Clearing the Conceptual Air
The catastrophic events and aftermath of September 11, 2001, have been so dramatic that they have led to the understandable conclusion that the world has entered an “age of terrorism.” People have experienced terrorism—by whatever definition—for centuries if not millennia, and there is increasing consciousness of the phenomenon, especially in the past three decades. Nevertheless, since September 11 consciousness of terrorism has taken a quantum leap. Contemporary terrorism now looms as a new and menacing situation without end, calling for new approaches on many fronts, including a new kind of preparedness and a new kind of struggle against it. In this report we intend both to confirm and to qualify this impression of novelty.
Even prior to the recent, supercharged attention to terrorism and terrorists, social scientists had difficulty grasping these phenomena analytically. A summary account of terrorism, published on the eve of September 11, complained that “terrorism is a contested concept that resists precise definition,” that it is not “a central element in major theories of war and conflict,” that it is difficult to “make comparisons or draw general conclusions,” and, above all, that “there is no comprehensive unifying theory of terrorism” (Crenshaw, 2001, p. 15605). We intend to qualify this gloomy diagnosis by bringing behavioral and social science knowledge to bear on many aspects of contemporary terrorism.
The operative passages in the charge to the panel on Behavioral and Institutional Issues (called here the DBASSE panel) read as follows:
The purpose of the DBASSE panel is to provide guidance on terrorism-related behavioral, social, and institutional issues to the federal government. It will operate under the aegis of a National Academy of Sciences-wide committee, the Branscomb-Klausner
Committee on Science and Technology for Countering Terrorism. . . . The panel is charged with writing a report that will include (1) a typology of terrorism; (2) an evaluation of the current state of knowledge and capacity for dealing with the most significant threats; and (3) a research agenda.
The DBASSE panel responded to this charge and prepared this report in the context of two closely related additional activities. The first is the panel’s contribution to the work of the parent Branscomb-Klausner Committee. Two of the panel members, Neil J. Smelser and Thomas C. Schelling, were members of the parent committee. That committee’s report is entitled Making the Nation Safer: The Role of Science and Technology (National Research Council, 2002). Chapter 9 of that report, “The Response of People to Terrorism,” contains much of the material in the “responses to terrorism” section of this report, and behavioral and social science contributions appear in many other parts of that report as well. Readers will note that the research recommendations in Chapter 9 of Making the Nation Safer are not identical to the recommendations in this report. These differences reflect the different authoring bodies, whose perspectives on research priorities were not uniform.
The second line of activity is a report by the National Research Council’s Panel on Understanding Terrorists in Order to Deter Terrorism, sponsored by the Defense Advanced Research Projects Agency (DARPA). That panel had the same chair as the DBASSE panel, and eight individuals were members of both panels. The DARPA panel concentrated more narrowly on the issues of what terrorists value and what kinds of sanctions, inducements, and policies can be brought to bear in influencing those valued elements. Its report is entitled Discouraging Terrorism: Some Implications of 9/11 (National Research Council, in press). Taken together, the materials in the three reports constitute a significant compilation of available behavioral and social science knowledge relating to terrorism, brought together in the period following September 11, 2001.
As the reader will appreciate, the charge to the DBASSE panel is extremely broad and permissive (“to give guidance on terrorism-related behavioral, social, and institutional issues”) and left the panel with great discretion with respect to what issues it would pursue. Furthermore, the parent committee gave full freedom to the panel to develop the report in any directions it chose.
Before sketching how we responded to the three subparts of the charge, we should point out that up to the present, the behavioral and social sciences—like many other lines of inquiry—have not focused much research on the characteristics of terrorists, the organization of terrorist activities, or the determinants of terrorism. We do not claim to know the reasons for this relative neglect, but among them are its relative unfamilarity as a form of warfare, its analytic slipperiness (see the discussion on definitions and typology below), and the fact that most terrorist activity, designed to be carried out in secret and done so if it is successful, yields a very skimpy supply of available empirical data. In recent months, of course, a river of research on terrorism has begun to flow, and a full torrent of concerned but nonscientific writing has appeared. All these circumstances have dictated that the two panels (DBASSE and DARPA) did what was possible in consulting research directly on terrorism, but also went to different and related research topics, such as disaster studies, the literature on social movements, deterrence analysis, and some work on international economic and political relations.
