Title Menu - The International Journal of Peace Studies.  
Return to Table of Contents Return to I J P S Home


Bjørn Møller

Just as "war is too important to be left to the generals" (Clemenceau, quoted from Wintle, 1989: 97), the preparations for war, i.e. the structure and deployment of the armed forces, are too important to leave to "professionals". Indeed, peace researchers and students of international relations should devote some attention to these matters, if only because they affect international relations, for instance the propensity for arms races and wars.

The following article is an introduction to the concept of "non offensive defence", i..e. form of defence that rules out the option of armed attack against other states. The concept originated in peace research circles and has been debated in Europe for more than a decade. It is now attracting attention in other parts of the world such as the Middle East, East Asia and Southern Africa.

1. ‘Classical’ Common Security

The concept of NOD, i.e. non-offensive defence, is intimately linked with that of common security. As such, it is part of the quest for a more appropriate conception of security that has characterized international relations theory and strategic thinking for several years (Møller, 1994a).

In principle there are three different modes of expansion (as well as any combination thereof) and different degrees of radicality.

A. ‘Procedural expansion’, which does not affect what security is all about, only points to other paths leading in the same direction. Realizing that the level of individual state is inadequate, it implies an expanded scope for the quest for security, say that of dyads of states, regional (sub)systems, or even the international system as a whole.
B. ‘Focal expansion’, which implies taking into consideration the security of other entities than states, for instance nations and other societal groupings (Wæver et al., 1993), or even individuals.
C. ‘Sectoral expansion’, which implies including other types of threats besides the military ones, i.e. new ‘sectors’ (or ‘dimensions’) of security: economic, environmental, etc. (Buzan, 1991).

Mode of Expansion
Degree of expansion
Individual state
The State
Incremental expansion
Dyads of states
(Sub)Regions or groups of states
Societal groups
Environmental, societal, cultural
Radical expansion
International system
Humanity Ecosystem

Important though they are, I shall disregard focal and sectoral expansion in the following, since CS and NOD are primarily forms of prodedural expansion.
When it was first promulgated by the Palme Commission in its 1982 report (Independent Commission, 1982), Common Security was conceived as simply a novel means to the traditional end of national security. It was presumably preferable to a unilateral quest for security which would almost inevitably be at the expence of one's adversaries. Common Security was thus conceived as an escape route from the ‘security dilemma’, implying that states could only make themselves more secure by making their adversaries less so (Herz, 1950; Jervis, 1976: 58-93; 1978; Buzan, 1991: 294-327). By thus placing the security of their adversaries in jeopardy, states tended to provoke a response that detracted from their own security so that, at the end of each interactive cycle, both sides found themselves less secure. This security dilemma manifested itself in, at least, three different dynamics:

• An action-reaction pattern in arms acquisitions, where one state armed out of fear of its opponent's armaments, yet by so doing made the latter reciprocate, resulting in an arms race (Rathjens, 1973; cf. Hammond, 1993; Gleditsch & Njølstad, eds., 1990).
• A risk of preventive war (one state attacking another in the belief that its opponent would otherwise do so at a later stage when force ratios were less favourable), or of pre-emptive attack resulting from crisis instability (Schelling, 1960: 207-229; Levy, 1987; Vasquez, 1993).
• Competitive alliance-building, where one world power sought allies (in a specific region or worldwide) in order to prevent its adversary from gaining a foothold, yet by so doing tended to push remaining states into the opposing camp (Snyder, 1984).

Even though there were competing interpretations of the actual implications of Common Security, as a minimum it entailed taking the security of one's respective opponent into consideration. Security was only obtainable for either side of an adversarial dyad if both sides enjoyed it simultaneously. In the words of the Palme Commission (Independent Commission, 1982: 138 and 9):

In the absence of a world authority with the right and power to police international relations, states have to protect themselves. Unless they show mutual restraint and proper appreciation of the realities of the nuclear age, however, the pursuit of security can cause intensified competition and more tense political relations and, at the end of the day, a reduction in security for all concerned. ... Restraint should be the watchword of all states; restraint, out of respect for the right of others to security, but also in selfish recognition that security can be attained only by common action.

Self-proclaimed ‘Realists’ (including the present author, see Møller, 1992) have seen this as little more than ‘cooperation among adversaries’ (Axelrod, 1984; Stein, 1990; Stein & Keohane, 1993). This presupposes neither any altruistic self-denial on the part of states nor elaborate institution-building, but is just plain common sense: States pursuing their national interests in a rational manner had better think at least two steps ahead, i.e. take into consideration how their respective adversaries might respond to whatever step they were contemplating. ‘Idealists’, on the other hand, tended to interpret Common Security as an admonition to sacrifice national interests on the altar of the common good, or even to (partially) relinquish sovereignty in favour of global governance.

2. Post-Cold War Common Security

Granting that (the ‘Realist’ version of) Common Security was sensible under the spectre of the nuclearized East-West conflict (where the nuclear standoff meant that neither absolute victory nor absolute security were any longer attainable), one might surmise that it is obsolete in the post-Cold War era. Also, it might be asked whether is was ever applicable beyond the ‘central conflict area’, for instance to what the nuclear powers viewed as ‘peripheral’ parts of the world.

On the one hand, the premise may no longer hold true that absolute victory is unattainable. Indeed, it may never have been true for all countries, but only for those closely coupled to either of the two nuclear superpowers, i.e. regarded by them as ‘vital’. In any case, it is no longer obvious why, for instance, China should be unable to decisively defeat Vietnam or Russia Estonia, i.e. why the security of great powers should necessarily be unobtainable at the expense of their smaller adversaries. This does not mean that unilateral security strategies are better than common security ones, only that they might be feasible.

On the other hand, the political climate seems much more propitious for Common Security thinking (and practice) than during the Cold War, and today roughly synonymous terms such as ‘cooperative security’ have become common parlance (Nolan, ed., 1994). Also, even though Common Security never logically presupposed institution-building or any relinquishment of sovereignty, it would, of course, benefit greatly from such a transcendence of the state-centred (‘Westphalian’) system. This has also become more imaginable than it used to be, at least judging by the number of scholarly books and semi-official works on the subject (e.g. Commission on Global Governance, 1995; Lyons & Mastanduno (eds.), 1995).

