The utility of typology

Typologies have been used throughout the natural and social sciences as a means of classifying apparently related phenomena. ² From the eighteenth-century Linnaean classification of the natural world of plants ³ and the nineteenth-century Mendeleev periodic table of the chemical elements, ⁴ to the twentieth-century classification of personality types in the social sciences, ⁵ their value has been widely recognised. ⁶

Typology has been defined as ‘a purposive, planned selection, abstraction, combination and (sometimes) accentuation of a set of criteria with empirical references’. ⁷ Typologies vary along a number of axes: from the relationship between the type and real-life experience; the degree of abstraction applied; the purpose of the type; the time-span to which it applies; its spatial scope; and the function required of it. ⁸ Their role has been characterised as descriptive, classificatory and explanatory. ⁹ A number of arguments have been made for their utility. These include the clarification they bring to inquiry; the parsimony they promote; their application to social action and policy; and their central relationship to theory.

One of the primary benefits of typology is the greater conceptual clarity they allow. ¹⁰ This is particularly important when it comes to terrorism and political violence, as it ‘frequently involves the interaction and effects of the actions of many persons and collectives involving a multiplicity of motivations, psychological effects and subjective evaluations’. ¹¹ Typologies therefore provide discipline, and enable inaccuracies in a conceptual approach to surface – above and beyond the basic purpose of allowing for the ordering of data. ¹²

Arguably, all knowledge is based on constructs, described as idealised representations of phenomena in the social world. ¹³ Their classification provides a storage system for evidence ¹⁴ – a kind of ‘data-container’. ¹⁵ These break down broad types of phenomena into more convenient units of analysis, ¹⁶ which is particularly useful in furthering the understanding of – in our case – terrorism and political violence. ¹⁷ Described as especially helpful in young research domains, ¹⁸ their ultimate utility with respect to conceptual clarity is said to be due to the fact that ‘concept formation stands prior to quantification’, since ‘we cannot know what to measure unless we know first what we are measuring’. ¹⁹

Associated with this is the parsimonious way in which typologies order complex data. ²⁰ The more straightforward a typology is, the greater is its utility ²¹ – although too much simplicity has its price. However, we have to keep in mind that – unlike in the natural sciences – exact recurrences of social phenomena are rare. ²² Therefore, the aim of constructing a typology – to reduce such events to ideal-types, without oversimplifying them – is highly valuable. ²³

A further advantage of typologies can be seen in their practical application to the social and political realm. Described as heuristics, ²⁴ typologies should provide the most salient information related to compounds of attributes ²⁵ considered important in the phenomena under scrutiny. This is especially pertinent in the realm of policy since increasingly nuanced explanations, in particular of new or developing situations, are facilitated by typologies. ²⁶ Even simple binary dichotomies such as ‘new’ and ‘old’ terrorism have been described as useful to some decision makers. ²⁷ Generally, application to practical policy problems is often considered the mark of quality of a useful typology. ²⁸

There is also a coherent relationship between typology and theory. ²⁹ This is manifest in a variety of ways: in their foundation and development; ³⁰ in the development of theory based on the typology; ³¹ and through the criteria of testing and falsifiability. ³² Some argue further, saying that typologies constitute theory in their own right, ³³ with some noting an iterative relationship between theory and typology. ³⁴

A number of reasons have been given for typology to be considered theory; specifically, that they identify constructs and their interrelationships, and that these can be falsifiable and testable. ³⁵ This last point is a particularly useful attribute of rigorously developed typologies. ³⁶ The multiple constructs identified in typologies invite a number of routes for testing hypotheses, for example through comparing phenomena of the same type and across types, as well as through applying instances that are outside the initially established framework. ³⁷ The latter might actually encourage the development of a new theory capable of incorporating the ‘outlier’. The utility of typology building has been usefully summarized by G.K. Roberts, ³⁸ who noted that typology should aim ‘to discover new relationships among things so ordered, to generate hypotheses, to lead on to the development of theories, and to identify areas for investigation’. ³⁹

However, it should be borne in mind that the utility of the typology is dependent upon the needs of the user. ⁴⁰ In the terrorism field, these are many, and include social scientists, counter-terrorism experts, politicians, lawyers and policy makers, all of whom may have different specifications of what makes a useful typology of terrorist groups. With this in mind, and against the criteria for utility identified above, a wide variety of typologies will now be considered, beginning by outlining some of those developed in other spheres of political violence, in order to position the endeavour of typological development in the field of Terrorism Studies in its wider context.

‘Typologising’ political violence

Political violence is a heterogeneous term covering a wide variety of phenomena. ⁴¹ The multiplicity of behaviours considered under its rubric make its classification a demanding task. The issue of definition and the wide, sometimes contrasting, characterisations of political violence employed are particularly problematic. ⁴² In response, a number of routes have been followed in the development of typologies of political violence.

Where data are of sufficiently high quality, the analysis of political violence as an aggregate category has been advocated. ⁴³ However, where this is not possible, and bearing in mind the variety of types of political violence perpetrated by state and non-state actors, focus on specific forms of violence has been recommended. ⁴⁴ This is considered likely to enhance the explanatory power of the conceptualisation, and often results in deconstruction, creating sub-types of political violence. ⁴⁵

Given the complexity of the phenomena, a simplistic, unitary conceptualisation of political violence is described as more likely to lie ‘in the eye of the beholder than in empirical reality’. ⁴⁶ Therefore, any typology of political violence may be considered a simplification, or exaggeration of reality. ⁴⁷ As discussed above, the utility of any such attempt should be seen, at the broadest level, in its ability to contribute to knowledge growth ⁴⁸ and its facilitation of new findings. ⁴⁹ This can include theory development, ⁵⁰ as well as serving as a basis for data collection. ⁵¹

In the field of political violence, typologies are thought to be a prerequisite for generalising beyond single cases. ⁵² This may be considered particularly helpful with respect to the analysis and prevention of conflict. ⁵³ Hence, typologies of political violence are more than research tools, and can be used in the context of defence. ⁵⁴ For example, where a relationship is posited between a group type and its choice of warfare, the forecasting of likely behaviour may be possible. ⁵⁵ Further, typologies of political violence are considered useful in policy making ⁵⁶ – specifically, in the way they can identify what is causally significant and downplay that which is not. ⁵⁷

Given these benefits, it is perhaps not surprising that the application of typology in the field of political violence is widespread. Typological efforts have focused on assassination, ⁵⁸ civil war, ⁵⁹ coups d’état, ⁶⁰ ethnic conflict, ⁶¹ genocide and politicide, ⁶² hostage taking, ⁶³ insurgency, ⁶⁴ looting, ⁶⁵ resistance, ⁶⁶ revolution, ⁶⁷ riots, ⁶⁸ suicide bombers, ⁶⁹ vigilantism, ⁷⁰ violent non-state actors, ⁷¹ war ⁷² and warlordism. ⁷³

In order to systematise these efforts and provide clarity to that which follows, an updated version of the typology of political violence developed by Schmid ⁷⁴ is now presented. This is illustrated in Table 3.1, and resembles the work of Crelinsten ⁷⁵ . (It appeared in Chapter 2 as Table 2.2, but is worth repeating here, for clarity.) It aims to position the forms of political violence in a wider classificatory system, and provide a broader context in which to position typologies of political violence and terrorism.

This spectrum of political action reflects the interplay between the forces of order and the forces of change in violent and non-violent opposition modes. It focuses on actors and forms of political violence. In this context, forms of violence may be considered tactics, strategies of action or outcomes. In circumstances where the rule of law is upheld, confrontation is seen through opposition politics, and is classified as free persuasive politics. In the second stage of the escalation, pressure politics are applied by one or both sides, which might or might not be legal and/or legitimate. Some forms of government repression might be legitimate; others might constitute ‘crimes of repression’. ⁷⁶ Some of the tactics available to those opposing an oppressive status quo can be qualified as sedition and subversion. ⁷⁷ On the other hand, the conceptualisation of forms of illegitimate behaviour by a ruling power against its constituents can be described as oppression and repression. ⁷⁸ However, in practice there is overlap between these concepts. ⁷⁹

Violent politics incorporates a variety of actions available, to varying degrees, to both sides of the conflict dyad, with the list of potential types of behaviour described as nearly endless. ⁸⁰ For those in power, a greater range of repressive tactics is available, differentiated by the scale of the response to the forces of change. In it most extreme form, national political conflict results in civil war and, where the opposition is victorious, revolution.

In the framework outlined in Table 3.1, the action and reaction of either party are positioned to correspond roughly in line with the strategies available to them. This enables the examination of the tactics in both sides’ repertoire, and the potential symmetries and asymmetries of particular conflict scenarios. Such a conceptualisation is useful, for it can reveal asymmetries in conflict waging, for example when a democratic regime is attacked by tactics from the repertoire of violent politics from the side of a non-state violent actor. Ekaterina Stepanova has also used the concept of asymmetry. ⁸¹ She delineates terrorist groups in line with two criteria: (1) the scope of a group’s aims with respect to global or local concerns; and (2) the position of terroristic action in relation to other forms of violent confrontation, and the extent to which they are used alongside alternative forms of violence. Using these criteria, three functional types of terrorism are identified. The first is ‘classic terrorism of peacetime’. Terrorism of this type is separate from any wider armed conflict, and includes ‘stand-alone’ left- and right-wing terrorism. Second, ‘conflict-related terrorism’ is defined as an ‘embedded’ tactic incorporated into asymmetric armed conflict. Here, those using terrorism are motivated by a particular cause, and their fight is usually confined to a particular region. They use terrorism as one tactic among others, such as guerrilla-type attacks targeting security forces and critical infrastructures. The third type identified by Stepanova, ‘superterrorism’, is described as global in scope, and is seen as a relatively new phenomenon. The aims of those falling in this category are said to be existential and non-negotiable. The term is applicable to Al-Qaeda and some other parts of the wider Salafist-jihadist movement. Characteristics of these three types of terrorism are said to be interactive and connected, with the possibility of types combining with one another, dependent upon circumstances. This fluidity, while disturbing on the one hand, is useful on the other, given the interaction of uses and users of terrorism in a dynamic real-world context. It also highlights the problem of static and strictly delineated conceptualisations of political violence, and the need for typologies to be able to incorporate multiple tactics and forms of political action.

One attempt to do this through the presentation of a spectrum of types of political violence has been developed by R. Dekmejian, and is reproduced in Table 3.2. ⁸² The scale is defined by the magnitude of the violence, and the direction of the violence, which is described as anti-state or state. In the broad category of anti-state political violence, individual terrorism encompasses people acting in isolation against the state and includes terrorist acts by ‘lone wolves’. Subnational terrorist actors whose target is the state are delineated by the author by cause. There are therefore three categories of subnational terrorist group: ethnic nationalists who are acting as a result of affiliation with a particular ethnic group; religious militants, a category that can include cults and fundamentalist zealots; and ideological radicals, including groups that hold extreme left- and right-wing ideologies. Transnational terrorism against states is defined by the location of its operation: where violence is enacted away from the group’s domestic base, or those with a constituency spread across nations, for example Al-Qaeda. State terrorism is said to outweigh the violence perpetrated by non-state actors, and includes state-sponsored terrorism attacking domestic and transnational targets. The most extreme form of political violence described in the spectrum is that of genocide and politicide, and is defined as being carried out by state actors.

