Terrorism
in a Hurry

The Whole Thing in One Page

Terrorism looks like an eruption of madness: a bomb, a gunman, a hostage, an aircraft forced from its route. That picture begins too late and ends too soon. The dead and injured are the immediate victims. The political target is usually elsewhere: a government expected to concede, a population meant to panic, a community invited to retaliate, supporters who need proof of strength, rivals who must be outbid, or peace that someone wants to wreck.

The central fact is the gap between the people harmed and the people addressed. Terrorism uses violence or its threat to change the behaviour of a wider audience. Fear matters, but publicity, interpretation and response matter more. A small organisation cannot defeat a state division by division. It may still provoke the state into repression, persuade civilians that government cannot protect them, frighten witnesses into silence, turn compromise into betrayal, or make its cause impossible to ignore. The physical attack is one move in a political exchange. Its meaning is made afterwards, through claims, coverage, rumours, funerals, speeches, police action and whatever a frightened public demands next.

That is why definition is difficult. Terrorism overlaps with insurgency, guerrilla warfare, assassination, hate crime, organised crime and state repression. Governments write legal definitions for prosecution and prevention. Scholars want categories that travel across countries and centuries. Movements reject the label for themselves and apply it to enemies. The argument is not verbal fussiness. A label decides which deaths count, which powers become lawful and whose violence is treated as politics rather than crime or war.

Tactics change with technology and opportunity, from nineteenth-century assassination and dynamite to hijacking, hostage-taking, suicide attacks, vehicle attacks and online incitement. The strategic logics recur. Groups try to wear opponents down, intimidate populations, provoke indiscriminate retaliation, spoil negotiations or prove themselves more committed than competing factions. An attack can fail operationally and still spread fear. It can succeed spectacularly and damage the cause that launched it.

Causes do not reduce to poverty, religion, mental illness or one personality. Political exclusion, occupation, civil war, state abuse, identity, ideology and grievance can create a field of possibility. Organisations then recruit, train, select targets, reward commitment and make violence appear useful. Individuals enter through different routes involving friendship, kinship, status, moral outrage, adventure, coercion, belonging or personal crisis. Millions share grievances. Few join clandestine violence. The missing bridge is usually organisation and opportunity.

Responses therefore need more than force. Intelligence and policing disrupt plots. Courts preserve legitimacy. Protective measures reduce opportunity. Financial controls and international cooperation constrict networks. Political settlements can remove an audience or grievance. Military power may be necessary against groups holding territory, yet it is a blunt answer to a dispersed campaign. Prevention is valuable when it follows behaviour and context rather than pretending to detect a terrorist type.

The hardest lesson is that the responder enters the mechanism. Panic, collective blame, torture, indiscriminate force and permanent emergency can grant an organisation the polarisation and recruitment it sought. Restraint is not passivity. It is the discipline of denying a small actor control over a larger one.

That is the book.

Why You Should Care

Terrorism tries to recruit you without asking. You may enter as a witness watching repeated footage, a voter demanding action, a traveller accepting a new security rule, a neighbour viewed with suspicion, an official deciding how much risk is tolerable, or a member of a community expected to answer for a stranger. The attack may occur far away. Its political work begins when other people change their behaviour because it happened.

The first reason to care is personal. Terrorism is meant to narrow your imagination. It invites the belief that nowhere is safe, no stranger is innocent and no ordinary freedom can survive without exception. Those conclusions may be the intended product, not an accidental emotional aftershock.

This makes terrorism unusually good at distorting scale. A rare danger can dominate attention because it is intentional, spectacular and morally legible. Storms do not hate you. Road deaths do not issue communiqués. A deliberate attack tells a story with an enemy, an audience and a demand, then leaves governments and media to decide how large that story becomes. Public fear can therefore rise while measured risk falls, or remain low while communities inside a conflict face repeated violence. The word describes events that are globally uneven and personally hard to price.

Current figures make the mismatch visible. One prominent event dataset recorded 5,582 deaths from terrorism in 2025, concentrated heavily in conflict-affected countries. The total was the lowest since that dataset began in 2007, yet it still represented thousands of particular lives and local political systems under strain. Western publics often encounter terrorism through isolated high-impact attacks and constant security architecture. People in parts of the Sahel, Pakistan, Nigeria or the Democratic Republic of the Congo encounter it inside insurgency, weak state reach, communal conflict and military response. One word covers different lived realities. A sensible response begins by asking which one is present, because a clandestine urban cell, a territorial insurgency and an isolated attacker do not yield to the same instrument.

You should care because the wrong mental model produces bad policy. Treat terrorism as madness and motives become irrelevant. Treat it as poverty and development spending becomes a security cure it cannot reliably be. Treat it as a religion and millions of peaceful believers become a suspect pool. Treat every attack as war and soldiers inherit problems better handled by detectives, courts, financial investigators, negotiators or local politics. Treat restraint as weakness and a state may deliver the overreaction its opponent designed the attack to obtain.

The subject also exposes a larger question about power. Terrorist organisations are usually materially weaker than the states they confront. Their leverage comes from choosing the time and place, hiding among ordinary life, exploiting open communication and forcing stronger actors to act under public pressure. This is asymmetric conflict in its sharpest form. A few people can impose costs on millions because modern societies are connected, visible and expected to protect their citizens everywhere.

None of this turns murder into cleverness or asks you to admire its perpetrators. Strategy explains conduct; it does not excuse it. Nor does understanding causes imply that victims should bargain with anyone who harms them. The moral facts can remain clear while the political mechanism is studied closely. Refusing to understand an enemy is satisfying for about five minutes. After that it becomes a gift to the enemy.

The useful promise of this book is narrower and harder. You will learn to separate event from campaign, tactic from cause, attacker from audience, and operational success from political success. You will see why people enter violent movements, why campaigns persist after failure, how they end, and why the quality of a response is measured partly by what it refuses to become.

The Core Ideas

The Immediate Victim Is Not the Whole Target

Nineteen hijackers murdered 2,977 people on 11 September 2001. The victims were in aircraft, offices, stairwells, streets and rescue services. The intended audience extended far beyond them. Al-Qaeda wanted the United States to feel vulnerable, Muslim audiences to see American power wounded, potential recruits to see daring and capability, allied governments to doubt the cost of alignment, and Washington to respond in ways that would widen the conflict. The attacks were physical facts and arguments made with bodies.

That gap between victim and audience supplies the cleanest mental model in the subject. Ordinary violence often aims to remove, punish or rob the person attacked. Terrorism usually aims to affect other people through the attack. The audience may be a state, a community the organisation claims to defend, a rival movement, investors, tourists, police recruits or an international public. Several audiences can receive different messages at once.

Fear is one route, not the whole route. An assassination may advertise penetration of a protected system. An attack on a place of worship may invite communal retaliation. A hostage crisis may buy time and attention. Violence against local officials may persuade villagers that the state cannot remain. An attack during negotiations may make compromise politically toxic. The intended effect might be obedience, polarisation, publicity, recruitment or a display of commitment rather than undirected terror.

This is why symbolic targets matter. A parliament, market, school, embassy, transport system or memorial carries meaning before it is attacked. The perpetrator borrows that meaning, then relies on institutions to distribute it. News organisations report; politicians speak; markets move; rumours outrun confirmed facts; security measures appear. Modern communication increases the possible audience, but terrorism predates mass media because word of exemplary violence can travel through sermons, pamphlets, funerals and frightened conversation.

Audiences are rarely one crowd. A government may read an attack as a test of resolve, while a threatened minority reads it as notice to leave and supporters read it as proof that resistance remains alive. These interpretations feed one another. A retaliatory speech changes recruitment material; a frightened market alters political pressure; a community's refusal to retaliate can ruin a provocation. The event does not carry one fixed meaning from attacker to target. Meaning is contested after the violence, which is why calm factual communication and the conduct of authorities belong inside security rather than beside it.

The model also corrects a moral danger. Calling victims a message can make their lives sound secondary. They are not. The strategic analysis describes the perpetrator's design, not the value of the people harmed. Good response begins with rescue, care, accurate information, investigation and dignity for victims. It then refuses to let the attacker's preferred story become the only public account.

A useful first diagram therefore has three boxes: victim, audience, demand. Who was directly harmed? Whose conduct was supposed to change? What change was sought? When those questions produce no political or ideological purpose, another category of violence may fit better. When they produce several audiences and incompatible demands, the campaign may be less coherent than its propaganda claims. Either finding is more useful than treating terror as a substance released at the scene.

The Label Draws a Political Boundary

There is no universally accepted definition of terrorism, but that does not mean anything can mean anything. Most serious definitions cluster around several features: violence or threatened violence; a political, ideological, religious or social purpose; an intention to intimidate or coerce beyond the immediate victim; and attacks on civilians, noncombatants or other protected and symbolic targets. Agreement weakens over who can commit it, which targets count and how much harm is required.

The United Nations has built treaties around particular offences such as hostage-taking, attacks on civil aviation and terrorist bombing, while efforts to agree one comprehensive convention have repeatedly stalled. Security Council Resolution 1566 described criminal violence intended to cause death or serious injury, or hostage-taking, where the purpose is to provoke terror, intimidate a population or compel a government or international organisation. It offered convergence without ending disputes about occupation, armed struggle and state force.

Domestic law answers a different question: what conduct should trigger investigation, proscription, special powers or punishment? The United Kingdom's Terrorism Act 2000 reaches beyond lethal attacks to serious property damage, danger to public health and serious electronic disruption when the required cause and coercive intent are present. That breadth may help authorities intervene before bloodshed. It also creates hard boundary questions about protest, sabotage, publication and association. A criminal statute must be applied by courts to particular facts. A historian needs a category that compares unlike systems without importing one country's politics into the past.

The largest dispute concerns states. The word entered modern political language during the French Revolution to describe terror practised through government. States have since used exemplary violence, disappearances, collective punishment and attacks on civilians to frighten populations into obedience. Yet many datasets and laws reserve terrorism for non-state actors, partly because state violence is already classified through war, repression, crimes against humanity and other legal frameworks. Excluding states can clarify one field of study. Pretending that states cannot terrorise falsifies the history of the word.

Legal categories also answer questions that the moral label cannot settle. International humanitarian law asks whether a conflict exists, who is protected and whether an attack distinguishes civilians from military objectives. Criminal law asks whether an offence and the required intent can be proved. A bombing can violate the laws of war without fitting a domestic terrorism charge, and a terrorism offence can occur outside armed conflict. Keeping these frameworks separate prevents one emotionally powerful word from replacing the exact rule by which conduct should be judged.

