"Religion causes wars" and "religion makes peace" can each produce a long list of examples — and the lists often draw on the same tradition and the same book. The better question is therefore not whether religion causes violence, but this: under what conditions does the same body of text get wired to mobilization, and under what conditions to a ceasefire? One counterintuitive fact — almost every tradition wrote its rules limiting war at the same time it wrote its passages about war; and the earliest modern laws of religious toleration came out of the places where the wars of religion burned hottest.
"One should not honour only one's own sect and disparage others. Acting thus, one helps one's own sect to grow and benefits the sects of others as well."— Ashoka, Major Rock Edict XII, 3rd century BCE
Treating "using religion to restrain violence" as a late invention of modern civilization is a common trick of chronology. Ashoka (reigned c. 268–232 BCE) issued in Major Rock Edict XII what is the earliest known public call for religious tolerance: it asks people to criticize other sects with restraint and not to over-praise their own, for a very practical reason — running down another sect in fact damages your own.
The Abrahamic side took the road of stated conditions. Augustine (354–430) hung the justice of a war on motive and authority; eight centuries later Aquinas, in the Summa Theologiae II-II q. 40 (1265–1274), compressed it into three conditions that must all hold at once: the authority of the sovereign, a just cause, and right intention — the ancestor of jus ad bellum and jus in bello in modern international law.
In Islam, jihad means "striving, exertion" and is not a synonym for external conquest. Classical jurisprudence produced a full set of rules of engagement: the marching orders attributed to Abu Bakr in the earliest legal sources forbid killing women, children and the infirm elderly, and forbid felling fruit trees or tearing down inhabited buildings. As for the much-quoted line "we have returned from the lesser jihad to the greater jihad," it comes from al-Bayhaqi's Kitab al-Zuhd al-Kabir, and al-Bayhaqi himself noted a weakness in its chain of transmission — so it should not be presented as the mainstream orthodox formulation, whereas non-combatant immunity is written plainly into the law books.
In the Indian traditions, Jainism pushes non-harm (ahimsa) to its limit, and the epics' "war according to dharma" (dharma-yuddha) prescribes matched combat and no striking at those fleeing or not fighting. The Sikh tradition wrote the last-resort principle directly into doctrine: Guru Gobind Singh's Zafarnama, the 1707 letter to Aurangzeb, states that when every peaceful means has been exhausted, it is then righteous to draw the sword.
Toleration as an institution likewise did not begin with secularization. The edict of the Transylvanian Diet at Torda in January 1568 made four confessions legal at once, one of Europe's earliest toleration laws; in the same century the Mughal emperor Akbar pursued "universal peace" (sulh-i kul) and abolished the poll tax on non-Muslims in 1564; and the Peace of Westphalia in 1648 ended the Thirty Years' War by granting three confessions legal parity.
The scale of restriction and hostility. Pew Research Center's 16th annual report on global restrictions on religion (published June 2026, data year 2023) covers 198 countries and territories: 58 had high or very high scores on the Government Restrictions Index, slightly below the record 59 in 2022 — but the global median held at 3.0, the highest since the series began in 2007. Countries with high or very high Social Hostilities Index scores rose from 45 to 55, with the median at 1.6 for a third straight year. Harassment is nearly universal: religious groups were harassed in 192 countries, Christians in 165, Muslims in 143 and Jews in 98 (up from 90 the year before, for a group that is under 1% of the world's population).
The magnitude and concentration of terrorism. The Global Terrorism Index 2026 reports 5,582 deaths from terrorism in 2,944 incidents in 2025, down roughly 28% year on year and a decade low — but the distribution is extremely uneven, with the Sahel alone accounting for more than half of global deaths. Deaths in the West rose from a very low base to 57, and since 2021 more than nine in ten fatal attacks there have been carried out by individuals with no organizational affiliation.