The DBASSE panel dealt with the subelements of the charge in the following ways:
(1) We addressed the issues of definition and typology directly. However, instead of being able to locate an authoritative definition already “out there” empirically, we found that no such satisfactory general definition could be found in the conceptual haze surrounding the literature on terrorism. We did develop a working definition that provided a guide for us, but we also went beyond and analyzed the unsatisfactory state of definitional affairs. We also decided to identify a number of dimensions of terrorism, along which its variations can be described and which yield the tools for a provisional number of types, rather than seeking a typology from the array of empirical phenomena that have been termed terrorism.
(2) We decided to organize the “state of knowledge” about terrorism by assembling our best understandings of the diverse manifestations and determinants of the phenomenon, as well as its consequences. However, our report does not claim to cover all the available and relevant research literature; rather, it focuses on material germane to the main line of the panel’s argument. In addition, we acknowledge that the current state of
knowledge does not provide bases for formulating specific and unequivocal responses to terrorist threats. We trust that following through on the panel’s proposed research agenda will improve that knowledge base.
In fact, there are important topics related to terrorism that the panel was not able to take on. These include an examination of technological advances that change the costs and benefits of terrorism as a strategic choice; terrorism and its deterrence as a signaling game with imperfect information; the role of the media, especially in the Middle East; and terrorism, and support for it, as a diffusion process at the level of social groups.
The panel also did not consider questions of responsibility for failures of intelligence, did not address such political questions as the types of trials appropriate for apprehended terrorists, or provide policy advice.
(3) We produced 18 research priorities arising from our substantive analyses. These are found at the end of the report.
DECODING THE MYSTERY OF TERRORISM: NEW OR OLD, FAMILIAR OR UNFAMILIAR?
Especially after the attacks on the World Trade Center and the Pentagon—but also in response to some earlier attacks—a vivid but very oversimplified conventional wisdom has developed about the contemporary terrorist threat. It portrays that threat as both new and unfamiliar. The following are the relevant ingredients of that view:
Most purposive human activities are directed toward reducing uncertainty in the environment as a means of coping with that environment. The aims of terrorist activities are to create, maximize, and continuously shift the parameters of uncertainty, confusion, insecurity, and fear.
The evolution of terrorism from World War II—through the phases of colonial struggles, hostage-taking, hijacking, assassination, explosive bombing, and suicidal vehicle and airplane bombing—shows a certain mercilessness in the perpetration of violence: any target, at any time, in any place, and by any
means. The overriding premium is on quickness, surprise, shock, and generating reactions of horror and terror.
Contemporary terrorism is often stateless and not territorially bounded. It operates largely in the form of territorially fluid networks that are relatively unreachable politically and diplomatically.
The contemporary terrorist mentality and culture, which are rooted in absolutist, either-or, good-and-evil world views, resist efforts to negotiate, because accommodation, bargaining, and mutually acceptable compromise are not envisioned as possibilities within many terrorists’ mental framework. A leading Islamic terrorist proclaimed that “God does not negotiate or engage in discussion.” In a similar vein, the former leader of Hezbollah stated, “We are not fighting so that the enemy recognizes us and offers us something. We are fighting to wipe out the enemy” (quoted in Hoffman, 1998, pp. 98, 96). A corollary of terrorism based on absolute religious principles is that it is resistant to mechanisms of peaceful influence and persuasion, to say nothing of conversion, because of the strength and rigidity of these principles.
This absolute, noncontingent mentality frequently stands side by side with other characteristics that can only be described as rational and adaptive: a sharp sense of reality and reality testing, inventiveness, maneuverability, ingenuity, and instrumental capacity that has been manifested in terrorist planning and execution of attacks.
Known military solutions can neutralize specific terrorist organizations and can reduce the probability of terrorism in the short run, but they cannot suppress or destroy the cultural and motivational forces that inspire terrorism. These forces are the complex result of cultural definitions, historically generated hatreds, international power relations, contemporary economic and social conditions, and doctrinal education. In the worstcase scenario, military defeat may intensify some of the forces generating terrorism. In all events, the limitations of military conquest bring to mind the analogy that a one-time weeding of the garden can be only a short-term solution.