“Terrorist” also turns a tactic into an identity. The same organisation may run elections, provide welfare, fight soldiers, extort businesses and attack civilians. Some insurgents use terrorism among other methods. Some terrorist cells cannot mount an insurgency. A state may negotiate with an organisation it still proscribes. A movement can abandon attacks without abandoning its political goal. Analysis improves when the noun is unpacked into actors, actions and campaigns.

The familiar line that one person's terrorist is another person's freedom fighter confuses cause with method. A cause can be just and a tactic criminal; an oppressive cause can use conventional warfare; a liberation movement can restrain its targeting; a government can commit atrocities while calling itself lawful. Political sympathy does not settle classification. Neither does revulsion. The disciplined question is what was done, to whom, for what wider effect, under which rules. The label should follow the evidence rather than perform the argument in advance.

Small Forces Seek Large Reactions

Terrorism is associated with weak actors because it avoids the contest in which weakness is decisive. A clandestine group usually cannot hold an open front against a capable state. It can choose a vulnerable moment, concentrate limited resources and impose political decisions on a timetable of its own. The purpose is not to match the stronger side's force. It is to make that force awkward, expensive or self-defeating.

An influential model identifies five recurring strategic logics. Attrition says: we can continue imposing costs until you leave, concede or lose public patience. Intimidation says: we can punish collaboration and the state cannot protect you, so obey us instead. Provocation seeks a harsh or indiscriminate response that radicalises the population from which the group wants support. Spoiling tries to destroy trust between parties approaching settlement. Outbidding uses violence to show that one faction is more committed, capable or uncompromising than its rivals.

These are not five species kept in separate cages. One campaign may combine them, and leaders may disagree about which one they are pursuing. An attack can be intended as attrition towards an enemy, intimidation towards local officials and outbidding towards another armed group. Its effects can also diverge from intent. Violence meant to inspire may disgust potential supporters. Repression meant to isolate a cell may validate its account of collective persecution.

The five logics explain why body counts alone misread campaigns. A low-casualty assassination can alter negotiations or succession. A foiled plot can produce publicity and costly protection. Repeated attacks on minor local targets can hollow out administration. Conversely, a spectacular mass-casualty attack can trigger organisational destruction and broad moral rejection. Physical scale and political effect are related, but not in a straight line.

This is also why secrecy and publicity coexist. The organisation conceals people, resources and preparation, then wants selected facts known. A claim of responsibility links violence to a demand and brand. Anonymous attacks may still intimidate, destabilise or encourage rumours, but they make bargaining harder because audiences cannot know whose capability or promise they are assessing. Some groups delay claims, issue false ones or let supporters infer authorship. Communication is strategic even when it is confused.

The most dangerous error is to assume that every perpetrator is a cool optimiser. Organisations contain zealots, careerists, frightened recruits, local commanders and leaders with poor information. Internal incentives can reward activity rather than results. Security pressure can shorten planning horizons. Competition can produce escalation that damages everyone. A tactic may persist because it sustains group cohesion or leadership authority even while failing politically.

Rationality therefore means responsive to incentives, beliefs and constraints, not wise, humane or successful. Terrorism can be instrumentally chosen and morally abhorrent. It can also be instrumentally foolish. The analyst's task is to reconstruct the expected chain: if we do this, who is meant to do what next? The responder's task is to break that chain at its weakest link.

That requires identifying the logic rather than answering all violence with the same gesture. Better protection and credible endurance challenge attrition. Reliable local security and witness protection challenge intimidation. Precision and legal restraint frustrate provocation. Protected negotiations and reassurance between parties reduce spoiling. Political access and constituency support can weaken outbidding. None is automatic, and one response may help against one logic while feeding another. The framework earns its place when it turns outrage into a question about sequence, audience and the next avoidable error.

The Organisation Is Larger Than the Attacker

The person who carries out an attack is the visible tip of a system. Behind organised campaigns sit people who recruit, move money, obtain documents, select propaganda, provide shelter, transmit orders, gather information, settle disputes and maintain contact with supporters. None of these functions needs a large bureaucracy. Together they convert grievance into repeatable violence.

Organisations vary. Some have ranks, departments and a recognisable command. Others operate as networks whose nodes have uneven contact. Franchises may borrow a global name while pursuing local wars. Insurgent organisations can govern territory, tax communities and field armed units while also using terrorism. Small cells may share ideology and occasional assistance without receiving detailed orders. Labels such as “centralised” and “leaderless” describe degrees, not switches.

Structure changes under pressure. Hierarchy can coordinate strategy and restrain unauthorised attacks, but it creates leaders and communication links that intelligence services can target. Decentralisation protects parts of the network, yet it weakens quality control and political coherence. A brand that inspires autonomous violence may gain reach while losing the ability to decide who acts in its name. Success at survival can become failure at strategy.

Support is broader than membership. A clandestine campaign may depend on a community without commanding it. Relatives may hide someone from loyalty or fear. Donors may approve a political cause while knowing little about operations. Businesses may be extorted. Diasporas can provide money, publicity or diplomatic pressure. Political parties and charities may sit near armed wings in relationships that differ across countries and over time. Collapsing all sympathy into complicity produces bad intelligence and collective blame.

The “lone actor” reveals the same problem at smaller scale. The term usually means that one person conducts the final attack without direct operational partners. It does not prove that the person formed beliefs alone, learned no scripts from predecessors or signalled intent to others. Empirical studies have found varied mixtures of isolation, online participation, prior contacts, leakage, personal crisis and imitation. Operational solitude can be socially crowded.

Financing is similarly easy to dramatise. Large territorial organisations need salaries, taxation, smuggling income and administration. Many individual attacks cost little. Financial investigation remains valuable because transfers, procurement and facilitators can expose networks, and sanctions can constrain larger organisations. It cannot work as a universal off switch. A campaign supported through local coercion or ordinary earnings may leave little spectacular money trail.

Prisons, political fronts, welfare bodies, encrypted channels and ordinary friendship groups can all become organisational environments without performing the same function. A prison may connect people while also removing them from operations. A political party may discipline an armed wing or provide it cover. A public charity may be independent, coerced or penetrated. Mapping names alone cannot resolve these relationships. Investigators need evidence about authority, knowledge and flow, because indiscriminate disruption can destroy the social ties through which communities share warnings and members find exits.

The organisation should therefore be mapped as functions and relationships rather than imagined as a pyramid. Who authorises? Who enables? Who supplies legitimacy? Which ties are voluntary, coerced or merely alleged? What happens if one person disappears? Leadership removal can collapse a young, personalised group and barely slow an older institution with succession rules. Research on decapitation reaches conflicting aggregate conclusions because organisations differ in age, size, ideology, environment and state pressure. The name at the top matters. The machinery underneath decides how much.

Causes Stack Rather Than March in a Line

A cause of terrorism must explain two gaps. Why does political violence arise in one place rather than another? And why, among millions exposed to similar conditions, do a few people join while most do not? Single-cause theories usually answer one gap badly and ignore the other.

Begin at the structural level. Civil war, foreign occupation, state collapse, severe repression, communal exclusion and blocked political competition can make clandestine violence more available or plausible. Geography, borders, weapons markets, prisons, migration routes and communication systems shape opportunity. These conditions do not mechanically produce terrorism. They alter the field in which organisations can recruit, hide and claim necessity.

Then comes the group level, often the missing bridge. Leaders frame injury as a shared injustice, identify an enemy, provide a moral language, offer roles and decide that attacking a wider audience may work. Organisations select members rather than receiving a random sample of the aggrieved. They train attention, reward loyalty and make previous violence create obligations for further violence. Competition with the state and with rival movements changes the route after it has begun.

At the individual level, pathways differ. Political outrage can matter alongside friendship, kinship, status, excitement, fear, revenge, belonging, coercion or a search for significance. Some people enter for one reason and remain for another. Ideological commitment may deepen after joining. Others disengage while retaining radical beliefs because fatigue, family, disillusionment, fear or changed relationships alter the cost of participation. Belief and behaviour are connected without being identical.

Poverty is a poor master key. Studies of particular organisations have often found members who were no poorer, and sometimes better educated, than surrounding populations. That may reflect political engagement, organisational selection and the skills clandestine groups value. It does not prove that economic conditions never matter. Unemployment, corruption, unequal services and lost livelihoods can shape grievance and state legitimacy in specific settings. The defensible conclusion is that poverty alone neither identifies likely attackers nor supplies a general causal law.

Mental illness needs the same discipline. There is no clinical terrorist personality, and organised groups often screen for people able to maintain secrecy and perform assigned roles. Some lone-actor samples show higher rates of mental disorder or crisis than tightly organised group samples. Diagnosis still does not explain political meaning, target choice or why distress becomes ideologically framed violence in one case and not another. Treating illness as the cause can stigmatise millions while missing behaviour that matters.

Religion can authorise sacrifice, define cosmic stakes, create community and supply texts or symbols. Secular ideologies have performed comparable work. Religious language also enters local struggles over power, territory, status and security. Explaining an attack by naming the perpetrator's religion stops at the label. The serious questions concern interpretation, authority, organisation, conflict and the alternatives available.

Routes often contain a precipitating event rather than a timeless root. An arrest, killing, humiliation, prison encounter, communal riot, failed protest or personal loss can make an existing narrative urgent. Even then the event is not a button. It is interpreted through relationships and available scripts. Retrospective accounts can overstate the one moment that now makes the life look coherent, while neglecting chance contacts and abandoned alternatives. The closer analysis gets to an individual, the less useful broad population averages become.

Causes therefore stack: conditions, narratives, organisations, relationships, opportunities and personal trajectories. No stack is identical. Prevention improves when it asks where a pathway can be interrupted rather than searching for a demographic type who can be found in advance.

Campaigns Learn, Compete and End

Terrorism has a history because methods travel. Nineteenth-century revolutionaries watched one another across borders, read trial speeches and learned that new explosives and mass newspapers allowed a few conspirators to command attention. Anti-colonial movements studied previous campaigns. Aircraft hijacking spread through imitation before security and international cooperation made it harder. Suicide attacks were adopted by organisations with different religions and goals. Online platforms accelerated the circulation of images, manifestos and models of fame.

A useful historical scheme divides modern terrorism into anarchist, anti-colonial, New Left and religious waves. It captures broad changes in energy, language and international connection. It should not be mistaken for tides that neatly replaced one another. Nationalist, racist, revolutionary, religious and state violence overlapped throughout. Local conflicts did not wait for a global period label. Technology diffused across causes faster than ideology did.

