigning for candidates supportive of their cause. But in many cases these sorts of efforts are inadequate or futile, for example when all major political parties are aligned on an issue. In these circumstances, nonviolent action is used as a way to demonstrate strongly held views, to put an issue on the agenda, to increase economic or political costs for the contrary policy, and to prevent implementation of current policy.

Some of the issues where nonviolent action has been an important component of campaigning in systems with representative government are opposition to testing of nuclear weapons, opposition to nuclear power, opposition to whaling, support for gay and lesbian rights, support for student participation in university decision making, opposition to supersonic transport aircraft, opposition to abortion and support for animal rights. Often it is difficult to separate the contributions of nonviolent action and conventional politics in these campaigns, but it is safe to say that in many cases campaigns would have had little success without nonviolent action.

In summary, nonviolent action has played a major role in opposing authoritarian regimes and their abuses, including in quite a number of dramatic overthrows of such regimes. Nonviolent action is not always successful in or against authoritarian systems, but no approach is uniformly successful. Nonviolent action also has played a key role in many movements within representative systems, often as a way of putting issues on the agenda for conventional politics or for increasing support for unpopular causes.

4. Participation in Social Change

Participation in social change refers to people being actively involved in promoting and deciding on social change. Of course, if social change occurs, then people can be said to participate in it, but that is a passive form of participation, equivalent to being a passenger in a bus in which the driver makes all the decisions.

High participation is important for three main reasons. First, when lots of people participate in social processes, there is a much greater chance they will move in socially beneficial directions, because a small group cannot carry the day over the opposition of the majority. Second, participation is a valuable experience in itself, found by most people to be satisfying. In other words, increased participation is itself socially beneficial, largely independently of the actual decisions made and directions taken. Third, participation gives experience to those involved that can improve the quality of future participation: ongoing change processes benefit from experienced participants.

With these observations, I will look briefly at levels and quality of participation using the three approaches to social change examined here, namely conventional politics, violence and nonviolent action.

4.1 Participation in Conventional Politics

In authoritarian systems, a single ruler or a small ruling group makes key decisions and the wider population is expected to obey, perhaps with token opportunities to contribute to decision making. The connection between authoritarian politics and low participation is such a reliable relationship that it would be possible to use the level of participation in a political system to judge how authoritarian it is. Nearly all authoritarian leaders are men, almost always from the dominant ethnic, religious or occupational group. Thus they come from only a narrow segment of society.

In representative systems, there are two main levels of participation. Elected officials participate intensively in decision making, often as a full-time job, dealing with a host of issues. In most societies with representative government, elected officials are mostly men, typically in their 40s or older, and commonly drawn from a restricted set of prior occupations (such as lawyers and political party officials, with very few nurses or truck drivers).

Most citizens who are not elected officials have a limited role in formal decision making. They may vote for candidates at regular intervals, though many choose not to vote. A few are involved more heavily, for example in campaigning at election time or writing letters about issues. Representative systems thus have a system of dual participation, very high for a small number and quite low for most of the rest.

In participatory systems - as the name suggests - nearly everyone can participate in decision making if they want to. For example, any member of the electorate can vote in a referendum. In a participatory workplace, workers decide on important matters, which may include working hours, jobs, products, and investment. Anyone who is a member of a relevant group can participate - but not non-members. In some participatory schools, students make key decisions about how they learn: age, sex and ethnicity are not inherent barriers to participation.

4.2 Participation in Violence

Most soldiers are young fit men. So are most guerrilla fighters and terrorists. Women are very underrepresented in most forms of violence (though this is changing somewhat in professional military forces). Children and old people are similarly underrepresented, and people with disabilities are rare.

Decisions about violent actions are usually made by an individual or small group. In armies, commanders make the crucial decisions; soldiers are expected to carry out the commands. When violence is used to challenge governments, planning is carried out in secret, in order to prevent the other side having an advantage.

Given that most participants in violence are drawn from a restricted segment of the population (young fit men) and that decisions about violence are made by a small number of people, it can be said that violence, so far as participation is concerned, is usually aligned with authoritarian politics.