While the use of such a spectrum of political violence is helpful, this proposal is limited by its heavy emphasis on terrorism. Of the five types of political violence descrbed, four of these are terroristic, with the fifth being politicide and genocide. This therefore leaves out a considerable number of types of political violence. In addition, the conflation of tactics, motivations, areas of operations and forms of organisation identified in the perpetrators aspect of the design makes comparison across groups less straightforward, and arguably reduces its analytical utility.

The conceptualisation presented in Table 3.1 resembles an approach proposed by Parmentier and Weitekamp, ⁸³ which also incorporates political crimes and serious violations of human rights. While issues of definition remain unresolved, political crimes are considered to be in line with the subjective and objective criteria identified in the Norgaard Principles. ⁸⁴ These describe six characteristics used to ascertain whether a crime is political: (1) the motive of the offender; (2) the context in which the act is committed; (3) the legal and factual nature of the act, including its gravity; (4) the political objective of the act: at whom it was directed; (5) whether it was carried out following an order from a group of which the actor was a member; and (6) the relationship between the act and the political objective – specifically, the proximity of the relationship and its proportionality.

Parmentier and Weitekamp ⁸⁵ describe a number of typologies of political crime, which could be superimposed, on the spectrum of political action outlined in Table 3.1. One typology defines two forms of political crime; purely political and related political offences, ⁸⁶ the second of which is said to incorporate elements of ‘common crime’ or non-political criminality. A second typology contrasts crimes in defiance of political authority, and those used to defend it. ⁸⁷ This was developed into crimes committed against the state, aiming to damage the establishment, and those committed by the state against those considered to threaten it. ⁸⁸ Similarly, Ross ⁸⁹ proposes crimes by the state (non-oppositional crimes) and those against the state (oppositional crimes). Some of these are also reflected in the behaviours listed in the violent regime and violent action elements of the spectrum of political violence described in Table 3.1.

However, according to Parmentier and Weitekamp, ⁹⁰ these various conceptualisations presented by different authors, ⁹¹ neglect two features of modern political violence: (1) the indistinct division between state and non-state actors, which can result in groups of the same type in conflict with one another; and (2) the international nature of much political crime. This is represented in Figure 3.1. In this conceptualisation, societies are considered pyramids of at least three layers: state apparatus, society organised in non-state groups and associations, and individuals. It is hypothesised that the international community is made up of many such pyramids. Political crimes are all criminal activities undertaken by people in any of these layers against people, institutions or organisation in any other of these layers, national and international, where the intentions, context and consequences are political.

Figure 3.1   Types of political crimes, according to S. Parmentier and E. Weitekamp.

Source: S. Parmentier and E.G.M. Weitekamp, ‘Political crimes and serious violations of human rights: towards a criminology of international crimes’. In S. Parmentier and E.G.M. Weitekamp (eds), Crime and Human Rights: Sociology of Crime, Law and Deviance, vol. 9. Oxford: Elsevier, 2007.

While useful in its consideration of multiple actors and their interrelations, this typology may be more accurately characterised as a typology of political actors, rather than political crimes. Its authors claim that it is more extensive than existing conceptualisations. This may be accurate with respect to the level of analysis applied, but it is not clear whether this is the case with respect to political violence itself. Since it fails to provide details of types of political violence, its explanatory and descriptive power is limited in this regard.

Typologies of crime, political crime and political violence

The scope of political violence is considerable, and involves a complex array of levels of analysis and overlapping classificatory systems. Thus, what is presented here provides a flavour of work in the field of typologies of crime and political violence. We will consider typologies of violent crime to position the forms of political crime against ‘common crime’; attention will also be paid to various forms of non-violent protest. Then, the focus will shift to types of political violence – specifically, typologies of assassination – to give an overview of a specific type of violent behaviour. With respect to the actors responsible for the various kinds of violence, typologies of violent non-state actors will be considered; particular attention will be given to warlordism. A number of conceptualisations of war will also be presented, to offer a wider context against which to consider terrorism. Alongside this, the various types of war crime and human rights violations carried out by states will be discussed to provide a balance to the behaviours of non-state actors.

In addressing ‘common crime’ or non-political criminality, the focus will remain on violent and serious crime to provide a broad comparator of conceptualisations in each type of behaviour. In the United States, serious violent crime includes rape, robbery, aggravated assault and homicide. ⁹² In terms of the British classificatory system, violent crime includes robbery, sexual offences, assault and murder. ⁹³ A basic typology proposed in the British Crime Survey (a social survey of victimisation) is based on the relationship between the victim and offender. There are therefore crimes described as domestic violence, involving members of the same household, including partners and ex-partners; mugging, or robbery/attempted robbery; stranger violence, where the victim and offender were not known to one another; and acquaintance violence, when the victim and perpetrator know one another at least by sight. ⁹⁴

Typologies of crime have been developed around a number of criteria. ⁹⁵ These include legalistic ones, where crime is most commonly classified by the degree of punishment it incurs; person centred ones, where criminal classification is on the basis of the personality of the offender; and social typologies, which consider the social context of the offender and the act. ⁹⁶ A comprehensive typology of criminality has been developed by Clinard and Quinney, ⁹⁷ and incorporates nine forms of crime. These include violent personal, occasional property, public order, conventional criminal, political criminal, occupational criminal, corporate criminal, organised criminal and professional criminal.

An adaptation of this ⁹⁸ describes the various types of crime as follows: violent crime, which causes bodily injury or death; property crime, involving the intentional deprivation of the right to property; public order crimes, which disrupt socially acceptable behaviour; occupational crimes, carried out during the course of employment; corporate crime, conducted while carrying out business or deriving from inappropriate business practices; organised crime, perpetrated on an ongoing basis by a group dedicated to its commission; professional crime, which provides the main source of financial support for the offender; and finally, political crimes, which violate the laws of government. In the original iteration, political crimes are described as often committed by those who receive support from a particular constituency, and depend for their acceptance on the degree of (il-)legitimacy ascribed to the government’s policies. ⁹⁹

To enable a contrast to be drawn between violent and non-violent political action, attention will briefly be given to forms of non-violent protest and sanctions. In 2005, Sharp and Paulson ¹⁰⁰ identified 198 manifestations, which are broken down into three groups. The first is non-violent protest and persuasion; this includes petitioning, picketing, holding vigils and participating in marches. The second form is non-cooperation and can encompass activities in the political, economic or social spheres. This is considered more disruptive, and in the social realm can include social boycott and excommunication. In the economic environment, two forms of activity are identified: economic boycotts and labour strikes. The largest class of political non-cooperation can include boycotting elections, civil disobedience and the breaking of diplomatic relations. The most disruptive form of non-violent action is non-violent intervention. This involves deliberately disturbing the status quo using psychological, social, economic or political means, and can involve such activities as sit-ins, non-violent obstruction and the activation of alternative social or economic institutions. It can be coercive but remains non-violent. ¹⁰¹

Where non-violence as moral and principled action is rejected, a considerable number of tactics become available. Some of these are identified in Table 3.1, including assassination. We discuss assassination here as an exemplar form of political violence to which the techniques of typology construction have been applied. Ben-Yehuda ¹⁰² approaches the issue from a temporal and strategic perspective, and utilises the definition of assassination originally proposed by Kirkham, Levy and Crotty: ¹⁰³ ‘an act that consists of a plotted, attempted or actual, murder of a prominent political figure (elite) by an individual (assassin) who performs this act in other than a governmental role’. ¹⁰⁴ It differentiates assassination from terrorism through the importance placed on the targeted character of the act, in contrast to the more random nature of terrorism. Neglecting motivation, the typology developed describes four categories: pre-planning, centring on the decision to carry out an assassination attempt; planning, where a decision to go ahead with the killing is made, but is not followed through; unsuccessful, meaning that an attack was attempted but no injury was incurred by the intended target; and successful, resulting in either partial success, seen in the injury of the victim, or full success, which sees the target killed. The author goes on to examine political assassinations by Jews in Palestine–Israel and delineates the motivation for the acts as revenge, or acting against traitors. Assassinations are described as an alternative form of social justice, ¹⁰⁵ and may therefore be considered as being on the opposite side of a continuum of political action with principled non-violent political action that might ostracise (ban, exclude) but not kill a political opponent at the other end.

Crotty proposed an alternative conceptualisation. ¹⁰⁶ This is built on five types of assassination: anomic, where private reasons motivate the killing of a political figure; elite substitution, an attempt to replace the victim with someone from an opposing group; tyrannicide, or the killing of a despot with the aim of installing a more acceptable leader; propaganda of the deed, which, via the killing, focuses attention on a wider problem; and finally, terroristic assassination. According to Crotty, terroristic assassination can exhibit a variety of forms and be driven by a number of motivations. These include random assassinations hoping to undermine belief in the government, or ‘targeted killing’ aimed at eliminating particular individuals or groups. This form of assassination can also be perpetrated against a larger number of people. In this case, it serves to demonstrate that the government is incapable of dealing with a rebel or insurgent threat, or is meant to undermine the support for the government through fear. It might also serve to try to gain support for the terrorists’ group or cause, or might be used as an instrument employed by a small group to intimidate and hold down a wider population.

Consideration will now be given to those who perpetrate acts of political violence. These are becoming increasingly diverse, and some of the perpetrators have been described as privatised violent actors, ¹⁰⁷ or violent non-state actors (VNSAs). ¹⁰⁸ These have been the focus of a number of authors; the interpretation of two will be presented here. This will aim to give an overview of those groups that may be considered most likely to act against the controlling power and use some of the forms of political violence outlined in Table 3.1. VNSAs are considered to flourish where there is an absence of capacity or legitimacy in the state. VNSAs are diverse in motivation, purpose and function, strength, financing, structure and in their relationship to the state. ¹⁰⁹ A number of key groups have been identified by Williams ¹¹⁰ and include warlords, militias, paramilitary forces, terrorist organisations and criminal groups. Many of these types of group are also catalogued by Sullivan, ¹¹¹ who expands on Mair’s ¹¹² original four types of privatised violent actor: terrorists, organised criminals, rebels or insurgents and warlords, adding private military companies, as well as pirates. Differences in objectives, motivation, strategy and ability are recognised between these groups; yet their common denominator is the instrumental use of violence. ¹¹³

In order to explore one category of VNSAs and provide a group-level comparison for the ensuing discussion of terrorism, some of the contributions to the literature on typologies of warlordism will now be presented. Warlords have been described as differing along a number of dimensions: the size and type of their military force, the degree of legitimacy they hold in a given local area, their sources of finance, and the relationship to their constituency, links to global networks and their relationship to the government. ¹¹⁴ One conceptualisation of warlords positions them along a continuum of armed factions based on the motivation and behaviour of the group. ¹¹⁵ It places ideological movements that wish to change society at one end of the spectrum, and absolute warlords at the other. Between these are groups claiming ‘their’ ethnic rights, whose objective is the promotion of the interests of a particular group, constituting a ‘lesser degree of warlordism’; and ethnic ideological ‘entrepreneur’ factions, where a higher level of violence is exhibited, and warlordism is a marked feature. ¹¹⁶

A further characterisation builds on the work of Hill, ¹¹⁷ who described three types of militia: freelance – most often small groups with little local loyalty; clan, which relates to a particular identity; and personal militias, which act at the behest of a particular leader, often based on ideology or ethnic group membership. Jackson ¹¹⁸ suggests that each type of militia typically has a different kind of leader: freelance militias have a gang leader; clan militias have a ‘traditional’ leader or ‘chief’; while personal militias have warlords at their head.