Campaigns learn internally as well. Security changes produce adaptation. Arrests alter trust. Failed attacks can reveal weakness or create martyrs. Splits generate more cautious factions and more violent ones. Peace talks redistribute status, which is why spoilers attack. An organisation may escalate to keep recruits, donors or media attention even when escalation moves its stated political objective further away.

Innovation is therefore interactive. A new tactic may spread because it appears successful, because coverage makes it imaginable or because protection has closed easier routes. States copy each other's screening, laws and specialist units; organisations then search for unguarded systems or cheaper methods. Diffusion does not prove central direction. Similar attacks can arise from common constraints and visible precedent. The contest is less an arms race towards ever greater sophistication than repeated movement between access, defence, publicity and adaptation.

Success has to be separated into levels. An operation can reach its immediate target. A message can dominate news. Recruitment can rise. The organisation can survive. A government can change a policy. The final political programme can be achieved. These outcomes do not travel together. Research comparing stated goals often finds poor strategic performance, especially where civilians are the main targets or demands are maximal. Yet failure to win does not make the tactic irrelevant. A campaign can delay peace, harden identity, provoke repression and keep leaders in command.

Endings are equally varied. Leaders are captured or killed. Groups negotiate and enter politics. Support collapses. Members age, quarrel or defect. Police and intelligence penetrate networks. States repress campaigns, sometimes at severe human cost. A movement wins a limited demand, changes into insurgency or crime, or survives as a label after losing most capacity. Several endings can occur in sequence.

The Provisional IRA combined a long armed campaign with political organisation and ended its ceasefire cycle through the 1998 Good Friday Agreement, later decommissioning weapons. ETA stopped its armed campaign in 2011 and disbanded in 2018 after sustained policing, political change and loss of support. Peru's Shining Path was badly damaged by the 1992 capture of Abimael Guzmán, though remnants persisted. Sri Lanka destroyed the Tamil Tigers' territorial organisation militarily in 2009 amid grave civilian suffering. No single template fits all four.

Campaigns should therefore be judged through trajectory rather than spectacle. Is capacity rising? Is the claimed constituency expanding or recoiling? Are rivals multiplying? Can the organisation replace skilled people? Has the political demand narrowed? The dramatic attack is a frame. The campaign is the film, and many films end through exhaustion rather than a final scene.

The Response Enters the Mechanism

A state under attack has duties that cannot be wished away. It must save life, stop continuing threats, identify perpetrators, protect evidence, inform the public and prevent repetition. Delay can kill. Passivity can invite further violence and destroy confidence. The question is not whether to respond. It is how to respond without converting the attacker's small act into the state's large mistake.

Different threats require different instruments. Intelligence and police work are central against clandestine networks because they identify people, plans and facilitators before violence. Criminal courts test evidence and make punishment public and bounded. Protective security can reduce opportunity and reassure users of exposed places. Financial measures constrain some networks. Border cooperation matters where people or resources cross jurisdictions. Prevention and disengagement work best when attached to credible local knowledge, observable behaviour and individual routes rather than broad suspicion.

Political action belongs in the same frame. A negotiable territorial dispute, an excluded constituency and a transnational apocalyptic movement do not present the same bargaining problem. Addressing a legitimate grievance does not reward every person who exploits it. Refusing any political channel can strengthen those who say violence is the only route. Negotiation can split coalitions, expose limited aims and create exits, but it can also confer status, frighten partners or be used for recovery. It is a tool, not a moral verdict.

Military force has a narrower but real place. When an organisation holds territory, trains openly, fields units and governs populations, policing may be impossible until control is broken. The defeat of Islamic State's territorial caliphate required armed force alongside local partners, intelligence and financial pressure. Territorial loss did not erase its affiliates, ideology or clandestine capacity. A battlefield victory can change the problem back into one for police, politics and administration. Keeping the military answer after the military target has changed is how means outlive purpose.

Overreaction supplies several terrorist strategies at once. Indiscriminate force can produce the provocation sought. Collective blame assists recruitment. Torture and unlawful detention damage intelligence, alliances and legitimacy. Sweeping emergency powers may migrate to conduct far from the attack that justified them. Saturated publicity can grant celebrity and spread scripts for imitation. These effects are not reasons for inaction. They are costs that belong in the security calculation.

Northern Ireland offers a blunt example. Internment without trial in 1971 was introduced as a security measure but was applied overwhelmingly against the nationalist community, based on poor intelligence, amid abusive treatment. It deepened alienation and is widely judged to have aided republican recruitment rather than isolating the Provisional IRA. Later progress depended on better intelligence and policing, political negotiation, institutional change and a route by which armed actors could stop without every underlying dispute being declared solved.

Resilience completes the response. Public authorities should communicate verified facts, distinguish communities from perpetrators, support victims over time and restore ordinary life without pretending risk has vanished. Citizens can refuse both panic and indifference. The aim is not emotional numbness. It is proportion: enough urgency to protect people, enough restraint to preserve law, and enough strategic patience to deny the attacker command of what happens next.

A strategy also needs measures and an end condition. Fewer attacks matter, but so do disrupted capability, public cooperation, lawful convictions, reduced support and the return of ordinary institutions. An emergency power that cannot state what evidence would justify its removal has escaped evaluation. A military campaign with no political condition for completion turns endurance into purpose. Review is not administrative tidiness. It is how a state detects when an instrument built for yesterday's organisation has begun producing tomorrow's grievance.

Core Idea 1 began with the distance between victim and audience. That distance is terrorism's leverage. It is also its vulnerability. The audience is not obliged to perform the role written for it.

How It Actually Works

Terror before terrorism

The word began on the side of the state. During the French Revolution, terror described a method of government: exemplary violence used to defend the new republic against enemies within and without. In 1793 and 1794 the Revolutionary Tribunal, emergency laws and public execution made fear an instrument of rule. “Terrorist” first named an agent or supporter of that system.

After the Jacobins fell, the word changed direction. Terrorism became an accusation against illegitimate violence, then increasingly against conspirators attacking states from below. The older meaning never disappeared. Governments and armed movements have continued to fight over who may use force, whose fear counts and whether violence is law, war, repression or terror.

This origin matters because terrorism was never a neutral object waiting to be discovered. It entered politics as a judgement about authority. The modern category grew while states were claiming a monopoly on legitimate violence and revolutionary movements were claiming a right to overthrow them.

The bomb, the assassin and the newspaper

Modern clandestine terrorism took recognisable form in the late nineteenth century. Cities grew, newspapers reached mass audiences, railways moved people and ideas, and dynamite gave small groups a new capacity for destruction. Revolutionaries could now produce an event in one street and make it a political fact across a continent.

Russia supplied the defining early campaign. Narodnaya Volya, the People's Will, emerged from a revolutionary movement frustrated by autocracy and failed rural agitation. It built a secret organisation around the assassination of Tsar Alexander II. After several failed attempts, members killed him in St Petersburg in 1881. They succeeded at the operation and failed at the revolution. The state executed leaders, intensified repression and broke the organisation. A dead ruler did not produce the popular uprising they expected.

Anarchists in Europe and the Americas developed the idea of “propaganda by the deed”: an exemplary act would expose state weakness, awaken the masses or trigger revolt. Presidents, monarchs and ministers were assassinated. Bombings reached cafés, theatres and public buildings. The movement was transnational in imagination but fragmented in organisation, which made responsibility and strategy difficult to control. It generated fear and copycats, then helped states build political police, extradition arrangements and specialist investigation.

Irish republicans added another thread. Fenian networks used bombing in Britain during the 1880s to place the Irish question inside British cities. Later republican movements combined guerrilla action, political organisation and attacks that different observers classified differently. The mixture would become common: terrorism rarely arrived alone.

Empire, nation and the problem of success

The first half of the twentieth century brought mass war, revolution and anti-colonial struggle. Violence against civilians and political assassination appeared inside conflicts whose main form was often insurgency. The terrorism label followed power. Imperial governments applied it to movements they denied the standing of armies; movements described themselves as national liberation forces and highlighted state coercion.

After 1945, several campaigns made the classification problem unavoidable. Zionist underground organisations attacked British authorities in Mandatory Palestine, including the 1946 bombing of the King David Hotel by Irgun. In Cyprus, EOKA attacked British rule while seeking union with Greece. In Algeria, the National Liberation Front used urban bombings, assassination and intimidation alongside rural insurgency. French forces answered with mass detention, torture, intelligence penetration and military control during the Battle of Algiers.

Algeria became independent in 1962, but the slogan “terrorism worked” explains too much with one cause. France faced a wider war, international pressure, domestic political crisis and the judgement of Charles de Gaulle that continued rule was untenable. FLN violence helped make Algeria costly and ungovernable while also killing civilians, coercing rivals and narrowing political alternatives. Tactical effect, contribution to pressure and full political causation are different claims.

The same warning applies elsewhere. Some organisations linked to successful national movements used terrorism. Many failed movements used it too. States sometimes conceded limited demands to stop attacks while refusing larger programmes. A flag raised after independence does not tell you which tactic caused the outcome or what else the campaign destroyed.

The age of the international spectacle

In the late 1960s terrorism moved onto television. Palestinian organisations, especially the Popular Front for the Liberation of Palestine, used international aircraft hijackings to bypass a local military balance and command global attention. The 1970 Dawson's Field hijackings produced days of live crisis and a vast audience. The murder of Israeli athletes at the 1972 Munich Olympics turned a protected global festival into a stage for political violence.

The tactic brought attention and transformed security. Airports hardened, screening expanded, states improved cross-border cooperation and international conventions criminalised attacks on aviation and hostage-taking. What appeared innovative became harder to repeat. This cycle would recur: open systems create opportunities; attacks reveal them; protection closes some paths and displaces effort towards others.

The period also contained many organisations that had little in common beyond method. The Provisional IRA and ETA pursued nationalist projects. The Red Army Faction in West Germany and Red Brigades in Italy came from revolutionary left-wing milieus. Right-wing terrorists attacked minorities, trade unionists, officials and democratic institutions. Latin American dictatorships used disappearance and exemplary violence on a much larger scale. The idea of one worldwide terrorist movement was no more accurate then than now.

Television changed incentives. A small operation could hold attention for hours, and negotiators had to manage both the people in danger and the public drama around them. Governments developed specialised police units and crisis procedures. News organisations began debating whether live coverage revealed too much, amplified demands or rewarded spectacle. The attack was becoming inseparable from its distribution.