4.3 Participation in Nonviolent Action

Many forms of nonviolent action can be said to be participatory systems. For example, just about anyone can join a rally or a boycott. If workers hold a stay-at-home strike or a work-to-rule campaign, every worker who wants to can join in.

In a nonviolent campaign using a variety of methods, there are usually opportunities for just about anyone to participate, including women, children, the elderly and people with disabilities. Some particular types of action - such as canoeing into the path of a nuclear vessel - require special skills and strengths, but in many such actions there are parallel opportunities for others, such as an on-shore rally. People with disabilities may not be able to join in some activities, depending on their disability, but usually there are ways they can participate, for example in boycotts.

Because nonviolent action is so often used to support the efforts of those who have been excluded, participation by those who are less powerful is often greater: women predominate in feminist actions; rank-and-file workers are far more likely to join a strike than are executives.

A related issue is the level of participation in making decisions about organizing nonviolent actions and campaigns. Some actions are organized by a small group; others involve a more participatory decision-making process. Many nonviolent activists are personally committed to fostering participation. It is both difficult and unusual for people to be coerced into joining nonviolent actions. Nonviolent action tends to be linked to participatory systems.

5. Means and Ends

It is often possible to distinguish between the means to achieve something and the goal to be achieved, namely the end. Sometimes these are compatible, such as the person who seeks to promote generosity by being personally generous - also thereby setting a good example. But in other cases the means and ends are incompatible or at cross-purposes, such as the parent who shouts at children to be quiet, or a government seeking world peace by building nuclear weapons.

The relationship between means and ends is often unclear, and in some cases it is difficult to figure out what the means and ends are. Nevertheless, it can be useful to examine means and ends, for two main reasons. First, when means are contrary to ends, advocates are subject to the charge of hypocrisy, which can undermine change efforts. Often, as in the case of domineering parents, actions speak louder than words. Second, when means and ends are compatible, the process of social change is safer: even if the final goal is not achieved, the steps taken are in the right direction. Means-ends compatibility provides a sort of fail-safe mechanism to prevent things going horribly wrong.

5.1 Means and Ends in Conventional Politics

In authoritarian systems, the means are authoritarian, namely government edict or bureaucratic imposition. Authoritarian means are contrary to what most people believe is a desirable society, namely one with freedom and justice. A dictatorship might negotiate a peace treaty, but if dissent is crushed, this is not a good sign for sustained non-aggression. Leaders of a large company might promote participatory teams, but if implementation is from the top, with executives maintaining ultimate power, the level of participation may prove illusory or transitory.

In representative systems, the means for choosing governments are political parties and elections and the means for governments to make decisions are through legislatures, executives and bureaucracies. Only some decision-makers are actually elected. High officials in government departments and courts are usually appointed. So representative systems are only representative to a limited degree, and often operate using authoritarian means, plus a few participatory mechanisms. One important non-representative system is the military: commanders are not elected by the troops. Nor, usually, are university officials elected by staff and students, hospital administrators by doctors and patients, or corporate leaders by workers and customers.

Because representation has such a limited domain, it can be difficult to separate the connections between representative means and ends from authoritarian sub-systems. It seems, though, that representative governments very seldom make an attempt to increase the use of representation into arenas outside of elected government officials. For the most part, there is a disjunction between voting as a means and a desirable policy as a goal: the goal, perhaps environmental protection, has no particular connection to voting as a means, because voting is an indirect method of bringing about change.

In participatory systems, the fit between means and ends can be fairly good, especially when the goal is something to do with participation itself. An example is a women's collective that seeks to empower women and uses consensus decision making.

5.2 Means and Ends in Violence

Violence is normally considered a means to an end rather than an end in itself. Very few people desire a world with high levels of personal violence or war. Those who use or threaten violence are usually saying, in effect, "do as I say, not as I do." This applies to parents who beat their children for getting into fights and to governments that maintain powerful militaries to protect the peace. The hypocrisy is sometimes striking, as in the case of a nuclear weapons state whose leader becomes indignant when another government seeks nuclear arms, or even goes to war to stop another government developing nuclear weapons.

One of the disadvantages in using violence for beneficial change is that if the effort does not succeed, all that has been accomplished is negative, namely the violence and its immediate consequences itself (killing, destruction, disruption) and, more generally, training in and acculturation to violence. Violence is the best illustration of the dangers of using means contrary to the ends.