3. Common Security and Nuclear Deterrence

As far as the narrow conception of Common Security was concerned, its military implications were rather obvious: What mattered were to make both sides of the adversarial dyad secure, thereby circumventing the security dilemma. States should thus be able to defend themselves, yet without thereby posing threats to other states' national defence, which is what NOD is all about (Møller, 1992).

In the nuclear realm, this translated in a rather straightforward manner into an advocacy of stable mutual deterrence, albeit with a distinct preference for as low levels of nuclear weapons and as few nuclear powers as possible, i.e. a dual form of minimum deterrence. One might even argue that this MAD (mutual assured destruction) situation was what gave rise to Common Security thinking in the first place: It obviously made no sense to strive for a nuclear superiority that was both unattainable (because of the measures-countermeasures game) and meaningless since a ‘more than total’ destructive capability was a contradictio in adjecto (There can be no such thing as ‘more-than-overkill’). What did make sense was to reduce the risks of inadvertent or accidental nuclear war, which required steps to improve crisis stability and, paradoxically, a deliberate abstention from threatening each other's second strike capabilities (Classical formulations of the theory of stable deterrence include Schelling, 1960; Jervis, 1989). One nuclear power's security thus presupposed that of its adversary.

Even though some authors have alleged so, Common Security does not presuppose an abolition of deterrence as such, which is indeed impossible. The nuclear ‘genie’ cannot possibly be put back into its ‘bottle’ by a ‘disinvention’ of nuclear weapons. Furthermore, it would be unfortunate to return (quia absurdum) to the age of prenuclear warfare where absolute victory was a meaningful goal. In this sense, Kenneth Waltz and his fellow ‘Realists’ are right in their pro-nuclear stance (Sagan & Waltz, 1995). Where they seem to err is, first of all, in assuming that nuclear deterrence requires actually deployed nuclear weapons, whereas the potential (‘blueprint’) may suffice for the benign ‘existential deterrence’ effect (Booth & Wheeler, 1992; Rotblat et al., 1993). It remains disputed whether this potential needs to be demonstrated by a nuclear test (as India's in 1974) or materialized in a stockpiling of nuclear components. The present author thinks not, but other advocates of NOD favour such a ‘recessed deterrence’ (Singh, 1993).

Secondly, ‘Realists’ err in assuming that ‘more are better’, hence advocating a controlled and orderly proliferation. This is, indeed, a highly unlikely scenario. Either proliferation is stemmed, or it will probably occur in a ‘bursting dam mode’. Fortunately, the former seems the more likely in view of recent events: the shelving of nuclear programmes by Argentina and Brazil and (hopefully) North Korea, the appearance of the new species of post-nuclear states (South Africa, Belarus, Ukraine and Kazakhstan), and the indefinite extension of the NPT (Non-Proliferation Treaty) at the 1995 review conference (Reiss & Litwak, eds., 1994). However, additional safeguards against proliferation would be most welcome. The creation of nuclear-weapons-free zones (NWFZ) have therefore been advocated by most Common Security and NOD advocates, and they are now becoming more politically feasible than they were in the past (Redick, 1993). Indeed, declared NWFZ today include Latin America, the South Pacific, Southeast Asia, and Africa, and proposals for a Middle Eastern NWFZ seem to be taken seriously.

For the antiproliferation endeavour to be really succesful, however, it probably needs to address the demand side of proliferation, i.e. modify the conventional force postures (as well as, of course, the political context) so as to make states feel secure without nuclear weapons. This is where (conventional) NOD comes in, one of the main rationales for which has always been to make nuclear weapons superfluous, thereby paving the way for their abolition.

4. Conventional NOD

Horst Afheldt (the author who more than anybody else deserves the title of inventor of NOD) based his advocacy of what he called ‘defensive defence’ on a rather simple extrapolation from the above theory of nuclear deterrence. With a view to war prevention, the conventional armed forces should neither invite nor allow for pre-emptive strikes. As a consequence they should, first of all, pose no threat to the adversary that would provide him with a motive for pre-emption, i.e. they should be strictly and unmistakably defensive. Secondly, they should be so configured and deployed as to present a minimum of lucrative targets for pre-emptive (and especially nuclear) strikes by being dispersed in very small units, evenly distributed across the entire country (Afheldt, 1976; 1983). The former principle, which might be called the non-provocation principle, was the more important of the two, and formed the point of departure for the entire school of NOD proponents (including the present author).

NOD is simply the military implication of a (minimalist) Common Security policy: a defence that takes the security of the respective adversary into account by posing no threat to him, but which is still a military defence under national command and serving national interest (‘defined in terms of security’) by providing an adequate safeguard against aggression.

Before proceeding any further it does, however, seem advisable to clear away some conceptual confusion often encountered. First of all, there is a confusion between concept, terms and design principles.

• The concept could be defined in both absolute and relative terms: In absolute terms, i.e. as a goal, NOD signifies a military posture and strategy possessing sufficient defensive whilst completely lacking offensive capabilities. In relative, i.e. process terms, it could be defined as a simultaneous strengthening of defensive and weakening of offensive capabilities.
• There are several terms for this concept, among which ‘NOD’ is simply the best (in the present author's opinion who, as its inventor, may however, be biased). Other often-encountered terms include ‘defensive defence’, ‘non-provocative defence’, ‘non-aggressive defence’, confidence-building defence, and structural inability to attack’. They are all synonymous, but highlight different aspects of the NOD concept.
• There are numerous alternative design principles, i.e. ‘models’ such as territorial defence, barrier-type defence, disengagement, etc. (vide infra). Even though most ‘model designers’ have had their favourite terms, this did not mean that these terms referred to particular models. They all referred to the same concept, on the best realization of which opinions have merely differed.

Secondly, there is some confusion stemming from a (deliberate or accidental) misrepresentation of NOD, mostly by critics. The most common mistake has been to locate the entire NOD discourse at the wrong level of analysis. Whatever critics and others may allege, NOD is not (repeat: not) about offensive versus defensive weapons, for the simple reason that both are misnomers. There is no such thing as a defensive weapon, i.e. a weapon that can only be used as a means to defensive ends. Under the right circumstances, even the most defensively looking weapon can be used for attacking, just as an aggressor, besides offensively looking weapons, also needs such as may look quite defensive: anti-tank and air defence weapons, to mention but two examples. Nor is NOD tantamount to a passive defence, i.e. one that is tactically defensive by allowing the adversary the first blot in each encounter—which would indeed be a recipe for defeat.