"How many wars in history were caused by religion" is a trap question. The most-quoted figure comes from the Encyclopedia of Wars: 123 of 1,763 wars, about 6.98%, classed as religious. But that classification appears only in the index to the third volume — the main text does not categorize wars that way — and the method has been repeatedly challenged. Data pointing the other way exists too: the Religion and Armed Conflict dataset (1975–2015), built to sit alongside the Uppsala conflict data, shows a marked rise over half a century in the share of conflicts where the parties themselves stated a religious issue at onset. The two numbers measure different things — one asks about the historical ledger, the other about contemporary self-description.
How church and state actually sit. By Pew's count for 2015, 43 countries have an official state religion (27 Islam, 13 a Christian denomination), around 40 more have no official religion but favour one in practice, and over 100 countries and territories have neither. Separation of religion and state is not the world's default setting; it is one arrangement among several.
The scriptures of nearly every large tradition contain both kinds of passage: some usable for mobilization, some for restraint. What decides which way things go usually lies not in the text but in three switches.
William Cavanaugh's The Myth of Religious Violence (Oxford University Press, 2009) argues that splitting "religious" from "secular" and holding the former to be inherently more prone to irrational violence is itself a modern Western product. The thesis is much contested, but it does flag one thing: ruling that a conflict "is religious" rather than "is political" already carries a judgement.
The "devoted actor" framework developed by the anthropologist Scott Atran and colleagues holds that values treated as sacred are not open to material trade-offs: buying a concession with compensation is not merely ineffective but often provokes a stronger reaction, because the offer to trade is itself read as an insult. Its companion is identity fusion, where personal and group identity coincide completely. It also explains why effective intervention is usually not a better offer, but loosening the social networks that bind people together.
Misconception: most wars in history were caused by religion.
The much-quoted 6.98% comes from an index classification in the third volume of one encyclopedia; the main text makes no such categorization. But the reverse claim — that religion is never a cause — fails too: in contemporary conflict data, the share of parties stating a religious issue has risen.
Misconception: "jihad" simply means war against outsiders.
Jihad means "striving"; classical jurisprudence discusses war and also lays down non-combatant immunity. Meanwhile the most widely circulated "greater/lesser jihad" line has a chain of transmission that al-Bayhaqi himself marked as weak. Taking half of either side is dishonest.
Misconception: some traditions are peaceful by nature and others warlike.
Doctrinal tendency and historical record do not move in step: traditions that write non-violence most thoroughly have had adherent communities drawn into violence under particular political conditions, and traditions known for strict rules of engagement have also produced long arrangements of peaceful coexistence. Conditions explain more than labels.
Misconception: when religion recedes, violence recedes with it.
The largest organized killings of the twentieth century were carried out in the name of secular ideologies, and the great majority of recent fatal attacks in the West come from individuals with no organizational affiliation. The language of mobilization changes; the machinery need not.
This issue and the last are two sides of one question. Day 23 concluded that who holds the right to interpret often matters more to a tradition's direction than what its scripture says. This issue is the costliest application of that line — the same passage, handed to different structures of interpretive authority, yields either mobilization or a ceasefire. The differences in conceptions of the divine from Day 21 barely register here: monotheistic and polytheistic traditions alike have had eras of tolerance and eras of persecution.
The institutional history of toleration is longer than the history of secularization. Torda, Akbar and Westphalia all happened in societies where religion was intensely present, and the driver was usually pragmatism after exhaustion rather than principle in advance. Toleration therefore does not accumulate on its own — it is negotiated again and again, and can be undone.
Two contemporary pressure points, pointing opposite ways. At the state level, Pew's median for government restrictions is at a record, meaning constraint increasingly comes from policy and enforcement rather than communal conflict. At the individual level, mobilization is shifting from organizations toward online networks and lone actors, which leaves the old path of "dialogue with religious leaders" unable to reach the people who matter most. The first calls for work on institutions, the second for work on belonging; there is almost no shared remedy.
This piece surveys, from a neutral and comparative standpoint, the rules each tradition holds on war, restraint and reconciliation, and how those rules have played out historically. It offers no praise, blame or ranking, passes no judgement on any living religious figure, and takes no position on any specific present-day conflict. Figures and dates are recent estimates from scholars and survey organizations and vary with definition and method.