The same can be said of the effects of economic sanctions against terrorism, such as blocking financial assets, choking off imports and exports, and assaulting the drug trade. These may achieve the same short-term results as military action does, but
even successful economic action does not suppress and may aggravate the forces that generate terrorist activity.
In this report we hope to demonstrate that this composite view, while plausible in many respects in assessing the present international terrorist situation, is likely to be myopic with respect to every one of the characteristics listed. In fact, every “constant” in the above description turns out to be a variable.
In addition to its own internal variability, contemporary terrorism shares some characteristics with a number of other types of historically known situations. Terrorist groups themselves call forth a number of other phenomena with which limited comparisons can be made:
crime and disorder, both domestic and international (e.g., the international drug trade),
cults, sects, witchcraft and satanic groups, suicide clubs,
extremist social movements, and
Reactions to terrorist situations and attacks call to mind:
plagues and famines,
natural and accidental disasters (e.g., floods, fires, storms, earthquakes, and major industrial accidents),
episodes of intense international crisis (e.g., the Cuban missile crisis of 1962), and
prolonged periods of international tension (e.g., the cold war).
Terrorism is like every one of these, but it is identical to none of them. One can therefore learn from all of them, but new understandings and modes of coping with the historically unique situation currently facing the nation and the world are needed.
ISSUES OF DEFINITION
In the panel’s search for an adequate definition of terrorism, we came upon a paradoxical—and to some degree, paralyzing—result. There are several ingredients of this result: (a) A proper definition of terrorism is an essential ingredient in understanding the phenomenon and in crafting responses to it. (b) Actual definitions are multiple, varying greatly in inclusiveness. (c) A reasonable but necessarily imperfect working definition is, however, possible. (d) The term “terrorism” is a stigmatizing concept; as a result, definers, labelers, and the labeled are eager selectively to exclude themselves and their own actions under the term and, correspondingly, to include others and their actions under it; the result is that “terrorism” is a politically contested concept. (e) Given this frustrating set of circumstances, a reasonable strategy is to abandon the search for the one, true definition of terrorism and (f) rely on a strategy of identifying relevant dimensions of terrorism, highlighting one or more of these according to its usefulness in understanding a given problem and in the interest of maximizing flexibility of responses to an obviously multifaceted phenomenon. We comment briefly on each of these ingredients.
THE NEED FOR DEFINITION
A working definition of terrorism is necessary because the government must respond to it in unambiguous and legal ways, and citizens and others—including the international community—must recognize those responses as legitimate and not capricious. A scientific and analytical approach may guide the government toward the adoption of definitions and policies that are operationally most useful.
There are three major dangers in erroneous definitions of terrorism. First, defining it too broadly will dilute and waste resources that could be put to more efficient preventive and defensive use. Second, a too broad definition may catch in its net persons not at all engaged in undesirable activity and may violate their constitutional rights or international conventions. Third, defining terrorism too narrowly means that people will
not be protected from unanticipated kinds of attacks. In principle, terrorism must be defined both sufficiently and efficiently.
ATTEMPTS AT DEFINITIONS
We attempt no comprehensive survey of definitions (Johnson, 2001; Ruby, 2002) but give a few to indicate their variations in scope and their problems. One writer commented, “A universally recognized definition will be elusive. One survey of leading academics revealed 109 different definitions of what constitutes ‘terrorism’” (Takeyh, 2002:70).
A simple and straightforward definition is that terrorism consists of acts designed to induce terror. True to the name of the phenomenon, it nevertheless includes a world of behaviors in which we are not practically interested (e.g., stalking and sadistic “mind games”).
Another brief definition is found in the Merriam-Webster Collegiate Dictionary: “the systematic use of terror, especially as a means of coercion.” This includes a broader intent, contained in the word “coercion,” but it also leans in the direction of vagueness and overinclusiveness.
In 1988 the State Department issued a definition of terrorism: “premeditated, politically motivated violence perpetrated against noncombatant targets by subnational groups or clandestine state agents, usually intended to influence an audience” (U.S. Department of State, 1988). This is a helpful definition, but the terms “politically motivated,” “noncombatant” and “subnational” seem unduly constrictive, given the known diversity of motives, targets, and organizational modes of terrorism.