The only meaningful distinctions between offensive and defence are to be found on higher levels of analysis:

• Those of operational art, strategy and grand strategy: It is largely a matter of the scale of (counter-)offensives. An offensive strategy envisages ‘taking the war to the enemy’, whereas a defensive strategy such as NOD does not, but limits its ambitions to repulsing an attack and restoring the status quo ante bellum (Kokoshin & Larionov, 1989).
• Those of military formations and national (or even alliance) military postures. They are offensive if they allow for implementing an offensive strategy, but defensive if they do not. This depends, of course, on the weapons profile, but logistics, personnel structure, etc. also play a role.

Furthermore, offensive and defensive should be understood as relational and relative terms. A state may possess offensive capabilities vis-á-vis one adversarial state, but not another (Vietnam e.g. posseses such capabilities vis-á-vis Cambodia, but not vis-á-vis China). Moreover, NOD is context-dependent in the sense that what constitutes an offensive capability in one context may not do so in another. Heavy armoured forces were thus almost the essence of offensiveness in Europe, because large-scale offensive could not be undertaken without them. Elsewhere, however, it may well be possible to invade neighbouring states with much lighter forces—and the heavy forces may be effectively immobilized by a lack of infrastructure (Møller, 1992: 147-152). Finally, NOD is a matter of degree, rather that an either/or. It can thus be seen as a process of reducing offensive capabilities and upgrading defensive strength—or of reversing a trend in the opposite direction.

A skillful application of the NOD guidelines might, in principle at least, make a dyad of adversarial states ultra-stable, both in terms of crisis and arms race stability, almost to the point of escaping the security dilemma completely. The clue to stability is not, as usually assumed, balance (usually understood as parity) but a reciprocal imbalance: Each side should be superior to the respective other, albeit only in the sense of being superior on the defensive to the other on the offensive: what has aptly been termed ‘mutual defensive superiority’ (Møller, 1992: 84-89).

DA > OB & DB > OAD:
D: defensive strength; O: offensive strength
A and B: Two adversarial states

5. Defensive Strength

Even granting that meaningful distinctions can be made between offensive and defensive strategies and postures, the question remains whether the two can actually be disentangled without detrimental effects on defence efficiency. In other words: Does the relinquishment of offensive capabilities for the sake of crisis and arms race stability inevitably come at the expense of defensive strength? If so, many states might be well-advised not to adopt NOD as their guideline, above all such as face a serious risk of premeditated attack.

Fortunately, as a general rule (with allowance for possible exceptions), it is possible to strengthen one's defences while building down offensive capabilities, simply because the defensive form of combat is inherently the stronger, as it was claimed by Clausewitz in his 1832 On War (1984: 358). However, it is only inherently so, and to make it actually stronger requires skills and specialization, which is what NOD is all about:

1. Against a breakthrough operation, a prepared defender, as a rule of thumb, requires only one third of an attacker's strength in order to prevail: the well-known ‘three-to-one rule’. There is little disagreement about the validity of this assessment (which may even be too conservative), yet considerable controversy about its applicability and relevance. Suffice it to say that it only applies to individual engagements, but neither to entire wars or campaigns, nor to such battles where the surprise factor enters into the picture (Mearsheimer, 1989).
2. Several material ‘force multipliers’ are available to defenders, constituting inalienable ‘home ground advantages’: Interior lines of communications and supply; the option of prepositioning munitions, POL (petrol, oil and lubricants) and other supplies in scattered depots; of building various types of fortifications and of constructing barriers, perhaps even of a certain landscaping.
3. A defender can disperse his forces to a much greater extent than the attacker, thereby becoming less vulnerable to concentrated attacks: the ‘no-target principle’ that has been central to most NOD proposals. This ‘dispersal principle’ only seemingly violates the almost sacrosanct ‘concentration principle captured in the Lanchester Square Law (for a critique see Neild, 1986)—since it does not rule out fire power concentration. Let us remember Liddell Hart's brilliant explanation of the dialectics of concentration:

‘The principles of war, not merely one principle, can be condensed into a single word: ‘‘concentration’’. But for truth this needs to be amplified as the concentration of strength against weakness’’. And for any real value it needs to be explained that the concentration of strength against weakness depends on the dispersion of your opponent's strength, which in turn is produced by a distribution of your own that gives the appearance, and the partial effect of dispersion. Your dispersion, his dispersion, your concentration-such is the sequence, and each is a sequel. True concentration is the fruit of calculated dispersion’ (Liddell Hart, 1974: 334).

4. The defender's command structures may be decentralized to a considerable extent (viz. Auftragstechnik), hence made more robust than the very hierarchical ones that an aggressor would tend to rely on (Grin, 1990)—especially if structured like a web, as envisioned by most NOD proponents.
5. The immaterial (‘moral’) advantages are considerable: The defenders will enjoy the support (morally, materially and otherwise) of the population.
6. The defenders will be able to exercise under more realistic conditions than a prospective attacker, namely on the very terrain where they will have to fight.
7. There are capabilities that defensive forces need much less than forces supposed to wage both offensive and defensive combat, or not at all. Examples are: logistics for long-range (i.e. ‘strategic’) mobility, i.e. air and sealift equipment; the ability to move about under enemy fire (requiring armour, mobile air defence, etc.); and long-range striking power (including command, control, communications and intelligence, i.e. C3I, systems). Not acquiring such capabilities (which happen to be very costly) allows for a considerable enhancement of defensive strength.
8. Even though there are no technological panaceas (a ‘defensive superweapon’, for instance), certain trends in the development of modern weapons technologies tend to benefit the defending side disproportionately: The revolutionary development in micro-electronics and integrated circuits, for instance, allows for miniaturization which may render major weapons platforms superfluous for defensive purposes, whereas they remain indispensable for offensive operations. Even though it would certainly be premature to write off the tanks, combat aircraft, or major surface combatants as obsolete, they may nevertheless be facing obsolescence (in the sense of declining cost-effectiveness) in the coming decades (Canby, 1984).