So far, consideration has been given to typologies of ‘common crime’, forms of non-violent political action, violent political crime, and the various actors who may carry out these forms of crime. We now turn to war, which can be found at the most extreme end of the spectrum of political violence.

To begin with, on the broadest level of analysis, differentiation between ‘new wars’ and ‘old wars’ has been suggested. ¹¹⁹ The change from old to new is supposed to have taken place in the 1980s and 1990s, and incorporates a number of emerging themes in the analysis of war and political violence. It specifies that a new form of conflict can be waged with different actors, tactics and goals. Here, combatants and non-combatants are difficult to define, and aims revolve around ideological cleavage, rather than the traditional focus on conquest of territory. ¹²⁰ This temporal dimension to conflict development is utilised by Pfetsch and Rohloff, ¹²¹ who propose three types of conflict. First, they distinguish international conflicts between states, enacted primarily from 1945 to the late 1960s; second, there is violence perpetrated because of ethnic, religious, regional and national power positions, manifest predominantly until the 1990s; and third, there is what is described as the reprivatisation of violence carried out at a subnational level by gangs and organised criminals.

The actor as the predominant focus of analysis can be found in a number of typologies. One long-established example is offered by the Correlates of War Project (CoW) of David Singer. Recently amended to reflect the changing nature of war, ¹²² the CoW project now classifies three types – interstate war, extra-state war and intra-state war – as opposed to the two types of international war and civil war in Singer’s previous conceptualisation. An alternative approach to an actor-led characterisation is that developed by Vasquez, ¹²³ which proposes to distinguish between ‘wars of inequality’ and ‘wars of rivalry’. The first of these are between unequal parties, and the second between more evenly matched opponents. A typology that combines the type of actor and the purpose of the conflict identifies four types of war: ¹²⁴ (1) interstate or non-interstate; (2) revolution and/or ideology that aim to change the nature of the state; (3) identity or secession, working to change a particular group’s relation to the state; and (4) factional, which includes coups d’état and warlordism.

Rather than simply distinguishing between old and news wars, and following a temporal approach to the classification of war and conflict, Reed ¹²⁵ proposes five generations of warfare. The last of these purports to describe the type of violence practised by Al-Qaeda, thus positioning one of the most recent manifestations of terrorism on a continuum of war mutations (see Figure 3.2). This is an evolving model of conflict, incorporating changing domains, adversaries, force and objectives in a multidimensional representation of war. Five generations are posited, the difference between them being attested to via the question of whether an army from one generation, without significantly larger forces, can defeat an army from a subsequent generation. If not, there has been a ‘dialectically qualitative shift’, and the arrival of a new generation of warfare is implied. ¹²⁶

Figure 3.2   A generational typology of war and conflict, according to Reed.

Source: D.J. Reed, ‘Beyond the war on terror: into the fifth generation of war and conflict’. Studies in Conflict and Terrorism, 31(8), 2008, pp. 684–722.

Beginning with the Peace of Westphalia in 1648 following the end of the Thirty Years War between Catholic and Protestant rulers in Europe, the generations move through the industrial revolution (second generation); the Second World War (third generation); and asymmetrical warfare as used by Mao Zedong and the North Vietnamese in the Vietnam War among others (fourth generation). This is considered to evolve into fifth-generation warfare as practised by actors such as Al-Qaeda, the arrival of which is held to be marked by 11 September 2001. Characteristics of the proposed fifth generation are the almost virtual presence of (often networked) actors in what is described as an ‘omnipresent battlefield’; and second, the fact that fifth-generation actors do not necessarily use military force, but use both multiple-kinetic and non-kinetic types of force.

This conceptualisation boldly positions political terrorism in its manifestation as irregular warfare in a wider historical sweep. By incorporating objectives, domains of conflict, the nature of the actors and the force they apply, it boldly incorporates elements that are separate or excluded in some other typologies. Two common methods of delineating types of war seen in Reed’s typology are through the actors and their objectives. One typology that utilises both has been developed by Holsti, ¹²⁷ who outlines four categories of war: state against state; decolonising wars of national liberation; internal wars fought for ideological reasons; and state–nation conflicts, including wars carried out by particular ethnic, linguistic and religious groups.

Interstate war as described through these typologies has, historically, been one of the most extreme forms of political violence, although it is, in principle, governed by the rules or customs of war codified in the Hague and Geneva Conventions. ¹²⁸ Broadly speaking, international humanitarian law covers two avenues of managing the violence enacted in war: ¹²⁹ rules regarding behaviour during hostilities (jus in bello), covering the means by which warfare is carried out, especially with regard to non-combatants and soldiers incapable of fighting; and the right to declare and go to war, for example as (collective) self-defence (jus ad bellum). ¹³⁰

Where the laws of war are not adhered to, war crimes or other breaches of the laws of war, including crimes against humanity, may have been committed. ¹³¹ According to the 1945 International Military Tribunal of Nuremberg Charter, ¹³² these may include

murder, ill-treatment or deportation to slave labour or for any other purpose of civilian population of or in occupied territory, murder or ill-treatment of prisoners of war or persons on the seas, killing of hostages, plunder of public or private property, wanton destruction of cities, towns or villages, or devastation not justified by military necessity. ¹³³

There are three types of breaches; ¹³⁴ first, non-serious breaches or common violations, the sanction for which is administrative. Then, there are serious breaches, which come in two forms: war crimes, which states are at their discretion as to whether to prosecute; and grave breaches, or serious war crimes, committed against people protected under the Geneva Conventions, for which prosecution is mandatory. ¹³⁵

Political crimes carried out by the ruling power can be described alternatively as violations of international humanitarian law or human rights violations, depending on the context – whether they are committed in a time of war or in peace time. ¹³⁶ Human rights violations have been conceptualised in a number of ways. One legalistic approach posits three generations of human rights. ¹³⁷ The first generation, based on the Magna Carta, the American Bill of Rights and the Rights of Man of the French Revolution, includes freedom of expression, conscience and religion, freedom of the press, freedom of assembly, free correspondence and freedom from arbitrary detention. The second generation, said to have emerged from the nineteenth-century class struggles, incorporates economic, social and certain cultural rights, including the right to work, health care, education and participation in cultural life. Finally, the third generation includes ‘candidate’ rights for groups based on the ‘equal rights of peoples’. ¹³⁸ These include the right of peoples to existence, the rights of minorities to their own culture, religion and language, and the rights of people to humanitarian assistance.

An alternative typology of serious human rights violations is that proposed by Parmentier and Weitekamp. ¹³⁹ This is displayed in Figure 3.3 and uses the quantitative measure of frequency of the act, and the qualitative measure of gravity. This may be considered useful for conceptualising degrees of human rights violations. However, the classificatory device of gravity and frequency would require more detailed explanation to maximise its utility. This may lead to a more fully developed typology that delineates types of human rights violations, rather than the less specified spectrum of seriousness.

Figure 3.3   A (non-exhaustive) typology of serious human rights violations.

Source: S. Parmentier and E.G.M. Weitekamp, ‘Political crimes and serious violations of human rights: towards a criminology of international crimes’. In S. Parmentier and E.G.M. Weitekamp (eds), Crime and Human Rights: Sociology of Crime, Law and Deviance, vol. 9. Oxford: Elsevier, 2007.

In considering only a sample of typologies of crime, non-violent protest and political violence other than terrorism, the variety in levels of analysis, and the differences in the features of particular oppositional phenomena considered important by analysts, are indeed wide-ranging. War, for instance, is ‘typologised’ by some authors along multiple dimensions such as purpose, adversary, force and domain, while other authors focus on the types of actors and the objective of their violence. In moving the discussion on to non-state actors, the multiplicity of types operating that have force or violence in their behavioural repertoire was illustrated. This was further highlighted by a more detailed examination of warlords, of which a number of types were presented. Finally, consideration was given to war crimes and human rights violations, to balance the discussion by focusing on the illegitimate or illegal actions of those in power.

Against this background of violence other than terrorism, we now turn to terrorism itself. In the previous discussion, terrorists were represented either as a fifth generation of virtual networked players utilising new forms of force ¹⁴⁰ or as one of a variety of regime and violent non-state actors. ¹⁴¹ Given the potential consequences of terroristic behaviour, and the complexity of the interrelationships between types of political violence, the conceptualisation of terrorism is arguably one of considerable importance. This is illustrated in the numerous attempts made to address this question over the years. The contested state of the definition of terrorism (as illustrated in Chapter 2) makes the relationship to other forms of political violence a complex one. With this in mind, a review of typologies of terrorism will now presented. What follows is an outline of efforts to develop a typology of terrorism as represented in the chapter by Schmid, Stohl and Flemming in the second (1988) edition of Political Terrorism: A New Guide to Actors, Authors, Concepts, Databases, Theories and Literature ¹⁴² This serves to position the typology debate in its historical context before we move on to more recent typological work.

Pre-1988 typologies of terrorism

The typologies identified by Schmid, Stohl and Flemming were delineated into a number of classes, based respectively on the type of international terrorism, the type of actor and the type of political orientation, together with multidimensional and purpose-based typologies. An outline of those considered most useful by the authors of the 1988 review is given, alongside their proposals for typology.

Demarcations of international terrorism through a legislative lens were described by two authors, and focused on the relationship between the actor, the state where they were based, and the victim. Dror ¹⁴³ divided international terrorism into three kinds: imported terrorism, transient terrorism and extra-territorial terrorism. The first was terrorism enacted ‘by and on behalf of aliens’; the second described violence ‘by and against aliens’; while the third considered terrorism targeting ‘external representatives and the symbols of the democracy’. From a prosecutorial point of view, these are useful distinctions. A similar legalistic description was provided by Waugh, ¹⁴⁴ who distinguished between spill-over terrorism of foreign nationals against foreign individuals or property; integrated internal terrorism, where either the target or the terrorist is indigenous; and external terrorism, where the terrorist act takes place outside the boundary of the target government. However, while these are useful for legal purposes, their utility for social science is less obvious, as these categories are unable to contribute to our understanding of terrorism or explain the behaviour of terrorists. Enhancing understanding and explanation of the phenomenon under consideration, however, are key attributes of a useful typology. Yet the difference in approach, illustrated by two legalistic attempts of typologising, highlights the problem of developing a universally helpful, inclusive typology of terrorism and terrorists.