Long campaigns and political exits

Northern Ireland showed how a campaign changes while its name stays fixed. The Provisional IRA emerged amid communal violence, discrimination, the deployment of British troops and a split in republican organisation at the end of the 1960s. It combined attacks in Northern Ireland, bombing in Britain, punishment within nationalist communities and a claim to fight a war of national liberation. Loyalist organisations attacked Catholic civilians, often using sectarian terror to alter where people could live and whom they could trust. State forces, police, intelligence agencies, parties and governments all became actors in the same system.

Security policy moved from military dominance and internment towards intelligence-led disruption and reformed policing, though abuses and contested killings remained part of the conflict. Republican and loyalist organisations were penetrated but not erased. Political channels developed through public and secret contact. The 1985 Anglo-Irish Agreement gave the Irish government a consultative role. Ceasefires in the 1990s opened negotiations that produced the Good Friday Agreement in 1998, creating power-sharing institutions and changes to constitutional claims, policing and prisoner arrangements. The settlement did not prove that violence had been necessary, nor that policing had been sufficient. It created a route from armed organisation to politics while majorities in both jurisdictions endorsed new rules.

ETA followed another uneven decline. Founded under Franco's dictatorship, it continued killing after Spain democratised and the Basque Country gained extensive autonomy. Its violence increasingly lost legitimacy among the population it claimed to represent. French and Spanish police cooperation restricted sanctuary; arrests and intelligence weakened leadership; civil society mobilised against murder; political alternatives remained open. ETA announced the end of armed activity in 2011 and dissolved in 2018. The campaign ended after the original regime had vanished, which is a warning against assuming that an organisation's survival proves its stated grievance still explains its conduct.

Suicide attack and sacred authority

From the 1980s, suicide attacks became a durable tactic across distinct conflicts. Lebanese organisations used them against foreign and local forces. The Liberation Tigers of Tamil Eelam, a secular nationalist movement in Sri Lanka, used suicide operations systematically and created a specialised unit. Hamas and Palestinian Islamic Jihad adopted them in the 1990s. Al-Qaeda later used teams prepared to die inside a transnational campaign.

The tactic solved organisational problems at a terrible moral cost. An attacker who expected to die could continue until interception, signal commitment and remove the need for escape. The death also created a martyr story for supporters. Yet organisations still had to recruit, train, authorise and frame the act. Suicide terrorism was never evidence that strategy had disappeared. It joined organisational calculation to individual sacrifice.

Religious authority altered the language of legitimacy, enemy and time. The Iranian Revolution, the Soviet invasion of Afghanistan and the seizure of the Grand Mosque in Mecca all occurred in 1979 and helped make that year a convenient marker for a religiously framed period of political violence. The marker is useful and crude. Sikh, Jewish, Christian, Buddhist, Hindu and Islamist extremists have all used religious identities in different conflicts. Secular movements persisted. Local power and territory remained beneath global claims.

Aum Shinrikyo showed another route. In 1995 the Japanese cult released sarin on the Tokyo subway, killing and injuring commuters. Its apocalyptic doctrine, wealth and scientific pretension did not fit the standard image of an insurgent constituency. Terrorism could emerge from a closed movement seeking catastrophe rather than a negotiable national demand.

Al-Qaeda and the war made from the response

Al-Qaeda grew from networks formed around the anti-Soviet war in Afghanistan, then reoriented towards the United States and allied governments. It bombed American embassies in East Africa in 1998 and attacked the USS Cole in 2000. On 11 September 2001, its hijackers attacked New York and Washington and brought down a fourth aircraft in Pennsylvania after passengers resisted.

The attack achieved surprise, destruction and unmatched publicity. It also brought the overthrow of the Taliban government that had sheltered al-Qaeda, the loss or capture of many members, tighter finance and aviation controls, and years of intelligence pressure. Osama bin Laden was killed in Pakistan in 2011. Judged as an operation, 9/11 succeeded. Judged as a route to al-Qaeda's declared political order, the result is far harder to call a success.

The United States and allies named a global “war on terror”, invaded Afghanistan and built new systems of surveillance, detention, intelligence sharing and border control. Some disrupted plots and reduced opportunities. Others produced lasting legal and moral damage. Torture, secret detention and indefinite imprisonment weakened claims of lawful distinction. A tactic had been granted the status of an enemy that could be fought everywhere, which made the limits of the campaign difficult to state.

The quieter problem was measurement. A prevented attack leaves no public event against which a power can be judged, while every failure is visible. Institutions therefore had incentives to retain tools whose marginal value was hard to prove and whose costs accumulated elsewhere. Temporary architecture became normal travel, policing and data practice, even as the organisations and methods it had been built against changed.

Al-Qaeda's structure also changed. Central leaders still issued guidance and sought major operations, but regional affiliates and self-starting cells became more important. The 2004 Madrid train bombings and the 2005 London transport attacks were carried out by locally based networks with varying external contacts and a wider jihadist influence, rather than centrally dispatched teams on the 9/11 model. This created a harder intelligence problem: the ideology could cross borders faster than command, while local friendships supplied trust. It also exposed the danger of treating every sympathetic statement as operational membership.

The 2003 invasion of Iraq was not an al-Qaeda operation, but it transformed the movement's environment. State collapse, occupation, prison networks and sectarian conflict created opportunities that Abu Musab al-Zarqawi's organisation exploited. Al-Qaeda in Iraq attacked coalition forces, Iraqi institutions, Shia civilians and rivals, often seeking retaliation that would widen sectarian war. Its violence repelled many Sunni communities, contributing to the tribal and American campaign that badly weakened it from 2007. The organisation survived its setback through new names, altered leadership and another war next door.

Islamic State, territorial rule and dispersed imitation

The Syrian uprising and civil war after 2011 opened space for the Iraqi network to expand. In 2014 Islamic State seized Mosul, declared a caliphate and governed millions of people across parts of Iraq and Syria through administration, taxation, coercion and mass atrocity. It was more than a terrorist cell and less than a normal state. Terrorism served territorial war, sectarian cleansing, intimidation and global recruitment.

Its propaganda made violence portable. Professionally edited media displayed rule, belonging and brutality to different audiences. Foreign recruits travelled to its territory. Affiliates adopted the name in existing conflicts. Supporters elsewhere were encouraged to act with limited direction. The organisation could therefore lose central control while continuing to produce attacks, claims and imitation.

A military coalition and local forces dismantled the territorial caliphate, culminating at Baghouz in Syria in 2019. That defeat mattered. It removed population, revenue, training space and the strongest proof behind the promise of a functioning state. It did not eliminate clandestine cells or affiliates. In parts of Africa and Asia, the Islamic State label became attached to local wars with their own histories and grievances.

Al-Qaeda followed a different adaptation. Its leadership had long mixed central ambition with regional relationships. Pressure encouraged affiliates and looser inspiration. The two movements competed over authority and method while sharing enemies and some ideological ancestry. A single “global jihadist” box concealed organisational rivalry, local autonomy and disagreement over how much violence against Muslim civilians would destroy support.

The attacker without a headquarters

Far-right terrorism also moved through transnational imitation. The 2011 attacks in Norway and the 2019 attack on mosques in Christchurch joined local hatred to an imagined global audience. Manifestos, live or recorded media, online subcultures and references to earlier attackers helped create a chain in which perpetrators sought both killing and posthumous influence. Racist, antisemitic, anti-Muslim, misogynist and anti-government ideas mixed differently across cases.

Calling such perpetrators lone wolves made them sound self-created. Many lacked direct command, but they operated inside social fields of texts, jokes, enemies, heroes and tactical precedent. Some communicated with others, announced intent or left warning behaviours. Others remained difficult to distinguish from the far larger population expressing ugly ideas without planning violence. Prevention faced a low base-rate problem: broad traits create too many false positives, while specific conduct may become visible late.

The same blurred edge appeared beyond a single ideology. Online inspiration, conspiracy theories, personal grievance and fascination with violence could combine without a stable doctrine. British legal review of 2024 noted that terrorist violence was shading into violence pursued for its own sake. The category still required an ideological or political cause, but the cause could function as glue around motives that were personal and unstable.

The current map

The global map in the mid-2020s is dominated less by spectacular attacks in wealthy capitals than by violence embedded in conflict. The Global Terrorism Index 2026 recorded 5,582 deaths and 2,944 incidents in 2025, with a heavy concentration in Pakistan, Burkina Faso, Nigeria, Niger and the Democratic Republic of the Congo. Its figures come from one event dataset and one definition, but the broad picture is clear: terrorism now causes most recorded deaths where insurgency, weak state reach and armed competition already exist.

Europe presents another scale. Europol's 2026 report counted 45 completed, failed or foiled attacks reported by European Union member states for 2025: 22 completed, 20 foiled and 3 failed. It recorded 486 terrorism-related arrests. Those figures combine jihadist, left-wing and anarchist, right-wing, separatist and other legal categories. They are useful for policing trends, not directly comparable with a worldwide conflict dataset.

The Hamas-led attack on southern Israel on 7 October 2023 combined mass killing of civilians, hostage-taking and assaults on military positions. The subsequent war in Gaza caused catastrophic civilian harm and a wider regional crisis. The sequence showed the audience-and-response mechanism at extreme scale. It also showed why terrorism cannot substitute for analysis of armed conflict, occupation, state force or international law. The tactic sits inside a history larger than itself.

Across these settings, the operating sequence remains recognisable. An actor selects violence for effects beyond the scene. Organisations or milieus supply meaning and capability. Audiences interpret under pressure. States respond through several institutions. The response changes recruitment, legitimacy, protection and the next set of opportunities. Then the campaign adapts, fragments, ends or moves elsewhere.

How we know

Terrorism data are built rather than found. Researchers must decide whether an incident was political, whether a target was civilian, whether an actor was a state, whether an attack inside civil war belongs in the dataset and how to code disputed responsibility. Different legal systems and databases therefore produce different totals. Foiled plots are visible mainly where authorities report them; abandoned ideas and successful prevention are harder to observe.

Clandestine organisations leave selective records. Court files overrepresent disrupted networks. Propaganda exaggerates strength and coherence. Intelligence archives may remain closed for decades. Interviews with former members are shaped by memory, self-justification and who survived. Famous attacks receive more evidence than routine intimidation in remote places.

Causal claims are harder still. Violence, repression, poverty, occupation and state weakness influence one another. A campaign that ends after a leader's death may already have been declining. A concession may follow attacks, elections, war weariness and diplomacy together. The best evidence compares cases, methods and levels, then keeps setting-specific findings inside their setting. Confidence should be strongest about mechanisms and weaker about universal recipes.