6. NOD Models

How effective NOD will be depends, of course, on which particular model (or mixture of models) would be selected for implementation. Just as they are not equally defensive, NOD models are also not equally effective, and their suitability may be context-dependent (On the panoply of models, see Møller, 1991; 1995a).
Most of the following models have either been designed for a particular context (West Germany during the Cold War), or remain very abstract. Were one to simply transpose them to quite a different setting (as has occasionally been done by NOD critics) one is bound to arrive at absurd results. Furthermore, as all abstract defence models, NOD models are not to be confused with actual defence plans, for which they are merely conceived as abstract politico-military guidelines. Whereas such modelling is a perfectly legitimate task for ‘armchair strategists’, actual defence planning should remain a prerogative for professionals, i.e. general staffs, albeit under political control.

Even though the variety of NOD models is immense, they may be grouped into three ‘archetypal’ models (A-C), of which most concrete proposals are derivatives or blends, including models D and E below.

A. Area-covering territorial defence, along the lines of the seminal proposal of Horst Afheldt (1976; 1983), or the (less puristic, but presumably much more effective) spider-and-web’ model of the SAS (Study Group Alternative Security policy, to which the present author belongs). The latter envisages a combination of a stationary, area-covering defence web with mobile forces (‘spiders’), including some mechanized and armoured units. Even though the latter are per se suitable for offensive operations, they should be integrated with, and thereby made dependent on, the stationary web to such as extent as to be very mobile within, but virtually immobilized beyond the confines of the web (SAS, 1989; SAS & PDA, 1994).

B. ‘Selective area defence’ or ‘stronghold defence’, as suggested by members of the SAS group, especially for the Middle East and other regions with low force-to-space ratios and/or long borders. This would imply concentrating a state's defence on certain areas that are politically important (typically the approaches to the national capital or other major population centres) and/or which together will allow for a cohesive defence. The fire coverage afforded by the units in the strongholds will channel an attack, thereby making it more manageable for the mobile forces.

C. Strictly defensive forward defence, for instance by means of a ‘fire barrier’ and/or by means of fortifications and fixed obstacles along the border. In a European context, this would tend to be a rather capital-intensive type of defence, relying to a very large extent on automated fire by high-technology means. Without mobile land forces capable of taking and holding ground this would still be strictly defensive, even if boosted by capabilities for long-range air or missile strikes. For less technologically developed and/or less affluent countries there are numerous other strictly defensive barrier technologies available.

D. The ‘missing link approach’, based on the notion that an otherwise offensive force posture might be rendered strictly defensive simply by omitting one or several components (for instance long-range and/or mobile air defence capability, mobile anti-tank defence, or river-crossing equipment) presumably without critically weakening defensive strength.

All these models have their strengths and weaknesses, hence the attraction of combining elements of them. The mode of combination that has attracted the greatest attention in Europe has been the following, which may also prove most immediately relevant for many Third World conflict regions (the Middle East, Indochina, India-Pakistan, India-China, China-Russia, and Korea, for instance).

E. Disengagement, implying the withdrawal of certain forces (usually the most offensive-capable ones) from the border area to rearward positions, combined with a forward defence by strictly defensive means: typically tantamount to a tank-free zone in the border region, which should be defended by means of infantry armed with anti-tank weaponry. The attractions of disengagement derive from the fact that it would eliminate options of surprise attack and contribute to confidence-building. The depletion zone would simply serve as an early warnin device, since the deployment of proscribed weapons and forces into the zone would alert the other side to the impending attack and allow him to mobilize and prepare for combat. The same logic might suggest the following:

F. ‘Stepping down’, implying a lowering of the general level of readiness. Forces might be cadred (e.g. through a switch to a reserve system) or otherwise prevented from launching surprise attacks, say by a further subdivision of military formations, a separation of munitions from weapons, weapons from platforms, or troops from equipment.

However, the advantages resulting from disengagement as well as from ‘stepping down’ have to be weighed against the risk of potential malign interactions in a crisis period. If the forces withdrawn from the forward line were those possessing the most substantial offensive capability (as in most proposals), to redeploy them into the zone for defensive purposes in an intense political crisis might easily be misinterpreted as preparations for an attack. Counter-intuitive though it may be, crisis stability may thus require that offensive-capable forces be stationed close to their envisaged combat positions and maintained in a high state of readiness, whereas the unmistakably defensive forces could safely be cadred and stationed in the rear.

There are no universally applicable blueprints for NOD postures and strategies, and certainly none that would suit Indonesia as well as Argentina, or Brunei as well as China. Nor have NOD advocates claimed that there were. Indeed one might argue that what NOD postures have in common is their diversity, i.e. their emphasis on specialization and adaptation to special geographical circumstances. Since forces should be designed to operate only within the national frontiers, they could adapt structurally to the features of the terrain, whether these be forests, jungles, marshes, mountains, or whatever. They would no longer need the ability to operate in a wide range of different environments, indeed such an ability would be downright undesirable from an NOD point of view. Specialization might thus serve as a confidence-building measure, i.e. as a manifestation of defensive intentions. It would magnify the difference between the strength of a state's forces fighting at home and abroad, thereby making it all the more plausible that they would fight on their home ground, i.e. defensively.

The land forces are what matters before anything else. Air and maritime forces serve merely auxiliary functions, because they can neither take nor hold ground. However, even though they would constitute no genuine offensive potential on their own, the synergies between the different services imply that they may serve as powerful ’offensive force multipliers’—as demonstrated by the Israeli victory in the 1967 June War and that of the U.N. Coalition during the Gulf War (Freedman & Karsh, 1991). Hence, the maritime and aerial dimensions have to be taken into account when contemplating how to implement NOD. As a minimum, navies and air forces should be compatible with, preferably supportive of, the strictly defensive nature of the ground forces. To elaborate on the imoplications of this criterion is, however, beyond the scope of the present article (for preliminary analyses, see Møller, 1989; Booth, 1994).

7. NOD and Collective Security

Even the most skillful application of the NOD principles described so far cannot solve one fundamental problem, namely that states come in different sizes. Large countries need long-range mobility merely in order to cover their own territory, but this mobility almost inevitably gives them some ability to invade small neighbouring countries.

The aforementioned spider-and-web approach can go a long way towards solving this problem, by making ‘strategic mobility’ a function of political boundaries rather than geographical distances (the mobility of, say, Russian forces may be greater on Russian soil than in Estonia). However, it will not solve the problem entirely, leaving small countries with big neighbours in an uncomfortable situation. Of course, their predicament is not caused by any hypothetical shift to NOD, but neither would the latter completely solve the problem. Small states hence need some kind of underpinning of their conventional defence, preferably one that does not include the acquisition of nuclear weapons, say in the form of alliances and/or collective security arrangements.

