(c) 2003 Stephen E. Sachs <contact me>

The Changing Definition of Security

by Stephen E. Sachs

International Relations
Merton College, Oxford
Week 5, Michaelmas Term 2003


What are the principal difficulties involved in attempts to define 'security'?  How has the concept been extended beyond traditional concern with the military security dilemmas facing states?


A traditional definition of the state, often attributed to Max Weber, required as a necessary condition the effective monopoly on the use or licensing of violence within a given territory.  The security of states was therefore threatened by any change that might threaten that monopoly of violence--whether through external invasion or internal rebellion.  In the Westphalian world of (internally) strong states, there is less danger of internal conflict, and the international system is marked by conflicts among states rather than within them.  Since 1945, however, many of the most significant threats to state security have been internal, rather than external, a shift which has only accelerated and which may have profound consequences for the conduct of international relations.

As the predominant concerns of security strategists have changed, however, there has also been a more fundamental rethinking of the very framework of state security.  If many of the newly created states of the formerly colonized world are still quite weak, perhaps the security of the state apparatus--which may, after all, be the oppressive tool of an elite--ought not to be as significant a concern.  A new concept, at times given the name of "human security," has been suggested to express the need of individuals for safety in other arenas of basic need--access to clean food and water, environmental and energy security, freedom from economic exploitation, protection from arbitrary violence by the police, gangs, or domestic partners, etc.  However, while this concept may be useful in indicating the variety of human needs that must be satisfied, it is far too expansive to be an effective policy goal, and does not offer an appealing alternative to traditional conceptions of security.

The need for a new understanding of security is revealed by the changing nature of war over the last 250 years.  In the 18th and 19th centuries, wars were generally short, lasting only two years or so between the declaration of war and the signing of the peace treaty.  Since the experience of the two World Wars, however, the nature of conflict has changed. Cross-border war has become a primarily "small- or medium-power activity," and thus the attention of great powers has been focused on other types of conflicts.[1]  Wars are often conducted 'unofficially,' without formal declarations of their beginning or end, and such conflicts may drag on for decades (as in the case of Ulster).[2]  Kalevi Holsti notes that security between states in many areas--the Third World, the former Soviet Union, etc.--"has become increasingly dependent on security within those states."[3]  In the Third World, the security threats to the state apparatus are far more frequently internal than external, especially given that many decolonized nations were formed containing substantial linguistic, cultural, or ethnic minorities with few ties to the state.

Many of the intrastate wars we have witnessed therefore concern questions of national liberation, unification, or secession--questions "of statehood and the nature of community within states."[4]  These "people's wars" often make no distinctions between soldiers and civilians, and thus result in extraordinarily high civilian death tolls.  Moreover, because they are not conducted by states which have limited goals and a strong interest in self-perpetuation as an organized group, the "ordinary cost-benefit analyses that underlie wars as a 'continuation of politics by other means' no longer apply."[5]  In some areas, the breakdown in order has been so severe as to create conflicts reminiscent of the Thirty Years' War, when "warfare seemed to escape from political control; to cease indeed to be 'war' in the sense of politically-motivated use of force by generally recognized authorities, and to degenerate instead into universal, anarchic, and self-perpetuating violence."[6]

The shift in the nature of conflict has also forced states to consider new ways of protecting their monopoly on violence.  One prominent example is the effort to prevent the proliferation of conventional and unconventional weapons.  States have few resources to defend against the catastrophic delivery of weapons of mass destruction, such as via small airplanes, ballistic missiles, or advanced "reconnaissance strikes."[7]  The growing reach and sophistication of international terrorism poses a further threat of intra-state violence that cannot easily be countered by traditional military organizations.  Thus, some are turning to different approaches, such as "cooperative security," in order to achieve traditional security goals.