What People Get Wrong

“Terrorism has a settled definition”

There is a stable centre and a disputed edge. Deliberate political violence against civilians or other protected targets, intended to intimidate or coerce a wider audience, sits near the centre. Disagreement begins over states, soldiers off duty, attacks on property, cyber disruption, resistance under occupation and violence inside civil war.

The dispute persists because definitions perform jobs. Police need offences precise enough for court and broad enough for prevention. Governments decide which organisations to proscribe and which foreign partners to support. Datasets need rules that can be applied across thousands of incidents. Movements fight the moral force of the label. Each purpose draws the boundary differently.

The correction is not that terrorism means whatever the speaker dislikes. It is that every use should disclose its rule. Who may commit it? Which targets count? Is motive required? Does state violence enter? A clear definition can be contested and still useful. A hidden one turns classification into propaganda.

The boundary also changes the trend. Include failed and foiled plots, attacks on soldiers, property damage or incidents inside civil war and the total moves. Two reports can describe the same year honestly while counting different objects. Before comparing numbers, compare the rules that produced them.

“Terrorists are mad”

The idea is comforting because it removes politics and social process. If perpetrators are insane, there is nothing to understand beyond pathology. Research has not produced a universal terrorist personality, and tightly organised groups often require members who can keep secrets, follow routines and maintain relationships. Psychosis is not a general explanation for coordinated campaigns.

That does not make mental health irrelevant. Studies of lone actors sometimes find higher rates of disorder, distress or crisis than studies of members inside organisations. The evidence varies with definitions, records and comparison groups. A diagnosis also leaves the main causal work undone. It does not explain why suffering became attached to an ideology, why one target was chosen or how violent scripts were acquired.

The correction matters because profiling illness creates false alarms and stigma. Prevention should respond to concerning behaviour, capability, intent, leakage and context, while mental-health care remains valuable in its own right.

Crisis can still alter timing, judgement and vulnerability to influence. That makes support and risk management relevant in individual cases without turning diagnosis into a political explanation. The hard work is to connect a condition to conduct through evidence rather than allowing a medical word to finish the story.

“Poverty causes terrorism”

Poor governance and lost livelihoods can contribute to conflict, resentment and recruitment in particular places. The mistake is to convert that possibility into a universal portrait of the attacker. Studies of Hezbollah members, Palestinian public opinion and cross-national event patterns found no simple rule in which poorer or less educated people were more likely to engage in terrorism. Organisations may select educated recruits because clandestine work demands reliability and skills.

The opposite slogan, that poverty never matters, is no better. Economic exclusion can weaken trust, make armed income attractive or interact with discrimination and state abuse. Evidence from one movement does not travel automatically to another continent or decade.

The correction is about levels. National poverty, local inequality, household income and the characteristics of selected members are different measurements. Development can be good policy without being sold as a direct anti-terrorism switch. A cause that explains nearly everyone in a deprived region and almost none of the variation in who joins is incomplete.

There is a further cost to the slogan. When schools, jobs and aid are presented mainly as security instruments, people may read ordinary public goods as surveillance or payment for obedience. Development should be judged on human and economic grounds, with any security effect demonstrated rather than advertised in advance.

“Religion explains religious terrorism”

Naming a religious identity can describe the language of a campaign without explaining its emergence. Texts and traditions contain resources for restraint, coexistence, hierarchy, sacrifice and war. Organisations select, interpret and enforce among them. The same religion is lived peacefully by millions and mobilised violently by a small movement in a particular conflict.

Religious framing can change strategy. It may enlarge the imagined community, make compromise look sinful, sanctify death or place victory beyond one lifetime. Those effects deserve analysis. They still operate through institutions, authorities, networks, politics and personal relationships. Secular nationalism and revolutionary ideology have also made violence sacred in practice.

The correction prevents two errors. One is apologetic denial that theology can matter. The other is collective suspicion of believers who share a label but neither the interpretation nor the organisation. Ask who interprets the tradition, for which audience, under what political conditions, and with what power to turn doctrine into action.

Internal disagreement is evidence, not noise. Clerics, families, scholars and rival movements may use the same sources to condemn an attack, narrow its targets or deny the perpetrator authority. A label such as Islamist, Christian or Hindu hides the contest that determines which reading acquires organisation and force.

“Lone actors act alone”

A lone actor normally lacks direct operational partners in the final attack. The phrase does not mean the person invented the ideology, trained without models, communicated with nobody or left no warning. Research has found lone attackers embedded to varying degrees in online milieus, friendships, prior organisations, grievance communities and chains of imitation. Some sought contact and were rejected. Some announced intent to relatives or acquaintances. Some were socially isolated but culturally connected.

The correction is not to invent a hidden command behind every individual. Many attacks are planned and conducted without organisational direction. It is to separate operational solitude from social formation. That distinction changes prevention. Investigators look for conduct and relationships rather than a headquarters that may not exist. Platforms and media consider how notoriety travels. Communities are not made responsible for every person who consumed the same material.

“Lone wolf” is vivid and misleading. Wolves are social animals, and political violence usually borrows a pack even when the attacker never meets it.

Leakage must also be handled soberly. Many people voice rage, make disturbing jokes or consume extremist material without attacking anyone. A warning becomes useful through specificity, change, capability and context. Retrospective certainty about an overlooked clue is easier than prospective discrimination among thousands of alarming fragments.

“Terrorism either works or it does not”

The answer changes with the outcome being measured. An attack may kill the intended person, obtain coverage, attract recruits, keep an organisation alive, win a concession or achieve a final political programme. These are different successes. Research that codes stated strategic goals often finds that campaigns perform poorly, especially when civilians are targeted and demands are expansive. Other work shows that terrorism can prolong wars, derail settlements or help factions dominate rivals.

Causation is also difficult. If a government withdraws after years of attacks, elections, military cost and diplomacy, assigning victory to terrorism alone is an argument, not an observation. If a movement wins office after renouncing violence, the armed campaign may have built leverage, damaged legitimacy or delayed the same outcome.

The correction is to specify actor, objective, time and counterfactual. Compared with what alternative? At what cost to the claimed constituency? A tactic can be effective at producing attention and disastrous at producing the society its users promised.

The claimed constituency belongs in the score. Leaders may survive abroad while residents endure raids, lost investment, communal suspicion and retaliation. An organisation can call continued relevance victory because it has changed the objective from governing to existing. Survival matters, but it should not be smuggled in as fulfilment.

“The hardest response is the safest”

Visible force answers fear with reassurance: someone is in control and someone will pay. Sometimes rapid force is necessary. The mistake is to use severity as a substitute for fit. A territorial armed organisation, a clandestine cell, a radicalising prisoner and an isolated attacker present different problems.

Indiscriminate action can widen the pool of grievance, overwhelm useful intelligence and validate claims of collective persecution. Torture can produce unreliable information and durable legal damage. Permanent emergency powers acquire uses beyond the crisis that created them. Conversely, weak investigation and unprotected targets invite repetition. The choice is not toughness or surrender.

The correction is strategic discipline. Judge a response by lives protected, plots disrupted, evidence secured, support denied, rights preserved and political exits created. Include second-order effects and the behaviour of rivals. The strongest state is not the one that can do anything after an attack. It is the one that can still choose what serves its purpose.

Time reveals whether toughness was theatre or strategy. A raid may disrupt a cell today and multiply silence around investigators next year. A broad power may be justified against one network and retained against unrelated conduct. Every exceptional measure needs evidence, review and a condition under which it ends.

Use It

Draw the three-part map

When an attack occurs, begin with victim, audience and demand. Record who was directly harmed, then list the people whose conduct was meant to change. The audience might include a government, a local community, possible recruits, rival factions, investors, tourists or people identified as an enemy population. Then state the demanded change, even if it is vague or impossible.

This prevents the scene from swallowing the campaign. It also exposes uncertainty. A group may claim an operation it did not conduct, change its explanation after the event or send incompatible messages to different audiences. Do not grant strategic coherence merely because a communiqué sounds certain.

The map guides response. Rescue and justice concern the victim. Public communication concerns the audience. Policy judgement concerns the demand and the incentives created by answering it. Collapsing all three into “do something” gives the perpetrator the advantage of urgency without the burden of a clear plan.

Keep a fourth column for evidence. Separate what investigators know, what the perpetrator claims and what commentators infer. Drama does not prove power, and a claim does not prove responsibility. The column makes uncertainty visible without making it paralysing.

Separate the actor from the tactic

Describe conduct before identity. Did the organisation attack civilians, assassinate an official, fight armed forces, govern territory, extort businesses, contest elections or provide services? Many movements do several of these. “Terrorist group” may be legally necessary and analytically insufficient.

This lens protects two distinctions. First, condemning a tactic does not settle every claim made by the actor. A movement can identify a real injustice and choose an indefensible method. Second, recognising legitimate grievance does not cleanse the method. Moral judgement becomes sharper when cause and conduct are not made to acquit each other.

The distinction also helps with change. An organisation can abandon terrorism while retaining weapons, enter politics while a splinter continues violence, or lose territory and become a clandestine network. Ask what the actor is doing now, not what a fixed noun permits you to assume.

Apply the same discipline to governments. A state can lawfully investigate one offence and unlawfully punish a population in the same week. “Counterterrorism” names a purpose, not a certificate of legality or effectiveness. Conduct still needs its own rule.

Ask which level contains the cause

When somebody offers a root cause, ask what outcome it predicts. Does it explain why one country has more terrorism, why an organisation adopted the tactic, why a community supplies recruits, or why one individual acted? Evidence at one level cannot answer all four.

Poverty illustrates the problem. A poor region may contain weak institutions and armed competition. Members of a selected cell may still be educated and employed. Both can be true because the organisation is a filter between population conditions and individual participation. Religion, repression, foreign intervention and social media require the same separation.

Then look for the bridge. Which network framed the grievance? Who offered a role? What made violence seem effective or obligatory? Which event accelerated movement from sympathy to action? Causes become useful when they identify a process that could have gone differently, not when they place a large label over the place where explanation should be.

Match intervention to that bridge. A policing problem may require evidence and disruption. A prison network may require safe management and credible exit routes. A discriminatory institution may require reform because it is unjust, with any recruitment effect treated as an additional question. One word, prevention, should not hide several different jobs.

Score success in layers

Use a simple ladder: operation, communication, organisation, concession, political end state. Did the immediate act occur as intended? Did the intended audiences receive the message? Did recruitment, support or survival improve? Did an opponent change policy? Did the final political order move closer?

