Jeff Lewis has a good article in the Bulletin of Atomic Scientists on minimal vs. robust deterrence. The latter, advocated by folks like Albert Wohlstetter, asserts that nuclear deterrence is a delicate flower that will collapse if not nourished properly by thousands of warheads in multiple launch configurations. The former suggests that states that can be deterred from launching a nuclear attack will be so deterred by even a minimal chance of second strike response. Jeff Lewis is squarely in the latter category, as am I. The cost of absorbing a nuclear response from even a small nuclear power is so high, even for continent-spanning states, that a nuclear offensive will appear virtually suicidal. France, for example, could easily have destroyed the industrial heart of the Soviet Union if the Soviets (who had presumptive superiority) had launched an attack, even excluding the likely response of the other nuclear powers.
The question continues to have some policy relevance. For example, the irritatingly stupid arguments about how we need the RRW (Reliable Replacement Warhead) program in order to make our deterrent capacity credible vanish if minimal deterrence is taken seriously. Similarly, if minimal deterrence holds then there's no need for the United Kingdom to pursue ridiculously expensive replacements for its Vanguard class nuclear ballistic missile submarines; the need to hide from Soviet attack submarines has vanished, and no conceivable aggressor will be more deterred by the submarines than by some other, much cheaper delivery system.
--Robert Farley