In the past, security strategy has often been focused on external threats, and more specifically external military threats (which therefore require a military response).[8]  Yet the nature of future conflicts may require that those concerned with preserving the state's monopoly on force look beyond such traditional categories as "material capabilities and the use and control of military force by states."[9]  Instead, planners must address problems such as "environmental pollution, depletion of the ozone layer, [global] warming, and massive migrations of unwanted refugees."[10]  These issues may only infrequently become the direct cause of conflict (as in the case of wars over scarce resources, such as water), but could easily produce conflicts through the mechanisms of economic decline and political instability.  As Jessica Mathews writes, in conflicts spurred by environmental degradation, "[t]he underlying cause of turmoil is often ignored; instead governments address the poverty and instability that are its results."[11]

While some have objected to including such threats under the "security" framework--"Defense of the nation against infectious disease," argues Lawrence Freedman, "is an altogether different problem than defense against ballistic missile attack"--those concerned with preventing and ending conflicts will have a responsibility to address these dangers at the same time as more traditional military threats.[12]  Questions of "security" are often given pride of place before other potential policy concerns (as in the case of Thomas Hobbes, who thought that none of the other goods in life could be achieved without it).  Thus, if overpopulation or economic discrepancies will be a source of future conflicts, they are worthy of the security planner's attention.  Should political will and energy be "focused predominantly on military solutions to the problems of national security," Richard Ullman writes, "the nonmilitary tasks are likely to grow ever more difficult to accomplish and dangerous to neglect."[13]

However, it is important to note that such a focus is not as radical as it might seem.  A new goal such as preventing the proliferation of biological weapons is still ultimately designed to prevent the non-state-sanctioned use of violence within a given territory.  Even if the intermediate goals or the methods employed to reach them are inherently international, the policy is ultimately rooted in a concern for national security.  Although such security concerns may employ non-traditional means, they do so in order to achieve traditional ends.

Some scholars, however, have argued that the rethinking must go even further, including the ends of security policy as well.  The effort to broaden security planning to include "human security" changes the terms of debate.  It goes beyond arguing that non-traditional problems such as environmental degradation are likely to create a security threat (by encouraging conflict) to claiming that such degradation itself constitutes a security threat--a threat to the quality of life of those in a polluted environment.

The theoretical difficulty with limiting the concept of security to the use of physical violence is that all economic and political relations are characterized by force, whether threatened or actually employed.  The possession of economic rights in a resource is constituted by a threat of legal force against those who would attempt to violate those rights.  Someone who has no economic rights to food, the argument goes, or no rights to other resources which can be traded for food, is prevented by force from obtaining food just as surely as someone who is deprived at gunpoint.

Thus, J. Ann Tickner quotes approvingly another author's definition of security "not only in terms of the internal security of the state, but also in terms of secure systems of food, health, money and trade."[14]  Tickner places a focus on "structural violence," which goes beyond physical violence to include "the indirect violence done to individuals when unjust economic and political structures reduce their life expectancy through lack of access to basic material needs."[15]  A secure society must therefore "promote a viable ecosystem while at the same time working towards the elimination of both physical and structural violence," an elimination that requires "dismantling hierarchical boundaries between women and men, rich and poor, and insiders and outsiders which have contributed to an exclusionary divisive definition of security."[16]

However, such a conception of "structural violence" sits uneasily with traditional concepts of force and violence.  First, it is unclear why such limitations are only "violent" if they are unjust.  There are many just uses of violence, as in the case of threatening force (even only the level of force necessary for arrest, trial, and imprisonment) against those who would violate individual rights--and these acts do not become less forceful because they happen to be just.  Second, such a definition raises the possibility of treating all unjust economic arrangements per se as cases of structural violence.  Given that sophisticated health care is expensive, any system that unjustly reduces the resources available to one group will result in a decline in the group's life expectancy from what it would otherwise have been.  Thus, if all unjust social arrangements are inherently violent, no unjust society is secure--and it is impossible to give a descriptive account of security without first establishing normative agreement on what constitutes a just economic and political system.  Indeed, Tickner goes on to state that "true security cannot be achieved until . . . hierarchical social relations and divisive boundary distinctions are recognized and substantially altered and until all individuals participate in providing for their own security."[17]