Alliances are either directed against somebody else (often for good reasons), i.e. by their very nature adversarial; or they are superfluous; or they may, finally, be something completely different in embryo, namely collective security arrangements. There is nothing wrong with adversarial alliances as such, but they tend to simply replicate the problems providing their raison d'être on a higher level, because alignment breeds counter-alignment. Instead of malign security dilemma-type interactions between states, we tend to end up with the same type of interactions, accompanied by the same risks and costs, only between alliances. This may well be a short-term solution for individual states, but it is hardly a long-term solution for the international system as such, nor for any of its regional subsystems.

Non-adversarial alliances directed against no serious external threat are utterly superfluous—which does not mean that they will automatically cease to be, because of institutional inertia. They face the choice between finding another credible threat against whom (or which) to direct the common effort of their members, or of undergoing a metamorphosis from alliances to collective security arrangements (Carpenter, ed., 1995). Such systems should preferably incorporate former adversaries, ideally be all-encompassing, if only within a region (Kupchan & Kupchan, 1991; Downs, ed., 1994; for a skeptical view, see Betts, 1992).

However, both alliances and collective security arrangements would require forces with longer effective ranges than required for national defence, indeed longer than would be permissible according to NOD criteria. Without such genuine strategic mobility, however, the mutual assistance commitments that alliances and collective security are made off would not be implementable, hence not credible. Other states would, furthermore, need substantial offensive capabilities for liberating territory already conquered by an attacker, especially because the latter, in his ‘attacker-turned-defender’ position, would enjoy some (but not all) of the aforementioned advantages of the defence, while the true defenders would have to fight offensively, i.e. at a disadvantage. NOD might thus rule out collective security-type operations as envisioned in the UN Charter (‘article 42 operations’) and in the Agenda for Peace, i.e. ‘peace enforcement operations’ (Boutros-Ghali, 1992).

This represents a real dilemma, yet fortunately one with a solution: The synergies inherent in combined arms forces imply that not individual force components, but only larger force ‘conglomerates’ are ‘offensive’. The mobile forces with shock power which are indispensable for collective security might thus be rendered incapable of offensive use on a national scale via multinational integration: Individual states might allocate some mobile forces (e.g. part of the ‘spiders’ of the aforementioned SAS model, or the aircraft carriers that are incompatible with NOD) to joint task forces. The latter could thus be provided with the requisite offensive capabilities while the participating states might remain individually non-offensive. The envisaged pledges to earmark forces for the UN might thus finally be met, and the Military Staff Committee might assume command over true UN forces, albeit only on a contingency basis.

Defensive Restructuring-cum-Collective Security

State A

Regional Collective Security

State B

National defence CS contribution 




CS contribution

National defence

Defensive forces Offensive component I(e.g. tanks)

Offensive-capable Task Force


Offensive       component II

(e.g. fighter-bombers)

Defensive forces








Defensive forces

Offensive component III

(e.g. aircraft carriers)



Offensive       component IV

(e.g. logistics)

Defensive forces

National defence

CS contribution

Offensive-capable Task Force


CS contribution

National defence








State C

 Global Collective Security

State D

8. Modes of Implementation

Should a state reach the conclusion that NOD (cum alliance membership or collective security) is a goal worth pursuing, it would face the question how to go about implemention. In principle, there would be three different modes of voluntary implementation, besides that of ‘enforced defensive restructuring’, as occasionally seen in the aftermath of major wars (Tanner, ed., 1992): negotiated arms control, strict unilateralism, and what might be called informal arms control.

NOD proponents have traditionally been skeptical (to say the least) about negotiated arms control, because of the many pitfalls inherent in this approach, above all the unfortunate emphasis on ‘balance’. Even in the bipolar environment, a balance was hard to define, inter alia because of the incommensurability between different types of forces and weapons, the asymmetrical structure of Eastern and Western forces, and the undeniable importance of unquantifiable factors such as the quality of weapons, the morale of troops and the reliability of allies. Balance would be even harder to recognize than to define, because of the propensity for worst-case analysis and ‘double standards’. States would, for instance, tend to compare their standing forces with an opponent's mobilizable potential, etc. Finally, even if it were to be definable and recognizable, balance would provide all concerned with both too little and too much: Too little, because a surprise attack might still overwhelm any defence, also in the case of evenly matched forces; too much, because a well-prepared defender could do with less than parity (vide supra). Until around 1987, this theoretically founded scepticism about the prospects of arms control seemed to receive empirical support from the meagre accomplishments of East-West arms control negotiations, hence the attraction of unilateralism.

Strict unilateralism has been recommended by several NOD proponents, who have recommended states to adopt a NOD strategy without further ado, simply because it is the most effective. Furthermore, regardless of the respective adversary's response, the situation would be stabilized because incentives for pre-emptive attack would be removed. The main problem with such suggestions was, on the one hand, that they were addressed to the wrong side, namely NATO, i.e. the clearly most defensively oriented of the two opposing alliances. What would really improve the situation in Europe would, however, have been a Soviet abandonment of its very offensive strategy. On the other hand, in the light of the Soviet intransigence until Gorbachev's take-over, it seemed futile to directly urge the Soviet Union to change strategy.

Hence the attraction of a mixture between arms control and unilateralism, in casu via a form of gradualism (George, 1988). This would imply making limited but unconditional conciliatory overtures, say in the form of a limited arms build-down, accompanied by an invitation to the respective adversary to reciprocate, with a promise to proceed along the same path if reciprocation would be achieved. In the East-West conflict, NOD might presumably have been one element in such a gradualist strategy of confidence-building, tension reduction and disarmament: By abandoning certain offensive elements in its own strategy and posture, NATO might presumably have induced the USSR to reciprocate in the form of a gradual abandonment of her offensive strategy. Ironically, since the advent of Gorbachev, the roles were inverted, when the USSR assumed the initiative and launched a ‘peace offensive’ that represented a textbook case of gradualism.

Mutatis mutandis, gradualist strategies might prove successful in other settings, either between pairs of adversaries, such as the two Koreas, or even in multipolar settings such as the Japan-China-Korea or India-Pakistan-China triangles. Indeed, they might be especially relevant in environments with a low institutional densitity such as Asia.