Add time and cost. A campaign may win a concession this month and lose public support over five years. It may preserve an organisation while devastating the community it claims to protect. A government may prevent attacks while creating powers that damage institutions long after the threat changes.

This lens makes debate less theatrical. “Terrorism works” and “terrorism never works” become testable claims rather than identities. It also reveals substitution. If terrorism is compared only with doing nothing, it can appear effective. Compare it with political mobilisation, strikes, civil resistance, guerrilla warfare, negotiation or international advocacy where those alternatives existed.

Write the counterfactual before announcing a result. Would the policy have changed anyway? Did violence accelerate, delay or narrow the settlement? Which other actors paid for the leverage? Exact answers may be impossible, but naming the missing comparison prevents chronology from disguising itself as proof.

Treat the response as the next move

Before supporting a response, write the attacker's hoped-for sequence. Perhaps the state is meant to punish a whole community, close political space, abandon an ally, cancel an election, overextend abroad or elevate a marginal group into the recognised voice of its cause. Then ask whether the proposed action completes that sequence.

This does not create a veto for the attacker. A policy can be necessary even when the enemy predicted it. The exercise forces a second question: how can the same security aim be achieved while reducing recruitment, polarisation and legal damage? Precision, evidence, review, explanation and a defined stopping rule are strategic assets.

Apply the test to publicity too. Repetition can inform the public and spread the perpetrator's chosen image. Naming can aid accountability and create fame. Silence can reduce amplification and invite rumour. There is no automatic rule. The useful standard is whether coverage increases understanding without turning the attack into a recruitment product.

Price rare risk honestly

Terrorism creates pressure for zero risk, a standard no open society can meet. Unlimited prevention would require unlimited surveillance, control and false positives. The harder question is which protections reduce expected harm enough to justify their financial, social and legal costs.

Begin with base rates. A warning sign common among millions but present in a handful of attackers is a poor screening tool. Combine indicators, update them with behaviour and preserve human judgement. Audit who is stopped, searched, referred or excluded, because a system can impose concentrated harm while reporting a small average burden.

Proportion does not mean minimising victims. It means protecting people without allowing vivid danger to consume every other value. Measure attacks prevented where possible, but also delay, displacement, public confidence, discriminatory effects and whether a measure still has a purpose after technology and threat have changed.

Give every expensive precaution an owner, a review date and a condition for change. Otherwise absence of attack becomes proof that the measure worked, while any attack becomes proof that more is needed. A policy that can be confirmed by every possible outcome is protected from evidence rather than protecting the public.

The limits

This framework cannot tell a police officer whether a particular tip is true, a minister whether a hostage concession will save lives, or a negotiator when an armed faction can keep a promise. Intelligence is incomplete and adversaries adapt. Ethical duties do not collapse into strategic calculation. Some acts remain wrong even when useful, and some lawful options remain foolish.

The audience model can also flatter perpetrators by making incoherent violence sound designed. Groups miscalculate, internal factions compete and individuals attach political language to personal cruelty. Not every public effect was intended. Causes reconstructed after an attack can become tidy stories assembled from selective evidence. Keep contingency and incompetence in view.

Finally, terrorism is one slice of political violence. A focus on non-state attacks can obscure larger civilian harm from war, repression and communal killing. The category earns attention because of its mechanism, not because its victims are the only victims who count.

The one thing to keep

Keep control of the next move.

Terrorism tries to turn another person's violence into your instruction. Be afraid of this group. Blame that community. Grant this organisation the status it claims. Give the state any power it asks for. Strike now. Withdraw now. Stop gathering. Stop travelling. Watch the image again. The commands differ, but the mechanism is constant: limited violence seeks authority over decisions made elsewhere.

The correct response is not a pose of calm. It is disciplined separation. Care for the people harmed. Find and stop those responsible. Examine the political claim without allowing it to justify the method. Use the instrument fitted to the organisation that exists, not the enemy imagined in a speech. Protect law because legitimacy produces cooperation and because powers built for emergencies rarely remain there. Address conditions that armed movements exploit without pretending that grievance makes violence inevitable.

A small actor gains leverage when a larger audience reacts as one frightened body. The permanent change should be this: after the shock, look for the script. Who wrote a role for you, and what happens if you refuse it?

Terms

Terrorism. Violence or threats used for political or ideological effects beyond the immediate victim, often through intimidation, coercion or attacks on civilians. Definitions differ over actors, targets and harm.

Political violence. Harm or coercion connected to struggles over authority, territory, identity or policy. It includes terrorism, insurgency, repression, communal violence, assassination and some warfare.

Civilian. A person outside armed forces or an organised armed group. International humanitarian law protects civilians from direct attack, though classification can be difficult in irregular conflict.

Noncombatant. Someone not taking a direct part in hostilities, including civilians and personnel who are wounded, detained or otherwise out of combat. Some definitions make it broader than civilian.

State terror. Exemplary or systematic violence by government agents intended to frighten a population into obedience. The modern word terrorism emerged from revolutionary state terror before shifting towards non-state violence.

Insurgency. An organised struggle challenging state authority through political mobilisation and armed action. Insurgents may govern territory and fight security forces; some also use terrorism, but the categories are not identical.

Guerrilla warfare. Irregular combat based on mobility, concealment and selective engagement with stronger forces. Guerrillas primarily attack military or governmental targets; terrorism primarily works through wider audiences and protected targets.

Violent extremism. A policy term for beliefs or movements supporting ideologically motivated violence. Its breadth can aid prevention, but vague use risks merging radical opinion with crime.

Radicalisation. A contested term for movement towards beliefs or commitments that may support political violence. Belief does not progress through one fixed sequence, and radical ideas do not reliably predict violent action.

Disengagement. Leaving violent activity, a role or an organisation without necessarily abandoning the underlying ideology. Changed relationships, fatigue, fear, family and disillusionment can all produce behavioural exit.

Deradicalisation. A change in beliefs away from political violence. Programmes may combine ideological dialogue, psychological support, education, family work and practical help, with uneven evidence.

Lone actor. A person who plans and conducts an attack without direct operational partners. The label says little about ideological influences, online contacts, prior group links or people who may have noticed warning behaviour.

Cell. A small operational group whose limited membership and compartmented contacts reduce exposure if one member is identified. Cells can receive central direction, operate semi-autonomously or emerge from existing relationships.

Network. A pattern of relationships through which information, trust, money, people or authority move. Networks can survive the loss of a node, yet weak central control may reduce coordination and restraint.

Affiliate. A regional or local organisation formally associated with a larger movement or brand. Affiliation can bring prestige and support while leaving strategy, command and priorities substantially local.

Safe haven. A place where an armed organisation can recruit, train, plan or recover with limited interference. It may result from sponsorship, terrain, civil war or weak state reach.

Propaganda. Organised communication intended to shape perception and conduct. Terrorist propaganda frames enemies, claims attacks, recruits supporters and interprets losses; counterterrorist communication can inform or unintentionally amplify it.

Assassination. The targeted killing of a prominent person for political purposes. Its limited physical scope can carry large symbolic effect, but removing one leader rarely guarantees the institutional change sought.

Hostage-taking. Seizing or detaining people to compel another actor, attract attention or secure exchange. It turns the continued safety of particular victims into pressure on governments, employers, families or publics.

Hijacking. Unlawful seizure or control of a vehicle, especially an aircraft, for coercion, movement or spectacle. Conventions and security changes later made aviation hijacking harder.

Suicide attack. An operation whose execution depends on the attacker's expected death. It can signal commitment and remove escape requirements, while remaining an organisational tactic shaped by recruitment, authorisation and propaganda.

Improvised explosive device. A non-standard explosive weapon assembled outside regular military supply. The term says nothing by itself about motive, target or whether an incident is terrorism.

Attrition. A strategy intended to convince an opponent that continuing resistance will cost more than concession. Repetition and apparent endurance matter because the group is signalling future pain, not one isolated event.

Intimidation. Violence used to make a population obey, stop cooperating or doubt state protection. Local officials, witnesses, voters and minority communities can be targeted so that fear reorganises ordinary behaviour.

Provocation. A strategy seeking an excessive or indiscriminate response that harms the population from which an organisation wants support. The state is invited to make the group's polarising story more credible.

Spoiling. Violence intended to break negotiations or trust between parties approaching settlement. Spoilers make cooperation look unsafe, expose moderates to accusations of betrayal and try to restore armed actors to the centre.

Outbidding. Competition in which factions use visible violence to prove greater commitment or capability to a shared constituency. Escalation can strengthen one organisation within a movement while damaging the movement's political prospects.

Counterterrorism. The combined use of intelligence, policing, justice, protection, finance, diplomacy, prevention, communication and sometimes military force to reduce terrorist capability, opportunity, support and political effect.

Proscription. A legal decision to ban or list an organisation, usually creating offences involving membership, support, finance or display. It can restrict networks while raising difficult questions about evidence, speech and political change.

Resilience. The capacity of people and institutions to absorb an attack, care for those harmed, restore ordinary life and adapt without granting the perpetrator lasting control over public behaviour or law.

Go Deeper

The clearest first book

Louise Richardson, What Terrorists Want: Understanding the Enemy, Containing the Threat (2006). Begin here for a disciplined account of motives, organisational survival and response written for a general reader. Richardson resists the comforting idea that perpetrators must be incomprehensible, then asks what they seek from revenge, recognition and reaction. The book grew from the post-9/11 debate and some examples now need updating, but its separation of explanation from excuse remains unusually useful. It is readable enough to continue straight after this one and broad enough to reveal where compressed arguments deserve a second pass. Keep a note of which motives explain entry, persistence or strategy, because the distinction is easy to lose.

The history and the machinery

Bruce Hoffman, Inside Terrorism, third edition (2017). This is the established large survey of definitions, historical development, organisations, tactics, media and the changing transnational threat. Hoffman is especially good on the way clandestine violence becomes communicative strategy and on the organisational requirements behind spectacular attacks. The coverage is dense and sometimes closer to a reference work than a continuous argument. Read it with the state-terror question visible, because its principal category is non-state terrorism, and remember that developments since 2017 have altered the current map without invalidating the historical structure. Its index also makes it a useful book to revisit case by case.

The strategic model

Andrew H. Kydd and Barbara F. Walter, “The Strategies of Terrorism” (2006). This journal article gives the five-part framework used throughout this book: attrition, intimidation, provocation, spoiling and outbidding. It treats terrorism as costly signalling and asks which audience each act is meant to persuade. The formal argument is compact but demanding, and the categories should be treated as an influential model rather than a complete law. Read it because it turns an appalling event back into a sequence of expected reactions, which is the first step towards preventing those reactions from becoming automatic. Test each category against a campaign you know and note where several logics overlap.