Alternative definitions of human security run into similar difficulties.  Ullman, for instance, defines a threat to national security as something that either "(1) threatens drastically . . . to degrade the quality of life for the inhabitants of a state, or (2) threatens significantly to narrow the range of policy choices available"--categories in which he explicitly includes the threat of earthquakes.[18]  Yet there is no reason why a potential "national emergency" must always be at the same time "a threat to national security"; threats such as floods or earthquakes, which are not deliberately inflicted by an external or internal agent, could be usefully described as the former but not the latter.  A region that is peaceful but prone to hurricanes does not face a security threat.  As Mohammed Ayoob notes, the "all-inclusive definition of security . . . runs the risk of making the concept so elastic as to detract seriously from its utility as an analytical tool."[19]

Indeed, there is a significant danger in defining security as including everything that's good in life--or everything that's considered 'necessary.'  If it were so defined, it would be impossible for there to be tradeoffs between security and other values, and policies could only represent choices for one type of security as opposed to another.  The word itself thus loses its effectiveness at delineating a particular realm of political priorities.  The most oppressive and exploitative dictatorship may, if sufficiently entrenched, appear stable and secure from the outside.  Although we might urgently desire a change in the form of government and greater respect for human rights, it is an empirical proposition, not a certainty, that such changes would reduce the risk of external adventures by the regime or internal non-state-sanctioned violence.

It cannot be merely assumed that all goods stem ultimately from the same source; some may very well be in eternal tension with one another, and there may be no easy resolution of the tradeoff between freedom and order.  A choice to redirect the U.S. defense budget towards environmental protection might, to some, produce a better world, but it would never produce a more secure world; as Isaiah Berlin wrote of liberty, we cannot pretend that a sacrifice is an increase in the thing being sacrificed.  Defining "security" in a narrow way does not mean that all other goals--such as freedom, welfare, or equality--must be subsidiary to it; rather, it allows us to compare these goals without asserting a priori that they can all be met at once.

The effort to redefine the ends of security planning, therefore, appears to be misguided.  There are many values that policymakers might pursue, but security is only one of them, and cannot encompass the whole.  While the traditional understanding of security must undergo substantial revision, these revisions are limited to the means by which security is achieved, which should be expected to change in an era when the nature of conflict is in flux.  Yet there is no reason to generalize the ends of security beyond the protection of the local monopoly of violence, and thereby to jettison the useful work that the concept still can perform.

[1] Kalevi J. Holsti, The State, War, and the State of War (1996), p. 25.

[2] Holsti, p. 20.

[3] Holsti, p. 15 (emphasis in original).

[4] Holsti, pp. 26-27.

[5] Holsti, pp. 37-38.

[6] Holsti, p. 28.

[7] See generally Janne E. Nolan et al., "The Imperatives for Cooperation," Global Engagement:  Cooperation and Security in the 21st Century (1994).

[8] Mohammed Ayoob, "The Security Problematic of the Third World," World Politics 43:2 (Jan. 1991), p. 261.

[9] Peter J. Katzenstein, "Introduction," The Culture of National Security (Katzenstein ed., 1996), p. 9.

[10] Holsti, p. 15.  Although these threats are undoubtedly significant, Holsti wonders whether at least part of this shift was due to professional, rather than political, factors: "the prospects of conventional war between industrial countries have declined precipitously.  If new threats are not identified, strategic planners may become redundant" (p. 15).

[11] Jessica Tuchman Mathews, "Redefining Security," Foreign Affairs 68:2 (1989), p. 166.

[12] Lawrence Freedman, "International Security:  Changing Targets," Foreign Policy (Spring 1998), p. 53.

[13] Richard Ullman, "Redefining Security," International Security 8:1 (1983), p. 153.

[14] J. Ann Tickner, "Re-visioning Security," International Relations Theory Today (Ken Booth and Steve Smith, eds., 1994), p. 180.

[15] Tickner, p. 187.

[16] Tickner, p. 194.

[17] Tickner, p. 193.

[18] Ullman, p. 133.

[19] Ayoob, p. 259.