Looking at means and ends is a useful antidote to the usual approach to violence, which is to assess whether it is or can be justified. William T. Vollmann - one of the few scholars presenting the case for violence - in his massive seven volumes Rising Up and Rising Down devotes the bulk of his attention to whether violence can be justified. But just because something can be justified does not mean it is the best option. This is especially true when looking at options at a collective, social-change level. Protesters might be justified in coming to a peace rally with metal bars or guns to defend themselves from attack, but the contrast between means and ends would be highly damaging to their cause.

5.3 Means and Ends in Nonviolence

When challenging the corporal punishment, torture or the war system, nonviolence has a near-perfect meshing of means and ends. As the saying goes, nonviolence is the means and nonviolence is the goal. In other words, the aim is to use methods of nonviolent action to bring about a world without violence, whether the violence is corporal punishment, torture or war. More generally, the promotion of nonviolence itself, as a tool of social transformation, is carried out without violence.

6. Suffering

Suffering includes the human pain and sorrow associated with death, injury, degradation, poverty and exploitation, and can be extended to damage to the environment and non-human animals. The process of bringing about social change often has a cost in the form of suffering - but so does the mere persistence of the status quo, which is always less than ideal. In terms of suffering, an effort to bring about change could be judged worthwhile, in utilitarian terms, if total suffering is reduced: the suffering caused by the change process would have to be less than the reduction in suffering over the long term as a result of the change.

If the current level of suffering is high, then change agents might judge that campaigns are worthwhile even if suffering greatly increases for a while, because the long-term benefits are so great. This is a rationale behind many efforts at revolution. But this calculation is only valid if the change effort is successful. If revolutionaries fail, they have caused additional suffering for no benefit. Likewise, there may be no net benefit in a revolution that leads to new forms of oppression. So, in terms of suffering, promoting change is a calculated risk, with the hope that the chance of lowered suffering, times the amount, is greater than the corresponding figure for increased suffering.

This calculus of suffering needs to take into account the track record of previous attempts. Of course the current situation may be different, but past events give insight into likely consequences. Part of the calculation should take into account the relation between means and ends: when they are compatible, there is far less risk that failure to achieve all the intended goals will cause excess suffering. A final consideration is that there is more to social change than suffering. Beneficial social change can be considered to be a value in itself, providing both experience in bringing about change - a learning process with long-term benefits - and inspiration and hope, essential qualities for the human experience.

Determining the amount of suffering in the status quo is difficult enough; estimating the level of suffering in a hypothetical new society is even more challenging. Supporters of the status quo will assess these matters very differently from reformers and revolutionaries. In discussing suffering here, the main aim is to introduce some key issues rather than make definitive statements.

6.1 Suffering and Conventional Politics

Authoritarian systems are often responsible for intense suffering. In the 1900s, three authoritarian systems caused the greatest number of deaths. The Nazi government in Germany killed perhaps 12 million people, half of them Jews, in death camps and massacres of civilians, and many others through its military operations in World War II. The Soviet government killed millions through direct purges and tens of millions more through starvation due to its policies. The policies of the Chinese communist government killed tens of millions through avoidable famine in the late 1950s, plus millions more through repressive policies.

That authoritarian governments cause suffering is well known, though often the scale is hidden from outsiders. Conventional politics within authoritarian regimes does not have a built-in process for reform, so the key question is whether efforts for change - via violence or nonviolent action - can succeed in reducing this suffering.

Authoritarian systems within organizations - namely bureaucracy - seldom cause death overtly, but they can cause suffering via alienation, bullying and humiliation. The daily life of many workers is far less satisfying than it could be, and authoritarian organizations contribute to this shortfall.

Representative systems are less likely to cause massive death. But there is still much suffering, including that due to poverty, discrimination, exploitation and alienation. Furthermore, representative governments sometimes wage destructive wars, such as the US war in Indochina, which caused some two million Vietnamese deaths. Representative governments are the key proponents of neoliberal economic policies, which critics say cause massive death and poverty in poor countries. Representative governments often support authoritarian regimes, such as when most western governments supported the Suharto regime in Indonesia (1965 to 1998), which was responsible for killing up to a million people in 1965-1966 and hundreds of thousands more later in East Timor. Several countries with systems of representative government - including the US, Britain and France - have significant arsenals of nuclear weapons which, if unleashed, could cause untold millions of deaths.