A number of typologies categorised as actor-based were summarized in the Schmid, Stohl and Flemming chapter of 1988, partly based on the difference between state and non-state actors. This allowed differentiation between establishment regime (state or governmental) terror, and non-state, insurrectional agitational terror, ¹⁴⁵ and the parallel differentiation between regime terror and an anti-regime siege of terror. ¹⁴⁶ These two classifications were considered unproblematic, but were juxtaposed with categorisations that implied judgements about the legitimacy of the actors, probably an insufficient criterion of classification. As we saw in Chapter 2, on definitions, some authors also use the concept of ‘terror’ for a certain type of state violence, and ‘terrorism’ for the non-state variant.

In consideration of typologies based on the political orientation of the group, a number of proposals were outlined. These included one differentiating between national or anti-colonial liberation movements, regional or separatist movements, social-revolutionary movements in industrialised countries, defensive associations to protect group privileges, and opposition movements in dictatorial systems where non-violent political change is blocked. ¹⁴⁷ Another, built on the difference between left- and right-wing terrorists whose target is the state, incorporated the following categories: ethnic, religious and nationalist groups; Marxist-Leninsts; anarchists; pathological groups or individuals; neo-fascist groups or individuals; and ideological mercenaries. ¹⁴⁸ At a broader level, delineation between nationalist-separatist (including ethnic); issue; ideological; exile; state and state-sponsored; and religious terrorism was proposed by G. Davidson Smith. ¹⁴⁹

To try to overcome ambiguity regarding the assignment of groups to types, as may be the problem with some of those identified above, ¹⁵⁰ Schmid, Stohl and Flemming ¹⁵¹ proposed categorising actors that target the state and actors that do not. This last category included paramilitary vigilante groups. It was proposed that where vigilantism, described as an understudied phenomenon, fulfils key criteria of terrorism (violence against civilians, or intended to intimidate or otherwise influence third parties), it should be included in typologies. Vigilantes come in three main forms: crime-control vigilantes; social-group-control vigilantism; and regime-control vigilantes. ¹⁵²

A further type of violent actor, who the authors argued should be incorporated into typologies of terrorism, is that of the ‘lone-wolf’ terrorist. ¹⁵³ Wilkinson ¹⁵⁴ subsumed this type under ‘sub-revolutionary’ terrorism, which includes manifestations of terrorism whose aim is less ambitious than state change. This was built upon by Shultz ¹⁵⁵ to produce a categorisation of revolutionary terrorism. Depicted as violence or threat of violence aiming to effect change, it incorporates the following categories: revolutionary terrorism, aiming for fundamental revolutionary change of the politicalsocial process; sub-revolutionary terrorism, which hopes to change parts of the political system; and establishment terrorism carried out by those in power against opposition to their rule. ¹⁵⁶

A number of authors have proposed multidimensional typologies. Merari ¹⁵⁷ considers a group’s target population and operational base, producing a 2 × 2 matrix of potential types of terrorism. Schmid and de Graaf developed a more comprehensive multidimensional typology in the early 1980s. ¹⁵⁸ They proposed a typology incorporating political orientation, motives and actors, as represented in Figure 3.4. Widely used in the literature, ¹⁵⁹ it introduced the category ‘single-issue terrorism’, which has since been applied to non-state actors with a narrow political agenda, such as those who bomb abortion clinics. In subsequent publications, Schmid added the category ‘religious terrorism’, which was absent in his original ‘basic typology of terrorism’ from 1982.

Figure 3.4   A basic typology of terrorism, by A. P. Schmid (1982).

A = state actor; a1, a2 = non-state actors; [b] = proxy actors; B = other state actor.

A final category of typologies reviewed in Schmid et al.’s Political Terrorism (1988) is that of purpose-based categorisations. Using a division into types of purpose, but not into a formal typology, Thornton ¹⁶⁰ split terrorist aims into morale building, advertising, disorientation, elimination of opposing forces and provocation of countermeasures by incumbents. This was subsequently expanded by J. Bowyer Bell, ¹⁶¹ who considered four types of revolutionary terror – organisational, allegiance, functional and symbolic – and two forms of manipulative terror – that which exploits and escalates the impact of the violence, and that which is used in a bargaining situation. Crenshaw ¹⁶² also builds on Thornton’s seminal work and presents a typology incorporating tactical considerations and proximate objectives. This is represented in Table 3.3.

The category of degree of discrimination was first proposed in Karber’s ¹⁶³ typology, which interprets terrorism as a symbolic use of violence as well as a means of communication. Here, the message changes in line with the level of discrimination, hence ‘the more general the target group, the more diffuse the message’. ¹⁶⁴ The degree of discrimination is juxtaposed with the influence of the terrorist act. This is categorised as ‘instrumental influence’, where there is an immediate impact on behaviour because of fear, and ‘affective influence’, which produces a long-term behavioural impact because of identification with the actor. The result is four types of terrorism: coercive bargaining, advertising and recruiting, social paralysis, and social conscience. Three of these were produced in a different effort, ¹⁶⁵ with the last category of social conscience being described as ‘mass casualties’, illustrating a considerable diversion from Karber. ¹⁶⁶

Shultz has developed one typology that incorporates many of the variables identified in the literature. ¹⁶⁷ It utilises causes, environment, goals, strategy, means, organisation and participation, and contrasts them with categories of terrorist group ideology: revolutionary, sub-revolutionary and establishment terrorism. This is a comprehensive approach, and produces an array of potentially useful features of the terrorist group’s origins, purpose and organisation. It is reproduced in Table 3.4. As can be seen, the variables identified are largely the same for (sub-)revolutionary and establishment, which makes it more of a mechanical classification than a content-rich typology.

Through this presentation of the state of the art around 1990, the variety of approaches, levels of analysis and degrees of methodological rigour employed is clear. As already observed, the utility of a typology is in large part dependent on the user and the question being asked of it. Goldaber proposed one illuminating typology that explicitly recognises this ¹⁶⁸ – a typology of hostage takers, presented in Table 3.5. Goldaber’s typology is structured around a number of questions such as: who is the hostage taker? What is the distinguishing characteristic or situation? These questions are tabulated against those whose motivation is described as psychological, criminal or political. These three categories are broken down into sub-categories; for example, ‘political’ includes social protestor, ideological zealot and terrorist extremist. Each type is then described in response to the questions posed. This allows the user to identify the question they need answering and to produce an outline response in line with the specifications of the typology. In our view, this is one of the very best typologies available, offering great usefulness to practising hostage negotiators.

A number of ways of presenting typologies has so far been proposed. A young Austrian scholar under the supervision of Schmid portrays one new endeavour in Figure 3.5. ¹⁶⁹ Löckinger’s typological tree organises the various forms of terrorism into four categories: actors, means and methods, motives, and, finally, by geographic range. These various types are further broken down into sub-types, which describe the major dimensions of terrorism and how they have been conceptualised. This typology helpfully provides a graphic illustration of the different ways in which terrorism may be considered, and allows the positioning of new work into the meta-framework, generating a dynamic representation of the manifestations of terrorism.

Figure 3.5   Löckinger’s typological tree of terrorism.

Source: G. Löckinger, Terrorismus, Terrorismusabwehr, Terrorismusbekampfung. Vienna: Ministry of Defence.

A similar outline of ways in which terrorism and its characteristics can be displayed has been developed by Boaz Ganor ¹⁷⁰ and is depicted in Figure 3.6. Ganor aims to combine categories utilised in terrorism typologies, and uses multiple dyadic juxtapositions as an organising principle. This is a combination of typologies, rather than a typology for the organisation of terrorist groups. As Ganor recognises, its explanatory power is reduced by the scope of the criteria it considers, as the more features of a group considered in any conceptualisation, the less helpful it is to the analyst. This leads to Ganor’s development of the ‘limiting variable’ typology represented in Figure 3.8.

Figure 3.6   Ganor’s classification of terrorist organisations by their characteristics.

Source: B. Ganor, ‘Terrorist organisation typologies and the probability of a boomerang effect’. Studies in Conflict and Terrorism, 31(4), 2008, pp. 269–283.

Figure 3.7   General pattern of causation among the structural causes of oppositional political terrorism.

Source: J. Ross, ‘Structural causes of oppositional political terrorism: towards a causal model’. Journal of Peace Research, 30(3), 1993, pp. 317–329.

Figure 3.8   Ganor’s typology of terrorist organisations.

Source: B. Ganor, ‘Terrorist organisation typologies and the probability of a boomerang effect’. Studies in Conflict and Terrorism, 31(4), 2008, pp. 269–283.

Applying Goldaber’s approach to typology development, what follows is an overview of typologies of terrorism developed in recent years, based on key questions like: who are the terrorists? Why do they utilise terrorism? Where do they stage their acts? When does it change? And how do they do it? If described in single terms and related to more traditional descriptions of typologies, this might be defined as the actor, motivation or purpose, location, time-frame, and organisation of terrorist groups. A number of multidimensional typologies have been developed; these will be allocated to the question deemed to be of greatest relevance to the typology. Thus, while this approach does not make it possible to draw absolutely neat boundaries between types of typology, the hope is that researchers will be directed to the most useful conceptualisations for their needs. We also hope that through this categorisation of the field, gaps in knowledge may be more easily identified.

Who are the terrorists?

As may be observed from the number of typologies outlined in the preceding discussion, actorbased categorisations are a favourite way of classifying terrorism. This is still the case, with a number of the more recent typologies also applying this level of analysis. A key variable is the relationship of the terrorist (group) to the state. Thus, Chakravorti ¹⁷¹ divides terrorism into three types: establishment, anti-establishment and criminal professional. The first identifies the rulers – those holding state power utilising the weapon of terror to enforce social cohesion through forced compliance based on the threat of violence. Anti-establishment terrorism, on the other hand, is directed against the government controlling authority, with criminal-professional terrorism included in this typology because of the often-claimed (and less often found) links between criminal groups and both establishment and anti-establishment terrorism. The very high level of abstraction and the inclusion of actors who may not use terrorism as a sole or primary route to goal attainment mean that its analytical usefulness is somewhat limited.

Looking at the broader global stage, Lizardo and Bergesen ¹⁷² consider the position of the terrorist group in relation to the world system. Thus, groups are either embedded in the structural core, on its periphery, or in between, in the semi-periphery. ¹⁷³ This identifies three actor–target dyads: (1) core actors against governments; (2) peripheral or semi-peripheral actors against peripheral or semiperipheral governments; and (3) peripheral or semi-peripheral actors against core states. These categories are utilised alongside ideological justification to outline three types of terrorism. Ideology is delineated according to historical period, and echoes Rapoport’s ¹⁷⁴ theory of four waves of terrorism. The outcome is a typology of three types. The first of these is described as ‘terror in the core’, where terrorist violence is carried out in core states as a sign of revolt. The second is ‘struggling against oppression’ and, according to Lizardo and Bergesen, is held to constitute most terrorist activity. Type 1 groups often grow from this type of violence, which tends to be located in the semi-periphery or periphery of the global system and targets local governments. The third type refers to ‘the transnational turn’, when groups in the semi-periphery attack core targets across international borders. First seen in the 1960s, when transnational terrorism was used against European countries in the framework of the Palestinian–Israeli confrontation, the current wave of religious terrorism is also assigned to this category. This has increasingly diffuse targets but aims to destroy some features of the implied world system.