9. The Past and Future of NOD?

NOD originated in the divided Europe of the Cold War, and was, from the very beginning, conceived as an escape from its quandary. One might even argue that it was intimately linked with the fate of Germany, in the sense that most NOD proponents were German, and most of those that were not nevertheless had Germany in mind as their target (Møller, 1991). From Germany, the idea spread to the rest of Western Europe, especially to the U.K., Denmark, the Netherlands and the United States (Møller, 1995a).

What placed NOD on the international agenda was, however, the unexpected endorsement of the idea by the Soviet Union in 1986/87, as one element in the ‘new political thinking’ of Gorbachev, that found its most significant expression in the unilateral reductions-cum-restructuring announced before the UN General Assembly in 1988 (MccGwire, 1987; 1991; Bluth, 1990). This Soviet (and subsequently Warsaw Pact) endorsement of NOD placed the topic on the East-West agenda. Inter alia, it paved the way for a Soviet acceptance of the Western approach to conventional force reductions, as eventually manifested in the mandate for the Conventional Forces in Europe (CFE) talks: to ‘reduce the capabilities for surprise attack and large-scale offensive action’, i.e. roughly what NOD advocates had urged for several years (except for the unfortunate exclusion from negotiations of maritime and nuclear forces).

The CFE Treaty of 1990 was thus rightly viewed as a success for the NOD idea, with its reductions of tanks, APCs (armoured personnel carriers), artillery, combat aircraft and helicopters (Sharp, 1991). Even though the CFE accomplishment would effectively have solved the military problem in Europe, it was rendered almost irrelevant by the simultaneous collapse of the entire eastern bloc. Without the systemic conflict, not even a militarily superior Russia would have been much of a (military) problem. Unfortunately, though understandably, NOD has therefore receded into near-oblivion in Europe:

• In the former Soviet Union interest seems to have shifted from the defensive restructuring, that was embarked upon in the Gorbachev period, to collective security arrangements and improved relations with the West. Nevertheless, the new
states have to establish priorities, hence to decide upon a military doctrine and strategic conceptions, in which connection future stability would benefit from a defensive orientation. There is little or no opposition to this notion, only a reduced saliency of the entire defence question, a small group of supporters notwithstanding.
• The situation is similar in the former Warsaw Pact countries, above all the Visegrad states (Poland, Hungary, the Czech and Slovak Republics). To the extent that they have made decisions about their future military orientation, they have (at least declaratorily) endorsed the notion of defensive doctrines. Some expertise on NOD does exist, both in the academic community, in the Defence Ministries and in the armed forces, and some concrete proposals for unilateral NOD-type restructuring and arms control have been developed (SAS & PDA, 1994).

Even though the topic is sorrowly underexplored, NOD would also seem to be relevant for what may become a very serious problem in the coming decade, namely the Greek-Turkish conflict that might well erupt into war, once again. Also, the relations between the ex-Yugoslav states after the peace agreement might benefit significantly from a determined reorientation of the armed forces from offensive to defensive strategic conceptions.

Until very recently, the NOD debate was nevertheless largely a European/North American debate. Not much has been written about regions beyond Europe, or by authors who are neither European nor American. It was in recognition of this limitation that the present author in 1993 founded the ‘Global NOD Network’, with the generous support of the Ford Foundation. The purpose was to ‘plant’ the idea in regions where it might presumably be relavant, and to explore the relevance, and the need for modifications, in these places via a dialogue with regional experts.

The experience so far allows the conclusion that there are, indeed, opportunities for defensive restructuring in several of the countries and regions singled out for investigation, albeit not to quite the same extent:

• In the Middle East the peace process will undoubtedly require substantial arms control agreements—especially as far as a settlement between Israel and Syria is concerned—in which connection NOD may prove a suitable guideline: for instance in the shape of an extended disengagement in the Golan. Furthermore, if or when a Palestine state is established, it will need a defence policy. It is, however, inconceivable that Israel would accept the emergence of a hostile and offensive-capable new state on the West Bank and Gaza Strip, hence imperative that Palestine adopts NOD as its military guideline (Møller, 1994b).
• In Southern Africa, the situation is very propitious for NOD-type restructuring, especially as far as South Africa is concerned, which possesses by far the most formidable offensive capabilities in the region. A powerful rationale for defensive restructuring is the desire to forge peaceful relations with the rest of the subregion, to become the leader of which South Africa seems predestined because of its economic potential (Møller 1994c). There is a considerable interest in NOD in the region, yet no widespread awareness of its concrete implications. Furthermore, there is a certain political support for the idea in the ANC and it has, in rather vague terms, been endorsed in the interim constitution as well as included in the. In March 1996, the aforementioned Global NOD Network arranged a conference on Defensive Restructuring in Southern Africa, which also demonstrated to a vivid interest in the topic.
• In East Asia, NOD is obviously relevant for the two Korean states, both as a possible contribution to a dètente that might pave the way for a gradual unification, and as a guideline for post-unification military build-down and restructuring (Møller, 1995b). Japan's defence policy is already very NOD-like, but it is not self-evident that it will remain so (Matthews & Matsuyama, eds., 1993), yet extremely important for regional stability that it does not develop the formidable offensive capabilities that it might, should it chose to be so. The same holds true for China, which is also largely defensive as of today, but may be heading for a critical expansion of its offensive power.
• In South Asia, NOD-type arms control involving India, Pakistan and China might also stabilize the military situation, thereby also mitigating risks of nuclear proliferation. Moreover, NOD has long had a constituency in India, and interviews conducted by the present author in January 1996 seems to be emerging in Pakistan as well (Singh & Vekaric, eds. , 1989).
• There is also some interest in NOD in South America. The recent abandonment by Argentina and Brazil of the ‘nuclear option’ and the gradual emergence of a regional institutional framework seem to open up some opportunities for defensive restructuring. Indeed, concrete and very detailed proposals have been published for an NOD-type defence of Argentina (Caceres & Scheetz, eds., 1995).