So a case can be made for change in many representative systems. The question is whether internal processes of representation provide the tools for making the changes. In many cases they apparently do not. In the examples cited - the Indochina war, neoliberal globalization, support for authoritarian regimes and nuclear arsenals - all major political parties have supported the policies, differing only in details rather than the basic thrust, even when the majority of the population has been opposed. So to bring about change, some challenge outside of the formal systems of elections appears to be needed.

Another shortcoming of representative systems is they cannot be relied upon to oppose authoritarian systems, such as repressive foreign regimes. Part of the problem in representative systems is that internal subunits are run on authoritarian lines, including government departments, corporations and some political parties.

It is unclear whether participatory systems could lead to major suffering, because there is not enough experience with them. There are no examples of referendums for launching an aggressive war or examples of enterprises run by the workers deciding to produce and market dangerous products. Participation seems to offer protection against seriously damaging decisions, though certainly not a guarantee.

6.2 Suffering and Violence

Violence obviously causes suffering in the short term. But it can reduce suffering in the long term when it quickly destroys an oppressive system and all its attendant damage. Critics would argue nonviolence was a very slow means to achieve independence in India: violence might have caused more deaths in the short term but by bringing independence sooner, reduced suffering in the long term. However, most examples show only the horrific suffering caused by violence and do not give much evidence of any long-term reduction in suffering.

The British as colonial rulers did not use much violence against nonviolent protesters in India. But they reacted very differently in Kenya, where the pro-independence Mau Mau used violent means. The British rulers set up numerous concentration camps, used torture systematically, executed a thousand prisoners and killed tens of thousands of villagers in crushing the rebellion. The Mau Mau, by using violence, seemed to legitimate counter-violence by the British, a dynamic that did not occur in India.

Armed struggles for independence from colonial rulers reveal a horrific death toll. In Vietnam, armed struggle against the French from 1945-1954 and against the US until 1975 led to more than two million Vietnamese deaths. In Algeria, the armed struggle against French rulers resulted in up to a million Algerian deaths.

There are some examples in which armed struggle appears to have overthrown a regime with relatively low loss of life, such as the 1917 Russian revolution and the overthrow of the Batista dictatorship in Cuba in 1959. In both these cases, the regime mainly collapsed internally; in other words, violence was not the key to regime change but rather was part of the articulation of resistance, which included unarmed noncooperation.

In Europe occupied by the Nazis, partisan resisters used armed attacks and sabotage. The Nazis responded with brutal reprisals, typically killing ten civilians for every one of their own soldiers killed. The partisan resistance had little impact on Nazi rule, which was ended by Allied military violence. (On the other hand, nonviolent resistance caused the Nazis more difficulties.)

In representative systems, violence has been used by urban guerrillas, for example in Italy, Germany, US and Uruguay. Violence in these cases has led to suffering by those targeted by the guerrillas and by government responses. There is no evidence from any such campaigns that suffering in the long term has been reduced.

In summary, violence as a tool for social change always causes immediate suffering, sometimes on a horrific scale due to counter-violence. In principle, violence-induced change can reduce suffering in the long term, but there seem to be few clear-cut cases where this has been the result.

6.3 Suffering and Nonviolence

Nonviolent action, by its very nature, limits the amount of direct suffering caused. By definition, no physical violence is done to opponents. However, boycotts and strikes can cause economic hardship and ostracism can cause emotional suffering.

Another important consideration is whether nonviolence, as a strategy, is an effective way of reducing suffering, namely in bringing about a social change that reduces suffering over the long term. The classic case is colonialism, in which colonized peoples were exploited and sometimes subjected to horrible brutalities, such as Belgian rule in the Congo in which millions died. The classic nonviolent struggle against colonial rule was in India, under the leadership of Gandhi from 1915 until independence in 1947. The number of people killed in the course of this massive struggle was quite low, a matter of hundreds (these were almost entirely Indians killed by the British, not vice versa).