The degree of political alienation of the terrorist group from the political process has also been used as a way of classifying such groups. ¹⁷⁵ In this conceptualisation, there is a continuum from activists via militants to terrorists, with each sub-type evincing different levels of alienation, termed medium, high and extreme. Rhetoric and action are used as measures for positioning the group on the spectrum. This use of a continuum is helpful in considering radicalisation processes, allowing phased movement through the scale. It allows the possibility of tracing groups in their progress towards or away from political alienation. If the measures of rhetoric and action were defined more concretely, this would allow the formulation of testable hypotheses as to when and how groups move from one category to another. This, in turn, might have very practical benefits for governments when they are deciding the threshold beyond which an organisation is judged to have become extremist, and likely to engage in terrorism.

The case of narco-terrorism is one that straddles two of the forms of terrorism identified in the typology exhibited in Figure 3.4. ¹⁷⁶ Some groups identified as carrying out terrorism linked to organised crime and others engaged in political terrorism are said to be working together, in particular with respect to narcotics. ¹⁷⁷ This ‘grey area phenomenon’ ¹⁷⁸ or crime–terror nexus ¹⁷⁹ describes the alleged – and in some cases real – logistical or operational links between ideologically motivated terrorist groups and organised criminal enterprises driven more by profit than by power, and is exemplified in the politicised concept of narco-terrorism. The utility of combining conceptual and operational responses to codify the concept of narco-terrorism has been discussed by Schmid. ¹⁸⁰ The view taken was that there are forms of cooperation between criminal organisations and between different terrorist organisations – and further, that some terrorist groups profit from engaging in criminal activities, while there is also some evidence of organised criminal groups using terrorist tactics. Finally, cooperation is seen between terrorist groups and organised criminal groups. However, these links are, in most cases, insufficiently strong to justify a category of narco-terrorism. Furthermore, the legal and conceptual ramifications are likely to result in an unhelpful stretching of the terrorism concept. ¹⁸¹

The category of narco-terrorism has been used especially by political operators in the field of counter-terrorism who sought an opportunity to infuse new life into the ‘war on drugs’ by linking it to the ‘war on terror’. Attempts to link Al-Qaeda’s core to drug trafficking, while numerous, stand on very shaky empirical grounds. ¹⁸² However, in the case of the FARC in Colombia, for instance, the evidence is more solid. In reality, terrorists often use criminal methods to keep their groups alive, and criminal gangs sometimes have recourse to terrorist methods to intimidate sectors of the state or the public. In most instances, however, terrorist groups use organised crime methods as an in-house money-making mechanism rather than risk endangering their security by entering dangerous liaisons with drug traffickers and the like. ¹⁸³

A number of typologies dealing with state or regime terrorism have been proposed, and most often represent the relationship between those carrying out acts of terrorism and sectors of the government. Thus, Pillar ¹⁸⁴ describes state sponsors of terrorism as identified in US law; enablers of terrorism; and cooperators in counter-terrorism efforts. The use of terrorism by states as a foreign policy tool has been described as consisting of five forms. ¹⁸⁵ These include coercive terrorist diplomacy, where overt but often implicit threats are made to demand compliance with a particular goal. Clandestine state terrorism, on the other hand, is covert and involves agents of the state perpetrating acts of terrorism against another state. A proxy group on behalf of the state carries out state-sponsored terrorism; surrogate terrorism is a more remote form of sponsorship involving the enabling of a group leading to an enhanced capacity to carry out terrorism. Finally, there is state acquiescence in the execution of terrorism which, subsequently, is not condemned or opposed by the state because of the existence of mutually shared interests between state and terrorist actors (for more details, see Chapter 4 on theories of terrorism).

This spectrum of support has also been described in the form and degree of assistance a state provides to terroristic behaviour. Richardson ¹⁸⁶ describes this spectrum as moving from complete state control, discussed in terms of state repression rather than terrorism, through the recruitment and training of actors who will act internationally; close control, where the state provides direction to a group; provision of training, finance and a safe haven to an autonomous group; and financial support alone, where the aims of the group and the state coincide.

Why do terrorists act?

The question of why terrorists act is possibly the broadest question, and incorporates typologies previously considered under the rubrics of motivation, cause, purpose, objective and aim. While these are not synonymous, their ability to inform understanding of the reasons underlying terrorist behaviour, and the need for parsimony of presentation, warrant their consideration as a group. There are a plethora of potential purposes and functions that have been attributed to, and claimed by, terrorist groups. An early catalogue of these found at least 20 factors discussed by authors, ¹⁸⁷ ranging from enforcement to obedience, projection of strength, acquisition of popular support and to win recruits, as well as provocation and to terrorise a particular audience. The first two pieces of work presented below are not strictly typologies but are noted here because they are considered to feed into the potential development of terrorist group types by identifying structural causes for group development.

At the broadest level, Ross ¹⁸⁸ looks at structural causes of political terrorism. Other potential causes are recognised as being psychological, and rational-choice-based drivers. However, the empirical focus here is structural. Through examination of the literature, case studies on the causes of terrorism and terrorist organisation development, the author identified, using terminology originally introduced by Martha Crenshaw, a number of permissive and precipitant causes. ¹⁸⁹ Permissive, or ‘deeper systemic conditions that pre-structure and facilitate the presence of the precipitant [causes]’, were found to be in a number of forms. First, geographical location in urban environments tends to facilitate terrorism more than rural settings. Second, the type of political system has an effect, with wealthy democracies being at greater risk of harbouring terrorists. Finally, and most importantly, the level of modernisation of a country was identified as significant. Precipitant causes were found to be social, cultural and historical facilitation of norms allowing terrorism’s development; organisational split and development between more and less extreme elements of a group; the presence of other forms of unrest acting as catalysts; support for the group; failure by counter-terrorist organisations; availability of weapons; and grievances (real and perceived). This last is considered the most important factor. The cumulative picture produced by these causal factors is displayed in Figure 3.7. This acknowledges the centrality of feedback mechanisms, and considers the context of the terrorist group, including supporters and opponents. It may also be considered to fulfil Roberts’ ¹⁹⁰ criteria of being able to identify new relationships among phenomena, having the potential for hypothesis generation and theory development, and being able to identify new areas for investigation. It also has the considerable advantage of being empirically based, giving it a solid foundation.

An alternative attempt to Martha Crenshaw’s mapping of the various causes of terrorism is that of Post et al. ¹⁹¹ Through consultation with experts in the field, they identify four broad categories of causes. These are historical, cultural and contextual features; key actors such as the regime, and opponents and supporters of the group; group and organisational characteristics, for example group processes and structures; and features of the immediate situation, or triggering events. These are deconstructed into 32 variables with key indicators of risk of group radicalisation subsumed within them. This comprehensive approach outlines a plethora of potentially important circumstances in the development of group violence. However, this thoroughness may make it unwieldy for analysts and, while it fulfils many of the criteria for being a useful framework of analysis, it may not be as helpful as more parsimonious conceptualisations. The authors did, however, apply the framework to types of terrorist groups as identified by Schmid and de Graaf, ¹⁹² with the addition of a category for new religions. This produced an overview of those factors considered important in radical group development per type of terrorist group, thus providing a base for empirical testing, and highlighting gaps of knowledge in the field.

A similar attempt to identify key variables was made by Sirseloudi and Schmid, and is represented in Table 3.6.

One concept that recurs in the typological literature is the role of ideology. Paul Wilkinson, ¹⁹³ for instance, categorised terrorist groups by political motivation or ideological orientation, listing them under the headings of nationalism, separatism, racism, vigilantism, ultra-left ideology, religious fundamentalism, millennialism and single-issue campaigns. In addition to these, the author proposed that state terrorism and state-sponsored terrorism be added to the list. Similar conceptualisations have been presented by Bruce Hoffman, ¹⁹⁴ who details ethno-nationalist, left- and right-wing, and religious terrorism. Vasilenko, ¹⁹⁵ in turn, distinguishes between political, separatists, nationalist, religious and criminal terrorism. Cronin ¹⁹⁶ identifies leftist, rightist, ethno-nationalist/separatist and sacred terrorism. These are all considered to have had prominence in particular time periods, with sacred terrorism being the most recent phase. Cronin highlights the difficulty of allocating groups to concrete categories, as they may have a number of motivating ideologies. However, the domination of one is not unusual, and is argued to be related to terrorist group behaviour, violence and development. This allows the possibility of hypothesis testing, and subsequent theory development, as well as directing potential interventions. The treatment of sacred or religious terrorism as a sub-category of ideological terrorism raises a new set of questions, the answer to which is dependent upon on how one defines religion – a concept almost as contested as terrorism, as the following list ¹⁹⁷ illustrates.

• A religion is a unified system of beliefs and practices relative to sacred things, that is to say, things set apart and forbidden – beliefs and practices which unite into one single moral community called a Church, all those who adhere to them (Emile Durkheim).
• Religion is (1) a system of symbols which acts to (2) establish powerful, pervasive and longlasting moods and motivations in human beings by (3) formulating conceptions of a general order of existence and (4) clothing these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic (Clifford Geertz).
• Religion is the sigh of the oppressed creature, the heart of a heartless world, just as it is the spirit of a spiritless situation. It is the opiate of the people (Karl Marx).
• Religion is the daughter of Hope and Fear, explaining to Ignorance the nature of the Unknowable (Ambrose Bierce).
• Religion is what the individual does with his own solitariness (A.N. Whitehead).
• Religions are organised systems that hold people together (John Bowker).
• Religion is the human attitude towards a sacred order that includes within it all being – human or otherwise – i.e. belief in cosmos, the meaning of which both includes and transcends man (Peter Berger).

If religion is defined with John Bowker as ‘organised systems that hold people together’, then the same can be said of nationalism. In that sense, some more chauvinistic forms of patriotism can be considered as secular religions – sharing with most religion a similar obscure core (the core concept ‘nation’ is undefined in international law, nor is there a legal definition of what constitutes a ‘people’). Fettweis ¹⁹⁸ identifies two types: nationalist terrorists and ideological terrorists. The first claim to represent a particular ethnic group and desire territorial gain, whereas ideological terrorists are motivated by political, religious or other ideas, aiming to change society. A number of hypotheses are outlined as a consequence of this, including the identification of strategies, tactics and terrorist group development, as well as potentially successful counter-terrorism approaches. Again, the level of analysis makes it amenable to empirical testing, and allows the generation and testing of hypotheses, although the potential multiple political ideologies subsumed within the ideological terrorist type may limit potential conclusions and their utility. Fettweis’s distinction, however, stands and falls with the definition of nationalism. If it is also seen as an ideology, it becomes problematic. On the other hand, some nationalism is, as indicated above, almost indistinguishable from religious cults. In those cases, aggressive nationalism is, de facto, the secular equivalent of religion, requiring people to sacrifice themselves for their country if necessary.