The United Nations has all along played an important role in the global NOD debate. As already mentioned, it was the Palme Commission (working under a UN mandate, albeit formally independent) that promulgated the concept of Common Security inspiring the entire NOD debate. Ever since, the United Nations Institute for Disarmament Research (UNIDIR) has been in the forefront of the debate, initiating and sponsoring research on possible global applications of NOD (UNIDIR, 1990), most recently with regard to the Middle East. Furthermore, in 1990 the General Assembly thus adopted a resolution on Defensive Security (Resolution 45/58-O). Without really defining this (not entirely fortunate) term, the resolution contained some very NOD-like formulations, such as the statement that ‘security concepts and policies should be aimed at enhancing security and stability at progressively lower and balanced levels of armed forces and armaments’ and that defence capabilities should reflect ‘true defensive requirements’. The UNDC, likewise, recommended a very NOD-like approach to regional arms control in its 1993 Guidelines and Recommendations for Regional Approaches to Disarmament (Mason, 1995).

In view of the above, it thus seems premature to write off NOD as a child of the Cold War unable to survive its demise. On the contrary, the time may be ripe for a switch to NOD in several regions around the globe, both as a guideline for national defence planning and for arms control negotiations.


Afheldt, Horst (1976): Verteidigung und Frieden: Politik mit militärischen Mitteln (München: Carl Hanser Verlag).

Afheldt, Horst (1983): Defensive Verteidigung (Reinbek: Rowohlt Verlag).

Axelrod, Robert (1984): The Evolution of Cooperation (New York: Basic Books).

Betts, Richard K. (1992): ‘Systems for Peace or Causes of War? Collective Security, Arms Control, and the New Europe’, International Security, vol. 17, no. 1 (Summer), pp. 5-43.

Bluth, Christoph (1990): New Thinking in Soviet Military Policy (London: Pinter, 1990).

Booth, Ken (1994): ‘NOD at Sea’, in Møller & Wiberg (eds.), pp. 98-114.

Booth, Ken & Nicholas J. Wheeler (1992): ‘Beyond Nuclearism’, in Regina Owen Karp (ed.): Security Without Nuclear Weapons? Different Perspectives on Non-Nuclear Security (Oxford: Oxford University Press), pp. 21-55.

Boutros-Ghali, Boutros (1992): ‘An Agenda for Peace. Preventive Diplomacy, Peacemaking and Peace-Keeping. Report of the Secretary-General Pursuant to the Statement Adopted by the Summit Meeting of the Security Council on 31 January 1992’, in Adam Roberts & Benedict Kingsbury (eds.): United Nations, Divided World. The UN's Role in International Relations, New Expanded Edition (Oxford: Oxford University Press, 1993), pp. 468-498.

Buzan, Barry (1991): People, States and Fear. An Agenda for International Security Studies in the Post-Cold War Era, Second Edition (Boulder, CO: Lynne Rienner).

Caceres, Gustavo & Thomas Scheetz, eds. (1995): Defensa no Provocativa. Una propuesta de reforma militar para la Argentina (Buenos Aires: Editora Buenos Aires).

Canby, Steven L. (1994): ‘Weapons for Land Warfare’, in Møller & Wiberg (eds.), pp. 74-84.

Carpenter, Ted Galen, ed. (1995): The Future of NATO (London: Frank Cass).

Clausewitz, Carl von (1984): On War, edited and translated by Michael Howard and Peter Paret (Princeton, N.J.: Princeton University Press).

Clemenceau, Georges (1841 1929), quoted from Wintle, Justin (ed.). The Dictionary of War Quotations (London: Hodder & Stoughton, 1989), p. 97.

Commission on Global Governance (1995): Our Global Neighbourhood (Oxford: Oxford University Press).

Downs, George W., ed. (1994): Collective Security Beyond the Cold War (Ann Arbor, Michigan: University of Michigan Press).

Freedman, Lawrence & Efraim Karsh (1991): ‘How Kuwait Was Won: Strategy in the Gulf War’, International Security, vol. 16, no. 2 (Autumn), pp. 5-41.

George, Alexander L. (1988): ‘Strategies for Facilitating Cooperation’, in idem, Philip J. Farley & Alexander Dallin (eds.): U.S.—Soviet Security Cooperation. Achievements, Failures, Lessons (New York: Oxford University Press), pp. 692 711.

Gleditsch, Nils Petter & Olav Njølstad, eds. (1990): Arms Races. Technological and Political Dynamics (London: Sage).

Grin, John (1990): Military-Technological Choices and Political Implications. Command and Control in Established NATO Posture and a Non-Provocative Defence (Amsterdam: Free University Press).

Hammond, Grant T. (1993): Plowshares into Swords. Arms Races in International Politics, 1840-1991 (Columbia, SC: University of South Carolina Press).

Herz, John M. (1950): ‘Idealist Internationalism and the Security Dilemma’, World Politics, vol. 3, no. 2, pp. 157 180.

Independent Commission on Disarmament and Security Issues (1982): Common Security. A Blueprint for Survival. With a Prologue by Cyrus Vance (New York: Simon & Schuster).

Jervis, Robert (1976): Perception and Misperception in International Politics (Princeton, N.J.: Princeton University Press)

Jervis, Robert (1978): ‘Cooperation Under the Security Dilemma’, World Politics, vol. 30, no. 2, pp. 167 214.

Jervis, Robert (1989): The Meaning of the Nuclear Revolution. Statecraft and the Prospects of Armageddon (Ithaca: Cornell University Press).

Kokoshin, Andrei A. & Valentin Larionov (1989): ‘Four Models of WTO-NATO Strategic Interrelations‘, in Marlies ter Borg & Wim Smit (eds.): Non-Provocative Defence as a Principle of Arms Control and its Implications for Assessing Defence Technologies (Amsterdam: Free University Press), pp. 35-44.

Levy, Jack S. (1987): ‘Declining Power and the Preventive Motivation for War’, World Politics, vol. 40, no. 1 (October), pp. 82-107.

Kupchan, Charles A. & Clifford A. Kupchan (1991): ‘Concerts, Collective Security, and the Future of Europe’, International Security, vol. 16, no. 1 (Summer), pp. 114-161.

Liddell Hart, Basil (1974): Strategy. The Indirect Approach. Second, revised edition (1967, reprint New York: Signet Books).

Lyons, Gene M. & Michael Mastanduno, eds. (1995): Beyond Westphalia? National Sovereignty and International Intervention (Baltimore: John Hopkins University Press).

Mason, Peggy (1995): ‘The United Nations Guidelines for Regional Approaches to Disarmament’, Disarmament, vol. 18, no. 2, pp. 49-71.

Matthews, Ron & Keisuke Matsuyama, eds. (1993): Japan's Military Renaissance? (New York: St. Martin's Press, 1993).