Nonviolence, when used against repressive rulers, has sometimes succeeded and sometimes failed, but in both cases nonviolence seems to have kept immediate suffering to a relatively low level. For example, nonviolent resistance to the Nazis in occupied Europe had some successes, for example preventing the teaching of Nazi doctrine in Norway and protecting Jews in Denmark. These efforts did not lead to major reprisals by the Nazis. The most famous reprisal against nonviolent activists was in Beijing in 1989, when the Chinese government killed hundreds of pro-democracy protesters following demonstrations in Tiananmen Square.   Even so, this was a small death toll compared to what routinely occurs in armed struggles.

7. Conclusion

Three main options for social change have been examined here: (1) conventional politics (divided into authoritarian, representative and participatory systems), (2) violence and (3) nonviolence. These options have been examined through four lenses: their track records; participation; means and ends; and suffering. A summary of the main conclusions is given in the table.

Table 1. Options for Social Change Assessed by Four Criteria

Option Track Record Participation Means/ends Suffering
Conventional politics: authoritarian Extremely limited capacity for internally driven improvement Highly restricted Low compatibility for progressive goals High
Conventional politics: representative Good capacity for self-transformation except when key power groups are challenged Imbalanced among decision makers; limited among citizens Limited compatibility Limited for some situations, sometimes high for others
Conventional politics: participatory Good capacity for self-transformation (but limited evidence) High Good compatibility, especially for participatory purposes Low (but limited evidence)
Violence Some successes against authoritarian systems, few in or against representative systems Limited, mainly to young fit men Low compatibility High
Nonviolence Many successes in all systems High High compatibility, especially for peace and conflict-related   purposes Limited

Overall, authoritarian systems are disastrous in themselves and have little capacity, through their own conventional politics, for beneficial change. Representative systems have a far better track record and have considerable self-transformative capacity, but only in certain areas. Representative systems are not reliable allies against abuses elsewhere: they often tolerate authoritarianism, both in other countries and internally within organizations. Participative systems are far better in themselves and for self-transformation, though experience with such systems on a large scale is comparatively limited. Of the two main approaches to change outside of conventional politics, nonviolent action is preferable in nearly every regard: it has a better track record, high participation, compatible means and ends, and tends to limit the level of suffering.

Given this conclusion, it may seem surprising that nonviolent action has such a low profile. There are several reasons for this. One is that most people grow up and live in societies in which authoritarian or representative politics are normal. Whatever is normal can come to seem inevitable. Students are taught how the system is supposed to work and the media give maximum attention to conventional routines, such as decisions by government leaders or election campaigning. Unconventional methods - violent or nonviolent - are commonly portrayed as transgressive, and often condemned. People, when they have a problem, are encouraged to use the system, for example to make a complaint to an official body, to contact a politician, or to go to court. They are not often encouraged to set up a personal picket, much less to flourish a handgun.

Another important reason why conventional politics is seen as the solution to problems is most people in positions of power in such systems achieved their power, and exercise it, through conventional processes. Therefore they have a stake in maintaining the system. Even those who come to power by unconventional means, such as a military coup, normally set about making their own power conventional. In short, conventional politics is the preferred tool of those with power. Little encouragement is given to modes of challenge that shake up current power structures.

Yet there remains a curious phenomenon: violence, with its abysmal track record, low participation, incompatibility between means and ends, and high suffering, is seen by many challengers as preferable to nonviolence. There are several reasons for this. One is a common but false belief that violence is necessarily superior in a direct contest with nonviolence, an assumption played out in histories of war and commentaries on contemporary events. Part of the problem is that understandings about nonviolent action are relatively new, with most of the important thinking stemming from Gandhi, his interpreters and his successors, in the past century, whereas violence has been familiar for millennia.

Another factor is that violence is a regular tool for conventional politics, in the form of police and military forces. When used by governments, violence becomes conventional and seems normal.

Another reason behind a preference for violence is that it privileges men, who are far more likely to use violence than women, and privileges the leaders of violent movements. Leaders of successful armed struggles are in a prime position to be rulers, often in authoritarian systems. In contrast, nonviolent campaigns are more likely to lead to representative or participatory outcomes, with less privilege for the leaders. Therefore it is predictable that men seeking personal power are more likely to be drawn to violence, whereas men and women with more altruistic motivations are more likely to be drawn to nonviolence. A famous example is Gandhi, who declined to be the first president of India.