A typology that combines ideology, goals and recruitment is that developed by Chasdi. ¹⁹⁹ The product is a three-dimensional ‘typology cube’ outlining 36 types of Middle Eastern terrorist groups. These describe the various interrelations between three types of ideology: Marxist-Leninist, religious and Palestinian nationalist; four types of goal: an Islamic state in Palestine and/or other areas in the Middle East, a secular Palestinian state, a Marxist-Leninist state in Palestine and/or other areas of the Middle East, and a religious Jewish state in Israel; and three types of recruitment style: a clan and/or disenfranchised person; followers of a charismatic leader; and Palestinians, Palestinian refugees and/or disenfranchised Palestinians. Eight types of actual group from the possible 36 were identified by the author and were given appropriate designations. This is a strong example of reducing the level of analysis to a region in order to enhance its practical utility. By identifying specific forms of violent actor in the Middle East, considerable gains are made in understanding and explanatory power, providing insight for social scientists and policy makers alike.

Another typology that takes as its subject a specific geographical area is that developed by Leman-Langlois and Broduer. ²⁰⁰ It is described as an ‘operational typology’ based on an analysis of all terrorist incidents, included failed attempts and hoaxes, linked to Canada in the period 1973–2003. It is reproduced as Table 3.7. By using two variables – the scope of desired impact and the time-frame the terrorists have in mind for justifying their violence – four types of terrorism are derived. Demand-based terrorists are generally small groups making specific demands for action. Private justice terror often wants retribution for an act previously carried out. Revolutionary terror demands change at a structural state level, normally directed at states outside Canada but enacted by diasporas within its borders, and restoration terror wishes to re-establish a particular historical period.

The utility of the time-directional category for the justification of action is not clear; and the degree to which this enhances conceptual clarity with respect to terrorist group behaviour is also not immediately obvious. In practice, whether one takes a past Golden Age or a future Utopia as reference point for contemporary action might make little difference. However, the category of desired impact is a helpful one. This conceptualisation also allows the inclusion of hoaxes, as these may be reasonably assumed to have an (albeit short-lasting) intended consequence, just as do fully fledged attacks.

Also responding to the question of why terrorist groups act, a ‘limiting variable’ typology has been proposed. ²⁰¹ Arguably one of the most innovative of the recent typologies, this one, developed by Boaz Ganor, classifies organisations by the variable that limits their terror activities. It includes four groups of organisation, and is displayed in Figure 3.8. The types of groups are delineated by the level of motivation and operational capability they have at any given time. Where these variables both exceed the terror-level threshold, terrorism will occur.

The underlying idea of this model is explained through the ‘terror equation’ and the ‘counter-terror equation’. The terror equation posits that an organisation’s activities are dependent on motivation levels to execute attacks, multiplied by the extent of the organisation’s operational capability to realise that motivation. Motivation is dependent upon a variety of factors, including an organisation’s goals and interests, internal pressures on its leadership, external pressures, morale considerations, the counter-terrorism policies of the states dealing with it, the results of a rational cost–benefit consideration of paths of action, and desire for vengeance. The extent of the organisation’s operational capability is influenced by a separate group of factors, including the offensive and defensive actions taken by its opponent, policy changes by a supportive state, a reduction in supporter assistance, financial difficulties, shortage of weapons, lack of infrastructure and problems with recruitment. Therefore, the counter-terror equation requires the implementation of offensive and defensive measures that will change the cost–benefit outcome of the decision makers in the terrorist organisation by impacting on its operational capabilities and the leaders’ motivations to attack.

These are useful and relatively underdeveloped aspects of typological inquiry, in particular due to the addition of the oppositional actors’ behaviour and its effects. However, the ‘limiting variable’, defined as level of operational capability and extent of motivation, hypothesised to impact on the likelihood of a group committing an act of terrorism, requires further specification as it is a conflation of many other variables rather than a single causal feature. Similarly, the terror and counter-terror equations need fuller explanation. Compressing the various interacting influences on terrorist groups is a valuable exercise. However, it remains to be seen whether these can be supported empirically, and whether it is possible to condense them in the manner implied through the equation metaphor.

Where do terrorists operate?

With respect to the location of terrorist activity, two points of reference recur in the literature: terrorists’ base, and their target. This is seen in the work of Dror ²⁰² and Waugh ²⁰³ referred to earlier, consideration of which appears not to have moved on significantly since their work. The only exception may be the emergence of ‘transnational terrorism’, increasingly incorporated into typological work after it had, in the 1970s, been pushed aside by ‘international terrorism’ – a formula that served well those who wanted to link it to ‘international communism’.

The concept of transnational terrorism is one considered by a number of authors. Reinares ²⁰⁴ highlights the considerable ambiguity over the delineation between international and transnational terrorism. International terrorism is said to encapsulate two aims: first, demands for changes to the structure and power nexus in whole regions of the world; and second, that the terrorist actors and/or their victims and/or ultimate targets are located in a number of different countries. According to one conceptualisation, transnational terrorism involves those who both originate and operate in more than one country over which the target state has no direct jurisdiction. The aims of transnational terrorists are of a lesser scale, aiming to impact upon only a small number of states. Thus, in this conceptualisation all international terrorism is transnational terrorism, while the converse is not true. Since former president George W. Bush vowed to stop all terrorism with a global reach (i.e. able to reach the US mainland), the term ‘global terrorism’ has become popular for describing Al-Qaeda and some of its affiliates. One of its drawbacks is that it implies for this particular organisation a far greater reach than it is capable of. On the other hand, it suggests, correctly, a link to the concept of globalisation.

That transnational terrorism is a new phenomenon is the thesis of Hough’s ²⁰⁵ typology of terrorism. He identifies three levels of terrorism, differentiated by objectives, targeting and geography, generating domestic, international and transnational types of terrorism. International terrorism is then divided into four types: first, ‘pure’ international terrorism, including its use as part of a domestic insurgency; second, that carried out by independent non-state groups or individuals; third, state-sponsored terrorism enacted by those other than the state; and finally, terroristic violence by state agents on the state’s behalf. The author concludes that transnational terrorism is qualitatively different from international terrorism, and constitutes a new type of violence. He uses it with particular reference to Al-Qaeda, and differentiates it from international terrorism because it does not necessarily link objectives to a particular geographic entity.

A similar approach is adopted by Wilkinson, ²⁰⁶ who distinguishes between international terrorism (between individuals or groups from nation-states) and domestic or internal terrorism (carried out by those from within a country). However, Wilkinson points out that where a terrorist group has a reasonably long life, this is a less useful dichotomy, as international backing is often implicated in domestic struggles in the form of financial and other support. This again illustrates the complexity of allocating terrorist groups to particular types, and fulfilling the criteria of exclusivity posited to identify a valid typology. ²⁰⁷ With the spread of ethnic, religious and linguistic diasporas, and the enhanced mobility of people in general, due to cheaper cross-border transportation, distinctions between domestic (national) and inter- or transnational terrorism have arguably lost much of their relevance.

The categories of international and domestic terrorism are expanded by Seger, ²⁰⁸ who also lists regional groups, special interest extremists, and lone terrorists practising leaderless warfare. These classifications are applied to deterrence. For example, Seger argues that special interest groups are deemed more predictable as they have a limited set of specific targets. In contrast, the lone (lone-wolf) terrorist is considered the least predictable and therefore the hardest to identify and potentially deter. The practical application of terrorist group type to potential deterrence is, however, problematic for transnational groups without identifiable territorial bases (which could be hit in retaliation, the fear of such retaliation potentially acting as a deterrent). It is also problematical for those suicide terrorists who welcome a martyr’s death as a route to paradise.

Charles Tilly ²⁰⁹ identifies two criteria by which to partition terrorist groups: whether or not they are specialists in the use of violent coercion; and whether they carry out acts of violence in their own territory or outside of it. This is represented in Figure 3.9 and illustrates four types of ‘terrorwielding groups’: militias, conspirators, autonomists and zealots.

Figure 3.9   A typology of terror-wielding groups and networks.

Source: C. Tilly, ‘Terror, terrorism and terrorists’. Sociological Theory, 22(1), 2004, pp. 5–13.

Militias may be acting on behalf, or in support of, government or, alternatively, be of the nongovernment variety, but in both cases will be specialists in coercive violence operating within their own country. Conspirators tend to carry out operations outside their home territory; autonomists are allied more closely to political groups, but sometimes use terroristic violence in their own areas; and zealots are similar to autonomists but, according to Tilly, tend to operate outside their territory.

A synthesis of the typologies just outlined is presented in Figure 3.10 and aims to provide a formulaic representation of actors and their potential targets. This utilises the concept of violent non-state actors (VSNAs) ²¹⁰ and juxtaposes them with state actors and other VSNAs. While this dichotomy is not clear-cut, it contributes to conceptual clarity and enables an outline of possible interactions to be identified. It views possible actor–target relations across three potential groups across various geographical boundaries. It assumes that terrorism can be carried out against other VNSAs as well as states, and requires a complete specification of terms. However, it may help frame potential routes for investigation in an inclusive and systematic way, assisting with future research and identifying undertheorised areas, or potential threats, once the role of diasporas and cross-border operations are also taken into account.

Figure 3.10   A typology of possible violent non-state actor and state actor interactions across transnational, international, regional and domestic boundaries.

Note VNSA = violent non-state actor;T = transnational; I = international;R = regional; D = domestic.

When does terrorism change?

Time-span-based or periodical approaches to terrorism have seen a number of proposals evinced. Perhaps the best known of these is Rapoport’s ²¹¹ ‘four waves of terrorism’ theory. As will be discussed in more depth in the theory chapter in this volume (Chapter 4), Rapoport identified four qualitatively different types of terrorism. First there was an anarchist wave beginning in the 1880s and lasting for almost 40 years, then an anti-colonial wave lasting for a similar length of time. This was superseded in the 1960s by the New Left wave, a shorter period that ended in the 1990s, with only a few older groups still remaining active. Most recently, a new, religious wave, which began in 1979 with the Iranian Revolution, is posited. The aim of all four types of terrorism is revolution, with a number of audiences implicated in the violence of the groups: foreign terrorist organisations, diaspora groups, liberal sympathisers and foreign governments.

Rapoport’s typology has been widely used and has already undergone partial empirical testing. ²¹² However, the potential for overlap between categories is acknowledged by the author, ²¹³ and highlights the complexity of terroristic phenomena and the associated problems with its characterisation. Also, the potentially static nature of the four-wave thesis has been highlighted, and it is said not to account for the context in which terrorist groups develop. ²¹⁴ As elaborated in Chapter 4, there has also been an attempt to identify a fifth, ‘tribal’ wave. ²¹⁵ This has its origins in the fourth wave, and is ‘identifiable as a wave of millenarian violence contemporaneous with the ongoing religious wave of terrorism’. ²¹⁶ Characteristics of fifth-wave groups include the extreme nature of the violence they commit. In particular, it is posited that should they manage to grasp state power, the result would – as in Cambodia under the Khmer Rouge – be genocide, the ultimate form of anti-civilian violence. However, the identification of such a fifth wave rests, for now, on shaky grounds.

