ABSTRACT
Schelling’s analysis is based on assumptions that are epistemologically implausible, inconsistent measured by his own ontological assumptions and marred with distorted empirical assessments. Thinking about constellations in a nuclear-weapons-free world is a complicated epistemological operation. We are trying to construct knowledge about a non-existing context. Principally, there are two methods available for doing so, both working with counterfactual methodology.2 Counterfactual forward induction starts from where we are and changes a variable, then constructs the world with this changed variable at time t + 1, changes the next variable, constructs the world in t + 2 and so on, until a plausible path to the chosen final state of the world is completed. Backwards induction starts with the final state f, constructs the state f − 1 in which all the preconditions are present to move towards f, and eliminates, in a logical sequence, these conditions one by one until we are back to our present world. It is obvious that either process is fraught with speculations, but it is the only plausible method to go from here to there. Schelling, however, uses a rough, even brutish counterfactual operation: he starts from the assumption that the parameters of a non-nuclear world remain unchanged compared to the present one but for one single element, the existence of nuclear weapons. However, the scenario in which all parameters remain the same while nuclear weapons disappear is epistemologically a non-starter. It is inconceivable that governments would move beyond minimum nuclear deterrence when they believe a war probable enough to make their security ultimately contingent on having nuclear weapons (if not immediately available, then at least quickly reconstitutable). Such governments would stop at perhaps 50, 100 or several hundred deliverable nuclear weapons and leave it at that, while something would be ‘left to chance’.3 It is obvious that states are not willing to renounce nuclear weapons under current political and military circumstances. What would drive them to do so if nothing changes? Schelling makes the same mistake as many other pundits of nuclear deterrence: assuming that everything can remain equal in the process of disarmament and that zero can still be achieved (funnily enough, those who ask for immediate nuclear disarmament start from the same assumption). Nuclear disarmament
with the aim of reaching a real zero is a large-scale effort at political reengineering. It is bound to progress in small steps that states undertake because they believe these steps enhance their security or, at a minimum, do not diminish it. Successful steps may enhance mutual trust and thus encourage the parties to go further, and may lead to new requirements (e.g. in transparency and verification) whose implementation leads to even more trust accumulated. The major powers would also undertake efforts to sort out the problems they have with each other. In addition, the lower they go in terms of overall nuclear weapons holdings the more common interest they develop in keeping third parties from crossing the nuclear threshold; at some point, the joint security interest in keeping the nuclear door closed will surpass the interest to steal a geostrategic march on their peers by offering protection to a would-be proliferator. Going down to low numbers and envisaging the possibility of zero enhances their stakes in order and stability. This moves them in the direction of a great-power concert which might provide an indispensable structure for a zero-nuclear world, as has been argued elsewhere (Müller 2010, 33-66). It is particularly implausible that it should be so easy to hide away fissile material for a small arsenal of nuclear weapons. It might sound credible at first but not on further consideration. No government would store weapons-grade material ‘in a refrigerator’ or ‘in a well’ and leave it at that, as Schelling appears to insinuate. Bomb-usable fissile material would be thoroughly guarded and fenced in order to prevent unauthorized actors from obtaining it. These security measures would have signatures that could be picked up by verification agencies who would enjoy greater authority and access rights than in the present world (because otherwise states would not lay down their nuclear arms). In order to ensure true reconstitution capabilities, governments would also have to maintain many technical experts, who would have to practise their reconstitution job lest they risk failing in the hour of truth. Such clusters of expert people practising their future breach of the rules in appropriate facilities would also leave a significant detectable signature. And that brings us back to the unrealistic assumptions – the verification system of a nuclear-weapons-free world would be geared towards picking up such signatures in order to fulfil its early-warning mission.4