Given that many people do not recognize the power of nonviolence, it is not surprising there are some common arguments used to dismiss nonviolence. It is revealing to examine these, because they show the way double standards and selective use of evidence are used to reach a preferred conclusion.

A common claim is that nonviolence only works against kind-hearted opponents, such as the British in India. It is true the British did not often use heavy-handed violence against nonviolent opponents, but the reason is that when they did use such violence, such as beatings of nonviolent protesters in 1930 during the salt satyagraha (campaign against the British salt monopoly), it rebounded against the British, generating much greater support for the independence campaign. The British were quite capable of using extreme violence, such as against the independence movement in Kenya, as noted earlier. If the British seemed relatively kind-hearted in India, it was in large part precisely because the independence movement was nonviolent.

The historical record includes many examples of the effectiveness of nonviolence against repressive rulers, such as successful nonviolent insurrections against dozens of dictatorships in Africa and Latin America. One of the most dramatic campaigns was against the Shah of Iran, who had a powerful military, used torture against opponents, and whose regime was supported by the governments of Israel, the United States and the Soviet Union, among others. The Shah's troops gunned down unarmed protesters, killing tens of thousands. Yet in 1979 the unarmed resistance to the regime was successful.

(The subsequent theocratic regime led by Ayatollah Khomeini was also repressive. This is the best illustration that nonviolent campaigns do not automatically lead to a better society. But that is also true of violence and conventional politics.)

A variant of the same claim about the ineffectiveness of nonviolence is that it would not have worked against Hitler and the Nazis. As noted above, nonviolence was used against the Nazis, often with success. But it is important to note that nonviolence was seldom even tried against the Nazis. For example, there was no organized effort by Jews in Nazi-occupied Europe to use nonviolent means to resist deportations to concentration and death camps. In many cases, Jewish leaders - believing Nazi lies about the purpose of deportations - cooperated in Nazi efforts to round up Jews for transport.

It is worth remembering that Hitler came to power in Germany in 1933 through conventional politics, namely elections. Furthermore, his regime was supported by many foreign governments; fascist movements and sympathies were common in other countries. The Nazis thus thrived in the conventional politics of the time.

World War II, from the point of view of the Allies, is often said to have been a war against fascism, but actually it was a war against particular regimes in Germany, Italy and Japan, especially against the aggressive wars launched by the German and Japanese governments. The Allies won militarily, but they did not carry the war to the fascist regimes in Spain and Portugal, which remained in power until the 1970s. So war was not a reliable tool for liberation against authoritarian systems.

In the 1930s and 1940s, when the Nazis were in power in Germany, nonviolence was still in its early stages of development as a strategic tool. So claims against nonviolence based on alleged failures in the first half of the 1900s should be treated with caution. Critics sometimes use a few quotes from Gandhi about what he would have done against Hitler as a pretext for dismissing nonviolence as impractical. Skills and understandings of nonviolence have developed greatly since Gandhi's time.

Another claim is that, in opposing injustice, it is necessary to resort to violence after nonviolence has been tried and failed. Actually, the opposite has occurred in several prominent cases. The anti-apartheid struggle in South Africa used violence, largely unsuccessfully, and then turned to nonviolence, which was successful in the 1990s. The armed struggle in East Timor was unsuccessful against the Indonesians in the 1970s and early 1980s, whereas the nonviolent struggle from the late 1980s was far more effective. In Serbia, the authoritarian government led by Slobodan Milosevic survived the country's massive bombing by NATO forces in 1999, but was brought down in 2000 by nonviolent protest.

Another way the power of nonviolence is minimized is by attributing social change to conventional politics or, sometimes, to violence. After Communist regimes in Eastern Europe collapsed in 1989, many political scientists gave most attention to structural factors, such as inefficiencies in command economies, or to conventional political factors, such as the policy of glasnost introduced by Soviet premier Mikhail Gorbachev, neglecting the crucial role of nonviolent protest. Similarly, accounts of the US civil rights struggles often give attention to the role of the federal government in bringing about change, such as Congress passing the 1964 Civil Rights Act, without sufficiently crediting the role of nonviolent struggle.