How do terrorists operate?

As with all the questions posed which structure the typologies presented thus far, the area of how terrorists operate offers broad scope for interpretation. Here, it will be approached from two perspectives. First, we shall consider the organisational make-up of terrorist groups; and second, we will look at it from a tactical point of view, looking at the methods and means utilised by terrorists. It is worth noting that the organisational style of analysis was absent from the review of typologies in the 1988 edition of Schmid et al.’s Political Terrorism. This indicates the evolving nature of investigation into terrorist organisations, and the increasing importance of understanding the internal dynamics of groups. ²¹⁷ The classification of terrorist groups by organisational structure has seen a noticeable evolution, arguably reflecting the changing nature of terrorist group make-up. From hierarchical organisations such as the Irish Republican Army (IRA), there has been increasing commentary on networks as appropriate organisational characterisation. An outline of recent efforts will be given here.

Arquilla and Ronfeldt ²¹⁸ first applied the concept of networks to terrorist groups. Crucially facilitated by the increasing availability and sophistication of technology, small groups are able to communicate and coordinate activities without a central command. Decision making and operations are decentralised and depend on constitutive consensus building, allowing for local autonomy. There are different forms, including the chain or line network, where information is passed along a chain; the hub or star network, where groups or individuals are linked into a central point; and the allchannel network, where all nodes are linked to one another. These are illustrated in Figure 3.11.

Figure 3.11   Basic types of networks (Arquilla and Ronfeldt).

Jackson takes a broadly similar approach. ²¹⁹ The genesis of this characterisation is the command and control authority relationships operating in terrorist groups. The premise is that organisations are held together by the links between constituent units, or nodes: individuals at a low level, and components of an organisation at a higher level. These enable the organisation to communicate, exchange information, convey authority and shape the behaviours of others in the organisation. Each node may refer to an individual, group, part of a group, organisation or even a state. ²²⁰ Three different kinds of authority are said to be exerted among group members: strategic, operational and tactical control. Three types of groups are thereby specified: tightly controlled groups, coupled networks and loosely coupled ones. Tightly controlled groups’ commanders exert strategic, operational and tactical control, and can plan complex operations. This is most often found in hierarchies. Coupled networks have looser relationships among units and less direct modes of communication. Loosely coupled movements exhibit a pursuit of general goals based on a common ideology. Described as ‘leaderless resistance’, their operations are characterised by an absence of control at the tactical level, weak influence at the operational level but a strategic embedding by opinion leaders in the social, religious or political movement.

Many authors, in the fields of both criminal organisations and terrorist groups, have embraced the network model rather uncritically. Various definitions of network exist and are inconsistently applied; it is also far from clear what constitutes, for an entity working in the underground, a resilient, functional and efficient network in organisational terms. Mishal and Rosenthal ²²¹ reject the network approach in consideration of Al-Qaeda and propose one based on a metaphor linked to drifting sandbanks: the dune organisation. This is based on time perception, chain of command and control, communication lines, and level of division of labour, and is dependent upon a ‘process of vacillation between territorial presence and mode of disappearance’. Inspired by the de-territorialisation brought about by globalisation processes, the dune organisation is almost random as it moves from one territory to another, changing its characteristics as it goes. While the dune metaphor has some explanatory power for the nebulous waxing and waning of underground formations between heartlands and diasporas, it does not reflect an organisational model.

A further typology is built on the analogy of business organisations. ²²² Here, terrorist groups are based on respective levels of centralisation of resources and centralisation of operations. These two axes result in four types of group, which correspond to business models: (1) hierarchy, (2) franchise, (3) venture capital, and (4) brand. Hierarchy uses centralised operations and resources and is characterised by tight lines of command. Venture capital has decentralised operations and centralised resources, often with isolated and insulated cells. Franchise groups combine centralised operations with decentralised resources. Brand groups have the lowest degree of centralisation, with decentralised resources and operations. They rely mainly on ‘ideological self-identification and ad hoc cooperation’. While underground terrorist organisations are hardly comparable to international business corporations, this typology is actually one of the better ones for explaining the spread of Al-Qaeda.

Kaplan developed a further typology, which combines a variety of features related to the manner in which terrorist groups operate. ²²³ This considers a number of features of terrorist action including tactics, degree of force, agency and geographic context. A graphical representation is found in Figure 3.12. Kaplan’s typology explicitly aims at developing a conceptualisation allowing ‘meaningful normative evaluations of terrorist acts or types of terrorism’. ²²⁴ It therefore includes a number of features considered most important for such judgements to be made. First, the typology raises the question of whether groups discriminate in line with judgements on legitimate and illegitimate targets, with legitimate targets considered those relating to the military, social infrastructure or symbolically important property. The typology then distinguishes lethal from non-lethal attacks, before moving on to consider whether the actor is a state or non-state entity. The normative relevance of the actor is argued to be an indirect measure, as it brings into focus the different ways in which state and nonstate actors are judged in terms of their rights and responsibilities in relation to the use of force. Finally, the typology distinguishes between the geographic context of the acts as to whether they are within or across international boundaries. These four sets of criteria generate 16 types of terrorism. This is argued to provide a normative framework against which the various forms of terrorism may be considered justified to a greater or a lesser degree. While this does not imply that there are justified acts of terrorism, it is claimed that it allows an assessment of the severity of the act, facilitating a more organised approach to the normative question of justification and harm. In this regard, Kaplan’s typology is unique in relation to the various typologies considered here, as no other approach considers the normative dimension so explicitly. However, whether acts of terrorism are lethal or non-lethal is often a matter not of terrorist intent but of the bad or good luck of prospective victims. Arguably, non-lethal terrorism cannot terrorise – which undermines support for this typology.

Figure 3.12   Kaplan’s typology of terrorism.

Source: S. Kaplan, ‘A typology of terrorism’. Journal of Political Philosophy, 6(1), 2008, pp. 1–38.

To narrow the focus of analysis a little more, a number of emerging threats and ‘modalities of terror’ have been identified. ²²⁵ These include eco-terrorism, narco-terrorism, agro-terrorism, biological, chemical and nuclear terrorism, cyber-terrorism, as well as suicide terrorism. Most of these concepts are subject to considerable debate over definition and characterisation. Often they emerge not from academic research but are propagated by think tanks and the security and consultancy industry. A brief overview of some of these will be given here to provide a flavour of the debate. In particular, consideration will be given to environmental terrorism and cyber-terrorism.

A distinction should be made between eco-terrorism, which has been described as the extreme element of the environmental movement, and environmental terrorism. Eco-terrorism is better described as a single-issue form of terrorism, ²²⁶ and will not be considered further here. The concept of environmental terrorism has been explained, in one conceptualisation, as the threat of environmental destruction, in peace or wartime, designed to create fear over the ecological consequences of the act. ²²⁷ This definition is identified alongside the development of a taxonomy of environmental destruction based on whether an act is deliberate or unintentional, whether it is undertaken during peacetime or in war, and whether the act has a symbolic quality or not. Symbolism is considered in two ways: primary symbolism, where environmental destruction is carried out in order to instil fear over the impact on the environment; and secondary symbolism, involving the generation of fear, but not necessarily specific to environmental consequences. Through combining these elements, eight types of environmental destruction are described, two of which are considered to constitute environmental terrorism: ‘deliberate, primary symbolism in peacetime’ and ‘deliberate, primary symbolism in wartime’. ²²⁸

This is a potentially useful framework under which to consider two other forms of political violence that are emerging in the public and academic discourse on terrorism: ‘forest jihad’ and ‘agro-terrorism’. The first refers to the potential threat of arson in the forests of the United States, Europe, Russia and Australia made on jihadist websites. ²²⁹ The stated aim of such an attack lies in the economic impact, but it would also serve to engender fear in the population. This would therefore fit into the deliberate, secondary symbolism in peacetime category of Schwartz’s ²³⁰ taxonomy, as the primary aim is not to create concern over the impact on the environment, but to impact more broadly on the states where it is carried out. So far, however, more forest fires are the work of unscrupulous real-estate developers than of rural terrorists.

Agro-terrorism may be considered to be poised between so-called bio-terrorism and environmental terrorism. It has been described as deliberately introducing a disease agent into the food chain, including directly into livestock. ²³¹ These are considered ‘soft’ targets ²³² that have the potential for infiltration at many stages of the food preparation process. Again, if one applies the taxonomy developed by Schwartz, ²³³ this would come under the same category as forest jihad, in that the primary aim is not destruction of the environment but to instil fear and inflict casualties. So far, terrorists have shown little interest in ‘agro-terrorism’, and the rise of the concept has perhaps more to do with post-9/11 scaremongering about the whole gamut of potential vulnerabilities of the US economy than with a close reading of terrorists’ writings regarding their targeting priorities. It sometimes appears as though the counter-terrorism lobby, looking for new ‘business’, appears to work unintentionally in parallel with some terrorists, exaggerating vulnerabilities to generate government programmes against such new threats – threats that the terrorist themselves had, until then, perhaps not considered making. With the media providing a platform for such lobbyists, political momentum is built up that can create additional funding for specific ‘terror fighters’.

A different kind of tactic is seen in the use of technology. Termed ‘cyber-terrorism’, the term is used (and abused) to describe a wide variety of computer- and internet-related hazards. The authors of this chapter have yet to see a single incident of cyber-terrorism where (1) civilians have deliberately been killed by cyber-attacks, and (2) the explicit purpose was to influence (impress, intimidate, coerce) third parties. There have been a few cases of cyber-murder (e.g. by changing, via computer, the strength of a medical prescription in a hospital pharmacy for a particular patient who consequently died from an overdose), but not, to our knowledge, cyber-terrorism. We find that the conceptual enterprise in this area is replete with competing approaches. ²³⁴ However, one attempt at a typology of cyber-terrorism is provided by Ballard, Hornik and McKenzie, ²³⁵ and is proposed to guide future research efforts. It incorporates four types of what are described as cyber-incidents, and is useful in addressing the various forms of behaviour considered part of the repertoire of ‘cyber-terrorism’. The first type of attack is information attacks, where electronically stored information, or computer systems, are damaged, altered or destroyed. Such attacks are considered the most common, and are the subject of most analysis. The second type is described as infrastructure attacks which aim to impact upon computer hardware or programming. Technological facilitation refers to the use of information technology to plan or incite terrorism and acts to facilitate ‘traditional’ terrorism or cyber-terrorism. Finally, the use of cyber-communications can assist fundraising and the promotion of terrorism as a mode of waging conflict. This use of the internet is well documented in this regard.