The response

Richard English, Terrorism: How to Respond (2009). English starts from a bracing limit: individual campaigns can end, but terrorism as a recurrent tactic cannot be abolished. He therefore judges policy by containment, protection, political understanding, discrimination between threats and preservation of the values being defended. The book draws strongly on Northern Ireland while ranging more widely, so its prescriptions should be tested against different organisational and conflict settings. Its age is now useful as well as limiting. You can compare its warning against an unlimited “war on terror” with what followed and ask which emergency choices proved temporary in name only. It is best read with a recent official threat report beside it.

Notes and Sources

The Whole Thing in One Page

Working definition. No comprehensive definition has secured universal agreement. The manuscript uses a family resemblance centred on politically or ideologically directed violence or threat, wider-audience coercion or intimidation, and civilian, noncombatant, protected or symbolic targets. This combines the strategic emphasis in Kydd and Walter with the definitional histories in Schmid, Hoffman and English. It does not claim that every law, dataset or scholar uses the same boundary.

Five strategic logics. Attrition, intimidation, provocation, spoiling and outbidding come from Kydd and Walter. Their model treats terrorism as costly signalling to an enemy or a population whose support is contested. The manuscript presents the categories as recurring logics that can overlap, not as an exhaustive taxonomy or proof that every attack follows a coherent design.

Causes. The rejection of poverty, religion, mental illness or personality as single master explanations rests on Crenshaw, Krueger and Malečková, Piazza, Victoroff, Horgan, della Porta, McCauley and Moskalenko, and Gill and colleagues. These sources differ in cases and method. The common defensible claim is negative: no simple demographic profile or universal linear pathway can identify who will commit terrorism.

Why You Should Care

Global figures. The Institute for Economics & Peace's Global Terrorism Index 2026 reports 5,582 deaths and 2,944 incidents for calendar year 2025, the lowest totals in the dataset beginning in 2007. It draws on Dragonfly's TerrorismTracker and its own coding rules. These are dataset-specific recorded totals, not a natural census of all politically motivated violence. The report also shows strong concentration in conflict-affected countries. Figures from earlier editions based on different underlying data should not be joined into one seamless series without checking compatibility.

Risk and attention. The claim that intentional, spectacular violence can dominate attention beyond its statistical frequency is a general mechanism, not a comparison declaring terrorism less important than another cause of death. Local exposure differs sharply. People living amid repeated insurgent violence face a different risk environment from distant publics encountering rare attacks through media and security systems.

The Core Ideas

Victim and audience. The distinction between immediate victims and wider targets is standard in strategic definitions of terrorism and is developed clearly by Hoffman, Richardson, Schmid, Kydd and Walter, and English. The manuscript avoids the stronger claim that fear is always the sole or principal intended effect. Publicity, intimidation, polarisation, recruitment, spoiling and organisational competition may matter more in a given campaign.

11 September 2001. The figure of 2,977 victims excludes the nineteen hijackers and follows the official 9/11 Memorial count. The account of al-Qaeda's organisation, preparation and American institutional failures draws on the 9/11 Commission Report. Claims about al-Qaeda's wider strategic expectations are controlled by captured documents analysed by Lahoud and by the organisation's own public declarations, while recognising that leadership intention and later consequence are not the same thing.

Definition and international law. United Nations Security Council Resolution 1566 identifies criminal acts intended to cause death or serious bodily injury, or hostage-taking, where the purpose is to provoke terror, intimidate a population or compel a government or international organisation, in the context of existing terrorism conventions. It is evidence of partial convergence, not a universally binding comprehensive definition. The United Nations has instead built a sectoral treaty system around acts such as hijacking, hostage-taking and terrorist bombing.

United Kingdom law. Section 1 of the Terrorism Act 2000 includes action or threat designed to influence government or intimidate the public for a political, religious, racial or ideological cause, covering serious violence, serious property damage, danger to life, public-health risk and serious interference with an electronic system. The statutory language has been amended over time and must be read with current case law and human-rights duties. It is used as one example of the wider preventive reach of domestic law.

States and terror. The word's French Revolutionary origin supports the historical reality of terror from above. Modern academic datasets often exclude state violence to create a more manageable non-state category; international criminal law, humanitarian law and research on repression classify much state harm elsewhere. The manuscript neither calls every abusive state act terrorism nor treats government status as proof that terror is impossible.

Organisation. The account of hierarchy, networks, franchises, cells and social support draws on Hoffman, della Porta, Cronin, Horgan and the organisational literature assembled by Schmid. The claim is functional: repeatable campaigns require relationships that generate trust, resources, authorisation, knowledge or legitimacy. It does not mean every sympathiser, relative, donor or community institution knowingly enables violence.

Lone actors. Gill, Horgan and Deckert's study of 119 lone-actor terrorists found varied social, behavioural and preparatory signals rather than a single isolated type. Later research has extended and contested parts of the evidence. “Lone actor” is retained as an operational description, with the limitation that online participation, prior contacts, imitation and leakage can make solitary execution socially connected.

Finance. The distinction between low-cost individual attacks and resource-intensive territorial organisations follows the difference between operation cost and organisational maintenance. Financial intelligence can expose facilitators and constrain networks, but the manuscript rejects the claim that all terrorism depends on large international transfers. The exact importance of finance is organisation-specific.

Causes at several levels. Crenshaw separates enabling conditions from precipitants and strategic organisational choice. Della Porta connects political environment, group interaction and biography. Horgan argues for pathways rather than profiles. McCauley and Moskalenko describe mechanisms that can operate at individual, group and mass levels. The manuscript's “stack” is a compression of this multi-level literature, not a new quantitative model.

Poverty and education. Krueger and Malečková found little support for a simple individual-level link in the cases they examined, and Piazza found no robust general cross-national poverty effect in his sample. These findings do not show that livelihoods, inequality or public services never affect recruitment or conflict. Population conditions, local opportunity and the characteristics of selected members are different measurements.

Mental health. Victoroff's review found the psychological evidence too weak for a universal profile. Organised-group samples and lone-actor samples are not interchangeable. Some research on lone actors reports elevated rates of diagnosed disorder or acute crisis, but diagnosis does not by itself establish ideological meaning, intent, target choice or causation. The manuscript therefore rejects both blanket pathologising and blanket denial.

Religion. Religious ideas can shape authority, sacrifice, enemy definition and time horizons. Gerges, Hoffman, Horgan and the wider radicalisation literature also show that organisations and political conflicts select and apply those ideas. The book does not treat religion as decorative, nor does it infer collective responsibility from shared identity.

Waves. Rapoport's four-wave model begins with the anarchist wave from 1879, followed by anti-colonial, New Left and religious waves. The framework illuminates generational diffusion and international energy. Dietze offers a different account of the nineteenth-century invention of terrorism through transnational imitation. The manuscript retains the wave model as a guide while making ideological overlap and local chronology explicit.

Effectiveness. Abrahms found low rates of success against stated policy objectives in his sample and argued that civilian targeting damages coercive effectiveness. Fortna found that rebel groups using terrorism in civil wars generally performed worse than other rebels on larger political outcomes. English separates several dimensions of success and shows why yes-or-no verdicts mislead. These studies use different universes, coding decisions and counterfactuals; none proves that terrorism has no effects.

Leadership removal. Jordan found older and larger organisations more resilient to decapitation. Price found stronger average mortality effects in a different dataset and design. The disagreement is retained because it changes policy: leadership removal is neither a universal cure nor a useless gesture. Age, bureaucracy, succession, popular support, state capacity and concurrent pressure affect results.

How groups end. Jones and Libicki coded 648 groups active between 1968 and 2006 and found that policing and intelligence or entry into politics accounted for far more endings than military force in their sample; only a small minority achieved victory. Their definitions, end dates and treatment of organisational change are contestable. Cronin provides a broader typology of decapitation, negotiation, success, failure, repression and reorientation. The manuscript uses the mechanisms and avoids presenting one dataset's percentages as universal law.

Responses and backfire. Dugan and Chenoweth's analysis of the Israeli-Palestinian setting and Gill and colleagues' study of counterterrorism killings and Provisional IRA bombings show that coercive and conciliatory measures can have different behavioural effects. These are setting-specific findings, not proof that restraint always reduces violence. They support the narrower claim that response effects must be measured rather than assumed from severity.

Northern Ireland. The account of internment, intelligence, political negotiation and institutional settlement draws on English, official records and the Good Friday Agreement. Internment began in August 1971 and was initially applied overwhelmingly to nationalists on defective intelligence. Its political effects remain debated in detail, but its alienating and recruitment consequences are widely recognised. The later settlement was multi-causal and cannot be credited to either armed pressure or security policy alone.

Military force. The argument is conditional. Where an organisation holds territory, fields forces and administers populations, armed operations may be necessary to remove those capabilities. Territorial defeat does not erase clandestine networks, local grievances, affiliates or ideology. The distinction is illustrated by Islamic State's loss of its last major Syrian enclave at Baghouz in 2019 and its continued activity through cells and regional branches.

Operating history

French Revolutionary origin. Hoffman and Dietze trace the word to the Reign of Terror and its agents. Later usage moved towards violence against states from below. The change was political and uneven rather than a single date on which the old meaning disappeared.

Narodnaya Volya. The People's Will killed Tsar Alexander II in 1881 after previous attempts. The assassination did not produce the expected popular revolution, and the organisation was broken by arrests and executions. Dietze and Rapoport supply the wider transnational history; the case illustrates the difference between operational success and political outcome.

Propaganda by the deed. The phrase covered several anarchist debates rather than one central command. Assassinations and bombings were copied across borders amid mass newspaper circulation and new explosives. The manuscript avoids treating anarchism as a single organisation or implying that most anarchists supported indiscriminate violence.

Anti-colonial campaigns. The King David Hotel bombing, EOKA and the Algerian National Liberation Front are included to show mixed campaigns, contested labels and the difficulty of attributing independence to one tactic. The Battle of Algiers involved FLN attacks and systematic French torture and detention. Algeria's independence followed a wider war and political crisis; no single event proves that terrorism caused victory.

Hijacking and Munich. The late-1960s and 1970s sequence follows Hoffman and the historical record of Palestinian organisations, Dawson's Field and the 1972 Munich attack. Security screening, specialist response and international legal cooperation changed the opportunity structure. The book gives no procedural detail that could aid imitation.