MccGwire, Michael (1987): Military Objectives in Soviet Foreign Policy (Washington, D.C.: Brookings).

MccGwire, Michael (1991): Perestroika and Soviet National Security (Washington, D.C.: Brookings).

Mearsheimer, John J. (1989): ‘Assessing the Conventional Balance: The 3:1 Rule and Its Critics’, International Security, vol. 13, no. 4 (Spring), pp. 54-89.

Møller, Bjørn (1989): ‘Air Power and Non-Offensive Defence. A Preliminary Analysis’, Working Papers, no. 2 (Copenhagen: Centre for Peace and Conflict Research).

Møller, Bjørn (1991): Resolving the Security Dilemma in Europe. The German Debate on Non-Offensive Defence (London: Brassey's, 1991).

Møller, Bjørn (1992): Common Security and Nonoffensive Defense. A Neorealist Perspective (Boulder: Lynne Rienner and London: UCL Press).

Møller, Bjørn (1994a): ‘Security Concepts: New Challenges and Risks’, in Antonio Marquina & Hans Günter Brauch (eds.): ‘Confidence Building and Partnership in the Western Mediterranean. Tasks for Preventive Diplomacy and Conflict Avoidance’, AFES-PRESS Reports, no. 51 (Mosbach: AFES-PRESS, 1994), pp. 3-49.

Møller, Bjørn (1994b): ‘Non-Offensive Defence and the Arab-Israeli Conflict’, Working Papers, no. 7 (Copenhagen: Centre for Peace and Conflict Research, 1994).

Møller, Bjørn (1994c): ‘The Concept of Non-Offensive Defence: Implications for Developing Countries with Specific Reference to Southern Africa’, in M. Hough & A. du Plessis (eds.): ‘Conference Papers: The Future Application of Air Power with Specific Reference to Southern Africa’, Ad hoc Publication, no. 32 (Pretoria: Institute for Strategic Studies, University of Pretoria), pp. 48-128.

Møller, Bjørn (1994c): ‘The Concept of Non-Offensive Defence. Implications for Developing Countries with Specific Reference to Southern Africa’, Working Paper, no. 10 (Copenhagen: Centre for Peace and Conflict Research, 1994).

Møller, Bjørn (1995a): Dictionary of Alternative Defense (Boulder, CO: Lynne Rienner).

Møller, Bjørn (1995b): ‘Common Security and Non-Offensive Defence: Its Relevance to the Korean Peninsula’, Working Papers, no. 7 (Copenhagen: Centre for Peace and Conflict Research, 1995).

Møller, Bjørn & Håkan Wiberg, eds. (1994): Non-Offensive Defence for the Twenty-First Century (Boulder: Westview Press).

Neild, Robert (1986): ‘The Implications of the Increasing Accuracy of Non-Nuclear Weapons’, in Joseph Rotblat & Ubiratan d'Ambrosio (eds.): World Peace and the Developing Countries. Annals of Pugwash 1985 (London: Macmillan), pp. 93 106.

Nolan, Janne E., ed. (1994): Global Engagement. Cooperation and Security in the 21st Century (Washington, D.C.: The Brookings Institution).

Rathjens, George (1973): ‘The Dynamics of the Arms Race’, in Herbert York (ed.): Arms Control. Readings from the Scientific American (San Francisco: Freeman), pp. 177 187.

Redick, John R. (1993): ‘Nuclear-Weapon-Free Zones’, in Burns, Richard Dean (ed.): Encyclopedia of Arms Control and Disarmament, vols. 1-3 (New York: Charles Scribner's Sons), vol. II, pp. 1079-1091.

Reiss, Mitchell & Robert S. Litwak, eds. (1994): Nuclear Proliferation after the Cold War (Baltimore: John Hopkins University Press).

Rotblat, Joseph, Jack Steinberger & Bhalachandra Udgaonkar, eds. (1993): A Nuclear-Weapon-Free World. Desirable? Feasible?. A Pugwash Monograph (Boulder: Westview Press).

Sagan, Scott D. & Kenneth N. Waltz (1995): The Spread of Nuclear Weapons. A Debate (New York: W.W. Norton).

SAS (1989). Studiengruppe Alternative Sicherheitspolitik: Vertrauensbildende Verteidigung. Reform deutscher Sicherheitspolitik (Gerlingen: Bleicher Verlag).

SAS & PDA (1994). Studiengruppe Alternative Sicherheitspolitik & Project on Defense Alternatives: Confidence-building Defense. A Comprehensive Approach to Security and Stability in the New Era. Application to the Newly Sovereign States of Europe (Cambridge, MA: PDA, Commonwealth Institute).

Schelling, Thomas (1960): The Strategy of Conflict (Cambridge, MA: Harvard University Press).

Sharp, Jane M.O. (1991): ‘Conventional Arms Control in Europe’, in SIPRI Yearbook 1991. World Armaments and Disarmament (Oxford: Oxford University Press), pp. 407-474.

Singh, Jasjit (1993): ‘Nuclear Weapons Threat’, in idem & Thomas Bernauer (eds.): Security of Third World Countries (Aldershot: Dartmouth), pp. 57-79.

Jasjit Singh & Vatroslav Vekaric, eds. (1989): Non-Provocative Defence. The Search for Equal Security (New Delhi: Lancer).

Snyder, Glenn H. (1984): ‘The Security Dilemma in Alliance Politics’, World Politics, vol. 36, no. 4, pp. 461 495.

Stein, Arthur A. (1990): Why Nations Cooperate. Circumstance and Choice in International Relations (Ithaca: Cornell University Press).

Stein, Arthur A. & Robert A. Keohane (1993): ‘Achieving Cooperation Under Anarchy: Strategies and Institutions’, in David A. Baldwin (ed.): Neorealism and Neoliberalism. The Contemporary Debate (New York: Columbia University Press), pp. 85-115.

Tanner, Fred, ed. (1992): From Versailles to Baghdad: Post-War Armament Control of Defeated States (New York: United Nations/Geneva: UNIDIR).

UNIDIR (1990): Nonoffensive Defense. A Global Perspective (New York: Taylor & Francis).

Vasquez, John A. (1993): The War Puzzle (Cambridge: Cambridge University Press).

Wæver, Ole et al. (1993): Identity, Migration and the New Security Agenda in Europe (London: Pinter).