It is doubtful whether cyber-terrorism and environmental terrorism as described above constitute ‘true’ terrorism. Most incidents of cyber-crime, hacking, etc. have nothing to do with terrorism, and while these activities may create economic loss and disruption, there are, so far, few if any signs that the generation and exploitation of life-threatening fear is the intent of those engaging in cyber-attacks. The application of the ‘terrorism’ description to increasingly diverse forms of behaviour has been called ‘conceptual stretching’. ²³⁶ The result of this is, among other things, a threat inflation that makes the concept of terrorism fuzzier than it already is. This can, in the end, only serve the true terrorists, who can then argue that terrorism is just a label or, as the critical theorists argue, a ‘social construct’. The victims of real terrorism deserve better than that, and gain nothing from such conceptual stretching. ²³⁷

A review of typologies of terrorism requires a balance with respect to forms of counter-terrorism. There are a considerable number of approaches to engaging with terrorism. These have been divided into categories such as anti-terrorism, counter-terrorism, terrorism consequence management ²³⁸ and, the most neutral, responding to terrorism. ²³⁹ Other conceptualisations include conciliatory or coercive counter-terrorism, ²⁴⁰ or, more broadly, ‘discouraging’ terrorism. ²⁴¹ Anti-terrorism is sometimes conceptualised as more reactive, while counter-terrorism is seen by many as more proactive and pre-emptive. However, since the US government reaction to the 9/11 attacks, the label counter-terrorism has ‘won’ over anti-terrorism; it is now used to include a broad variety of responses to terrorism, both proactive/offensive and reactive/defensive.

The ways in which terrorism is engaged include the categories politics and governance, economic and social, psychological–communication–educational, military, judicial and legal, the police and prison system, intelligence and secret services, and others. Schmid has elaborated these in a ‘toolbox of counterterrorism measures’. Within these broad groups, more detailed measures have been outlined. ²⁴² Schmid’s toolbox has been described as useful for identifying counter-terrorism approaches, ²⁴³ broadening the response repertoire beyond narrow kinetic military approaches.

With respect to typologies of counter-terrorism, one approach based on an analysis of the counter-terrorism policies of EU members has been proposed. ²⁴⁴ It is described as a representation of the various policies in operation rather than an exhaustive classification. It is based on the four elements of EU counter-terrorism strategy: to prevent, protect, pursue, and respond. The outcome is four types of counter-terrorism policy measured against these four strategic aims. These are, first, the maximalist response, which engages with all four aims equally; and second, the human agent approach, involving a greater focus on the individuals who carry out terrorism, which concentrates more heavily on the protect and pursue aspects of policy. The confrontational approach is the most aggressive and weights policy towards measures designed to pursue terrorists. Finally, the antagonistic approach balances all but the ‘prevent’ strategies, with equal focus on protect, pursue and response policy options.

There is agreement that the form of counter-terrorism applied should be dependent on the type of terrorist group a state or society faces. ²⁴⁵ In addition, the dynamic nature of the terrorist threat necessitates a flexible approach to the management and control of non-state terrorism. ²⁴⁶ It is argued that countermeasures should incorporate, inter alia, defence and intelligence, and not exclude compromise and conciliation – at least with the constituency the terrorists claim to speak for. ²⁴⁷ The utility of a good typology is clearly seen in the counter-terrorism domain, as exemplified in Goldaber’s typology of hostage takers. ²⁴⁸ Terrorism is context-specific and, as a result, the decision as to which policy to adopt to counter it should be built on a firm conceptual understanding of the range of options available. Typology can serve as an auxiliary instrument for this, and, where rigorously constructed, can be applied to match types of terrorist group and types of counterterrorism response. This would provide a firmer conceptual underpinning to terrorist engagement, currently often thought to be lacking in some countries. ²⁴⁹

Conclusion

This review of typological work in terrorism has shown that there is great variety in both the approaches and the conceptual lenses utilised. Arguably, this is because of the range of contexts in which terrorism occurs. From those whose state support affords them impunity, to those who have too few reservations about victimising the innocent, and those ‘true fanatics’ who believe their apparently noble ends justify the use of generally ignoble means – the range of motivations and circumstances in which terrorism is used is considerable. Further complexities arise, as terrorism is sometimes used as the sole instrument to further a group’s aims, while other groups use it as one instrument among several tactics that can vary in their legality, justification and impact. Since there is no typical psychological terrorist profile and since most terrorists are normal in a clinical sense (albeit not in a moral one), typologies of terrorism that take the terrorist actor as their main criterion for classification are bound to be of limited value.

The various typologies considering the motivation and purpose of terrorist groups incorporate, inter alia, structural causes, and political and ideological motivation. They respond to the question of why groups resort to terrorism, either with respect to their own internal justifications, or through the wider influences present in the environment. The debate on ‘root causes’ and terrorist group justifications is one, and it necessitates multiple levels of analysis, implicating historical, social, political, ideology, group and individual psychology. What would be most helpful is a layered approach allowing the exploration of different levels of analysis within a unified framework of hypothesised causes and justifications for terrorism. The difficulties in this endeavour are numerous and implicate a plethora of variables, making a parsimonious aetiological typology particularly challenging. However, the explanatory weight, and the possibility for theory generation, make it an aspiration worth pursuing in future typological efforts.

The typologies outlining the conflict zone from and in which terrorists operate broadly identify similar features: that terrorists can operate both within and outside their home territory, and that they can potentially do so at a number of levels ranging from domestic, through regional, to international and transnational terrorism and, finally, global terrorism. Lately, the resurgence of piracy off the coast of Somalia has led to the use of the term ‘maritime terrorism’. To the extent that civilian sailors are taken hostage and held for ransom, there are indeed similarities, but this is basically piracy, which is robbery of ships, not terrorism. Lately, it has been mainly extortionist and criminal, with hardly any traces of political demands. There is no doubt that location matters, especially when it comes to issues of jurisdiction and for determining who should take the lead in dealing with terrorist and other militant activists who transgress both geographical and moral boundaries. The increasingly high profile of cross-border terrorism makes this particular aspect of terrorist investigation one of considerable pertinence, with these typologies providing a platform on which to position debate.

The broad historical canon of those typologies focusing on when the phenomenon of terrorism changes may be considered useful for analysts looking at how terrorism evolves, and for the knowledge that may be gleaned from comparative analyses. However, the extent to which this uncovers new relationships among those items under consideration is not as clear, although the application of empirical testing to Rapoport’s thesis ²⁵⁰ indicates that it is well suited to the development and testing of hypotheses and highlighting areas for investigation.

The various attempts to devise new typologies looking at how terrorists operate may be of some use to counter-terrorists, as they enable them to identify potential organisational weaknesses. They might also sensitise them to the evolving nature of terrorist groups. However, given the level of specificity required to verify empirically these typological conceptualisations, sufficient information may be hard to gather, reducing their practical and theoretical utility.

In comparing the work of the 1970s and 1980s with more recent typological attempts, and with typologies of political violence other than terrorism, a number of features come to the fore. Similarities in the categorisation styles employed can be seen in conceptualisations using actors, purposes, political motivation, ideology and geographical reach as principal units of typology building. The terrorist–state relationship is seen in both new and old typologies, with establishment (government) and anti-establishment (by insurgents and other non-state actors) terrorism considered important differentiating criteria in a number of typologies. Similar categories are seen regarding the ‘why’ of political motivation, with social-revolutionary, nationalist, left, right, ethnic and religious drivers seen in most new and old categorisations. Comparisons can also be found in classifications by geographical base and target, with international and domestic terrorism, and their permutations seen across typologies developed from the 1980s to the present. Despite being at least seven times less frequent than national terrorism, transnational terrorism has, however, experienced a raised profile. Financial, political and other support is being drawn from a wider geographical pool by currently operative terrorist groups, reflected in the term ‘global terrorism’ (it might, however, be more appropriate to use the term ‘cross-border’ or ‘inter-regional’, since not even Al-Qaeda operates worldwide).

A number of developments have been seen in the review of more contemporary typological attempts, with an increased focus on root causes, temporal dimensions and organisational make-up. Some conceptualisations ²⁵¹ considered structural causes of terrorism, including historical and cultural factors as well as other actors’ influence, and those of supporters and opponents. Distinguishing between different periods of terrorism has also emerged as a useful approach, with Rapoport’s ²⁵² four waves of terrorism gaining considerable credence. A similar debate is emerging over the genesis of a ‘new’ type of terrorism. Finally, a new development has been seen in the analysis of the organisational make-up of the terrorist group, with a number of conceptualisations presented including networks, dune organisations, and analogies to company structures aiming to inform analysis.

In comparison to the brief review of typologies of political violence other than terrorism, similarities in style of categorisation may be observed. Considering the purpose of conflict, similar reasons were posited to terrorist motivations, including revolution, secession, factionalism, and ideological justifications. Actor-based constructs focused at a similar level of abstraction to terrorist groups, with the relationship between combatants described along interstate, extra-state, and intra-state lines, and a variety of other violent non-state actors (VNSAs) categorised, including terrorist actors, militias, criminal groups and warlords.

Finally, the debate about ‘new wars’ and ‘old wars’ in the security studies literature has its equivalent in Terrorism Studies. Interestingly, there appears to be some convergence of conceptualisation between war and terrorism. For example, the fifth-generation warfare posited by Reed ²⁵³ incorporates terrorist actors. Further, the concept of new warfare presented by Kaldor ²⁵⁴ includes characteristics familiar to terrorism scholars, including ideological rather than territorial drivers of violence, the targeting of civilians as a tactic, and the increasingly unclear line between combatants and non-combatants.

On examination of the ways in which typologies have been developed, the predominance of deductive approaches is evident. Of those reviewed, only a minority applied an inductive, empirical route; most often, they used secondary sources as material and drew conclusions based on their interpretation. Thus, the deductive approach has a dominant presence in the typological literature, although some conceptualisations have been supported by empirical work. ²⁵⁵ It is in this way, through the development of typologies, followed by comprehensive testing and further investigation, that the typological literature is likely to fulfil its potential to successfully describe, classify and explain terrorist behaviour. The inductive route may be considered particularly useful in this, as it allows a building-block approach, where further data may be applied to typologies in order to enhance or challenge them, thereby countering the possible reification of the subject. ²⁵⁶

The aspirational outcome of an inductive approach to typology construction might be a contribution to a cross-type typology, which includes all possible types in its field of reference. ²⁵⁷ It is this that has been described as typological theory. ²⁵⁸ At least one researcher has described the achievement of such a grand typology as ‘virtually impossible’. ²⁵⁹ The plethora of factors implicated in terrorist group development, operation and aetiology make a comprehensive typology indeed a complex challenge. However, if the goals of discovering relationships, generating hypotheses, developing theory and identifying new areas for investigation ²⁶⁰ are maintained, an iterative, developmental process, crucially informed by empirical testing, can and will provide valuable insights into useful categorisations of terrorist groups and their behaviour in broader contexts of multi-party violence and conflict. However, even as typology-building on terrorism reaches greater maturity, we cannot hope that it will ever achieve the interpretative and predictive power of the periodic table that Dmitri Mendeleev establishes for the chemical elements in 1869. However, in the absence of a universally accepted definition of terrorism and with the lack of a general theory of terrorism, typology construction can nevertheless be a useful instrument to advance our understanding of terrorism – provided it is embedded in a framework that looks at the conflict behaviour of the opponents of the terrorists as well and takes into account additional contextual factors. Theoretical progress in the field of Terrorism Studies will have to be based on typological progress, which, in turn, is based on conceptual progress.