Northern Ireland and ETA. English's Armed Struggle and broader historical scholarship support the Provisional IRA chronology. The Good Friday Agreement was signed in 1998 and weapons decommissioning followed. ETA announced a definitive end to armed activity in 2011 and dissolved in 2018. Its decline involved democratisation, autonomy, policing, cross-border cooperation, civil resistance and loss of legitimacy rather than one decisive mechanism.

Suicide attacks. Bloom, Hoffman and Pape document adoption across organisations and ideologies. The Tamil Tigers' systematic use is important because it prevents a religious monopoly on explanation. The manuscript describes organisational effects without discussing construction, access or target vulnerabilities.

Aum Shinrikyo. The 1995 Tokyo subway sarin attack demonstrates a wealthy closed movement with apocalyptic aims rather than a mass constituency seeking a conventional concession. Casualty descriptions are kept general because totals differ according to whether later deaths and degrees of injury are counted.

Al-Qaeda. The 9/11 Commission Report, Lahoud and Hoffman support the chronology from the Afghan-war milieu through the 1998 embassy bombings, USS Cole and 9/11. Captured documents show strategic concern with American intervention and Muslim audiences, but they do not establish that every later war or policy was designed or predicted in detail by al-Qaeda.

Iraq and Islamic State. Gerges supplies the organisational history from Zarqawi's network through al-Qaeda in Iraq and Islamic State. The 2003 invasion, institutional collapse, prisons, sectarian mobilisation and the Syrian war created opportunities, but no single factor mechanically produced the organisation. The loss of territorial control at Baghouz is distinguished from organisational extinction.

Far-right and mixed-motive violence. The Norway and Christchurch attacks are used to show imitation, manifesto culture and transnational audience construction. The United Kingdom's Independent Reviewer of Terrorism Legislation reported on cases in 2024 where online inspiration, personal grievance, fascination with violence and unstable ideology were interwoven. This is a current UK legal observation, not a global new category.

Current global map. The 2025 totals and country concentration use the Global Terrorism Index 2026. Europol's European Union Terrorism Situation and Trend Report 2026 recorded 45 completed, failed or foiled attacks reported by member states in 2025, including 22 completed attacks, and 486 terrorism-related arrests. These products count different units under different rules and are presented separately. They must not be compared as if one were a regional slice of the other.

7 October 2023 and Gaza. The Hamas-led attack included deliberate killing of civilians, hostage-taking and attacks on military positions. Israel's subsequent war in Gaza caused catastrophic civilian harm and wider regional escalation. The example is used only to show how terrorist tactics, armed conflict, state force, occupation and international law can coexist in one sequence. The book does not attempt a full history, contested legal judgement or casualty account of the subsequent war.

Evidence limits. Event data require judgement about motive, target, actor and responsibility. Official statistics depend on law and reporting. Organisational archives are selective, propaganda is self-serving, and disrupted cases are easier to observe than prevented ones. These limitations support careful scope, not agnosticism about well-documented attacks.

What People Get Wrong

Definition. The correction follows the stable-centre, disputed-edge approach in Schmid and Hoffman. Disclosure of coding rules is more useful than pretending that no classification is possible.

Madness. Victoroff and Horgan support the rejection of a universal pathological profile. Lone-actor evidence is treated separately because organisational selection and social support differ.

Poverty. Krueger and Malečková and Piazza support the rejection of a simple direct law. The text retains economic pathways where evidence and local mechanism justify them.

Religion. The correction distinguishes doctrinal resources from organisational selection and political setting. It rejects both theological reduction and the claim that ideas never matter.

Lone actors. Gill, Horgan and Deckert support the distinction between operational solitude and social formation. The “wolves are social animals” line is explanatory wordplay, not an empirical premise.

Success. Abrahms, Fortna and English support layered measurement. Any verdict depends on actor, objective, time horizon, alternative strategy and cost to the claimed constituency.

Hard responses. English, Dugan and Chenoweth, Gill and colleagues, United Nations strategy documents and criminal-justice guidance support a mixed instrument set and attention to legality and second-order effects. The book does not claim that every conciliatory measure works or every coercive measure backfires.

Use It

The practical lenses are analytic rather than operational. They help readers separate victims, audiences and demands; actor from tactic; causal levels; dimensions of success; first-order protection from second-order reaction; and vivid fear from measured risk. They do not direct surveillance, interviewing, target protection or tactical response. Professional decisions require current law, intelligence, specialist training and accountability.

Terms

Definitions in the glossary follow the manuscript's disclosed conceptual boundary, international humanitarian law where relevant, and the specialist literature cited above. “Radicalisation” and “violent extremism” are marked as contested because official use varies. “Civilian” and “noncombatant” are not interchangeable in every legal or scholarly setting. “State terror” is included to preserve the history and prevent a non-state research convention from becoming a claim about what governments can do.

Bibliography

Primary and official sources

European Union Agency for Law Enforcement Cooperation. European Union Terrorism Situation and Trend Report 2026. The Hague: Europol, 2026.

Hall, Jonathan. The Terrorism Acts in 2024: Report of the Independent Reviewer of Terrorism Legislation on the Operation of the Terrorism Acts 2000 and 2006. London: HMSO, 2026.

Institute for Economics & Peace. Global Terrorism Index 2026: Measuring the Impact of Terrorism. Sydney: Institute for Economics & Peace, 2026.

National Commission on Terrorist Attacks Upon the United States. The 9/11 Commission Report. New York: W. W. Norton, 2004.

United Kingdom. Terrorism Act 2000, section 1, current text checked 4 September 2026.

United Nations Development Programme. Journey to Extremism in Africa: Pathways to Recruitment and Disengagement. New York: United Nations Development Programme, 2023.

United Nations General Assembly. The United Nations Global Counter-Terrorism Strategy: Ninth Review. A/RES/80/286. New York: United Nations, 2026.

United Nations Office on Drugs and Crime. Handbook on Criminal Justice Responses to Terrorism. New York: United Nations, 2009.

United Nations Security Council. Resolution 1566. S/RES/1566. New York: United Nations, 2004.

Modern works

Abrahms, Max. “Why Terrorism Does Not Work.” International Security 31, no. 2 (2006): 42-78.

Bjørgo, Tore, and John Horgan, eds. Leaving Terrorism Behind: Individual and Collective Disengagement. London: Routledge, 2009.

Bloom, Mia. Dying to Kill: The Allure of Suicide Terror. New York: Columbia University Press, 2005.

Crenshaw, Martha. “The Causes of Terrorism.” Comparative Politics 13, no. 4 (1981): 379-399.

Cronin, Audrey Kurth. How Terrorism Ends: Understanding the Decline and Demise of Terrorist Campaigns. Princeton: Princeton University Press, 2009.

della Porta, Donatella. Clandestine Political Violence. Cambridge: Cambridge University Press, 2013.

Dietze, Carola. The Invention of Terrorism in Europe, Russia, and the United States. London: Verso, 2021.

Dugan, Laura, and Erica Chenoweth. “Moving Beyond Deterrence: The Effectiveness of Raising the Expected Utility of Abstaining from Terrorism in Israel.” American Sociological Review 77, no. 4 (2012): 597-624.

English, Richard. Armed Struggle: The History of the IRA. London: Macmillan, 2003.

English, Richard. Does Terrorism Work? A History. Oxford: Oxford University Press, 2016.

English, Richard. Terrorism: How to Respond. Oxford: Oxford University Press, 2009.

Fortna, Virginia Page. “Do Terrorists Win? Rebels' Use of Terrorism and Civil War Outcomes.” International Organization 69, no. 3 (2015): 519-556.

Gerges, Fawaz A. ISIS: A History. 2nd ed. Princeton: Princeton University Press, 2021.

Gill, Paul. Lone-Actor Terrorists: A Behavioural Analysis. London: Routledge, 2015.

Gill, Paul, John Horgan, and Paige Deckert. “Bombing Alone: Tracing the Motivations and Antecedent Behaviors of Lone-Actor Terrorists.” Journal of Forensic Sciences 59, no. 2 (2014): 425-435.

Gill, Paul, John Horgan, Laura Dugan, and Jeffrey Lovelace. “Counterterrorism Killings and Provisional IRA Bombings, 1970-1998.” Terrorism and Political Violence 29, no. 5 (2017): 809-830.

Hoffman, Bruce. Inside Terrorism. 3rd ed. New York: Columbia University Press, 2017.

Horgan, John. “From Profiles to Pathways and Roots to Routes: Perspectives from Psychology on Radicalization into Terrorism.” The ANNALS of the American Academy of Political and Social Science 618, no. 1 (2008): 80-94.

Horgan, John. The Psychology of Terrorism. 2nd ed. London: Routledge, 2014.

Jones, Seth G., and Martin C. Libicki. How Terrorist Groups End: Lessons for Countering al Qa'ida. Santa Monica: RAND Corporation, 2008.

Jordan, Jenna. “When Heads Roll: Assessing the Effectiveness of Leadership Decapitation.” Security Studies 18, no. 4 (2009): 719-755.

Krueger, Alan B., and Jitka Malečková. “Education, Poverty and Terrorism: Is There a Causal Connection?” Journal of Economic Perspectives 17, no. 4 (2003): 119-144.

Kydd, Andrew H., and Barbara F. Walter. “The Strategies of Terrorism.” International Security 31, no. 1 (2006): 49-80.

Lahoud, Nelly. The Bin Laden Papers: How the Abbottabad Raid Revealed the Truth about al-Qaeda, Its Leader and His Family. New Haven: Yale University Press, 2022.

McCauley, Clark, and Sophia Moskalenko. Friction: How Conflict Radicalizes Them and Us. Revised and expanded ed. New York: Oxford University Press, 2017.

Pape, Robert A. Dying to Win: The Strategic Logic of Suicide Terrorism. New York: Random House, 2005.

Piazza, James A. “Rooted in Poverty? Terrorism, Poor Economic Development, and Social Cleavages.” Terrorism and Political Violence 18, no. 1 (2006): 159-177.

Price, Bryan C. “Targeting Top Terrorists: How Leadership Decapitation Contributes to Counterterrorism.” International Security 36, no. 4 (2012): 9-46.

Rapoport, David C. Waves of Global Terrorism: From 1879 to the Present. New York: Columbia University Press, 2022.

Richardson, Louise. What Terrorists Want: Understanding the Enemy, Containing the Threat. New York: Random House, 2006.

Schmid, Alex P., ed. The Routledge Handbook of Terrorism Research. London: Routledge, 2011.

Victoroff, Jeff. “The Mind of the Terrorist: A Review and Critique of Psychological Approaches.” Journal of Conflict Resolution 49, no. 1 (2005): 3-42.