The American Heritage Dictionary contains a more elaborated definition: “unlawful use or threatened use of force or violence by a person or an organized group against people or property with the intention of intimidating or coercing societies or governments, often for ideological or political reasons.” Is not “unlawful” perhaps too constricting and “threatened use of force” too inclusive?
In these definitions there are recurring definitional characteristics: illegal use or threatened use of force or violence; an
intent to coerce societies or governments by inducing fear in their populations; typically with ideological and political motives and justifications; an “extrasocietal” element, either “outside” society in the case of domestic terrorism or “foreign” in the case of international terrorism. These ingredients provide a useful composite, which the panel has used as a practical guide. However, neither singly nor in combination do these ingredients provide an unambiguous and uncontestable definition. We now proceed to say why this is so.
For nearly half a century, philosophers and linguists have written about what is called an “essentially contested concept” (Gallie, 1956). The main idea is that some concepts are inherently incomplete, without being totally incoherent, and are filled out differently by individuals and groups who bring different backgrounds, beliefs, and political convictions to bear on them. Moreover, the meanings of such words change in emphasis over time. Words that have been named as falling into this category are “work of art,” “democracy,” “rape,” “poverty,” and “underclass.”
Terrorism seems to be such a word. It is a concept with great rhetorical power but limited scientific precision. What is “terrorism” to some may be called “freedom fighting” by others. Ambiguity as to what is to be included under the concept at its borders poses another conceptual difficulty. As an exercise to demonstrate these features of contestation, we ask the reader to scan the following list and try to determine which of them are terrorist acts and which are not (or, perhaps better, who would call each terrorism and who would deny it):
British and American firebombing of Dresden in World War II
Dropping atomic bombs on Hiroshima and Nagasaki in World War II
Sherman’s march through Georgia during the American Civil War
Palestinian suicide bombing
Israeli punitive strikes on Palestinians
Lebanese Phalangist militia attacks on Muslims
The Bay of Pigs
Haganah and Irgun attacks on the British in Palestine
The original Sicilian Mafia as an organization to protect the peasants
Organized crime, especially protection rackets
Lynchings, church bombings, and Ku Klux Klan activities generally
International drug trade and the financing of guerillas
The “reign of terror” in postrevolutionary France
Stalin’s purges of the 1930s
Frank and Jessie James
Not only are many of these acts ambiguous but the arrays of phenomena classified as terrorism differ according to the definition chosen. This exercise alone should demonstrate the havoc that essentially contested concepts wreak on attempts to be scientifically precise in defining the term.
TERRORISM AS A SET OF DIMENSIONS
One productive way to come to terms with these definitional conundrums is to forsake the effort to conceive of terrorism as a thing to be defined and, instead, to seek out a number of dimensions along which terrorist and quasi-terrorist activities fall. This exercise may or may not yield a definite typology, but in any event it provides a basis for avoiding historical and comparative myopia and considering the variability and complexity of terrorism and its fluidity of organization over time. Having understood these features, we may be in a position to make more intelligent and focused decisions about how to contend with it.
One of the more elaborate attempts to lay out dimensions is that of Schmid et al. (1988), who identify 10 common bases for classifying terrorism: (1) actor based, (2) victim based, (3) case based, (4) environment based, (5) means based, (6) political orientation based, (7) motivation based, (8) demand based, (9) purpose based, and (10) target based. Still other bases could be imagined. As a further exercise in thinking dimensionally, we produced a dimensional analysis, overlapping with that of Schmid et al., based on the questions: Who are the actors? What
are their actions? And what are the consequences of these actions? Each of the categories contained in these questions is broken down further, yielding the ingredients of a complex typology as well as a number of derived types. The results of this exercise, as well as some of its implications, are found in Appendix A.
The panel sees several potential advantages to this kind of dimensional thinking: it offers a more open-ended approach to the range and complexity of terrorist behavior than a single, fixed definition; it provides an avenue for disentangling the problems of meaning we have identified; and it provides the basis for developing a systematic comparative analysis of different manifestations of terrorism.