The comparison of conventional politics, nonviolence and violence is complicated by differences in investment and experience in each approach. Conventional politics has enormous resources to entrench, develop and promote itself, for example through taxation, laws and media coverage. Conventional systems, both authoritarian and representative, invest enormous sums in violence by maintaining large military forces, developing and purchasing sophisticated weapons, training troops, and funding research and development into ways to make systems of violence more effective technologically, organizationally and psychologically. By comparison, investment in nonviolence is very small. Until nonviolence receives the same level of funding and cultural support as conventional politics and violence, it is premature to dismiss it.

For promoting beneficial social change, conventional politics, violence and nonviolence each have strengths and weaknesses. It is unwise to choose between them based on a few examples or a superficial assessment. Instead, it is better to look systematically at how they operate. An outline of what is involved in such an assessment has been given here, with examination of track records, participation, means and ends, and suffering. A key conclusion is that nonviolence has more strengths than commonly recognized.

Social change is a work in progress, both social change as a process and understandings of how best to bring about beneficial change. There is still much to be learned. Anyone who wants to can participate in that learning.


I thank Lyn Carson, Truda Gray, Kurt Schock and Ralph Summy for valuable comments on drafts.


Carter A. (2005). Direct Action and Democracy Today, 298 pp. Cambridge, UK: Polity. [A thorough discussion of justifications for nonviolent action in representative systems.]

Churchill W. with Ryan M. (1998). Pacifism as Pathology: Reflections on the Role of Armed Struggle in North America, 176 pp. Winnipeg, Canada: Arbeiter Ring. [This is an attack on pacifism as a social change strategy.]

Karatnycky A. and Ackerman P. (2005). How Freedom is Won: From Civic Resistance to Durable Democracy, 49 pp. New York: Freedom House. [This is an analysis of transitions from authoritarian rule, comparing violence and nonviolence.]

Moore B. (1972). Reflections on the Causes of Human Misery and upon Certain Proposals to Eliminate Them, 540 pp. Boston: Beacon Press. [This is a careful analysis of suffering and ways to go about reducing it.]

Moyer B. with McAllister J., Finley M. L. and Soifer S. (2001). Doing Democracy: The MAP Model for Organizing Social Movements, 227 pp. Gabriola Island, BC: New Society Publishers. [This describes the process of social change through social movement action.]

Schock K. (2005). Unarmed Insurrections: People Power Movements in Nondemocracies, 228 pp. Minneapolis, MN: University of Minnesota Press. [This study uses social movement theory and nonviolence theory to examine six case studies in nonviolent struggle.]

Semelin J. (1993). Unarmed against Hitler: Civilian Resistance in Europe, 1939-1945, 198 pp. Westport, CT: Praeger. [This is an insightful study of resistance to Nazi occupiers.]

Sharp G. (1973). The Politics of Nonviolent Action, 902 pp. Boston: Porter Sargent. [This is the definitive account of the methods and dynamics of nonviolent action].

Sharp G. (2005). Waging Nonviolent Struggle: 20th Century Practice and 21st Century Potential, 598 pp. Boston: Porter Sargent. [This offers numerous case studies plus accounts of dynamics and strategy for nonviolent action.]

Vollmann W. T. (2004). Rising Up and Rising Down: Some Thoughts on Violence, Freedom and Urgent Means, 733 pp. New York: Ecco. [This is an abridgement of a seven-volume treatment of when violence is justified, with nonviolence dismissed.]

Weinstein D. (1979). Bureaucratic Opposition: Challenging Abuses at the Workplace, 145 pp. New York: Pergamon. [This gives the perspective that bureaucracy is a political form of organization, analogous to an authoritarian state.]

Biographical sketch

Brian Martin has a BA in physics from Rice University (USA, 1969) and a PhD in theoretical physics from Sydney University (Australia, 1976).

He is associate professor in the School of Social Sciences, Media and Communication at the University of Wollongong, Australia. He is the author of 12 books and hundreds of articles about nonviolence, dissent, scientific controversies, democracy, information issues and other topics.