One theme in the Stimson Center’s Escalation Control and the Nuclear Option in South Asia (2004) is that deterrence is an abstract notion that sometimes fails real world tests. Crisis management and escalation control then become paramount, as was the case during the limited war in 1999 between Pakistan and India after Northern Light Infantry troops crossed the Kashmir divide in the heights above Kargil.

Kargil was one of four crises between India and Pakistan from 1990 to 2008 in which the United States served as the essential crisis manager. Washington’s well-used playbook is all about preventing the outbreak of hostilities. There is no reliable playbook for escalation control once a crisis transitions to hostilities between nuclear-armed states.

The subject of escalation control in limited warfare received considerable attention during the Cold War. Here’s a shoebox sampler of quotes:

“Although undesired escalation obviously does not occur all the time, the danger is always present. The room for misunderstanding, the pressure to act before the other side has seized the initiative, the role of unexpected defeats or unanticipated opportunities, all are sufficiently great – and interacting – so that it is rare that decision-makers can confidently predict the end point of the trajectory which an initial resort to violence starts.” — Robert Jervis, The Illogic of American Nuclear Strategy (1984).

“While not incurring a serious risk of an immediate all-out response, and while causing some physical attrition of the enemy’s power to move on the ground, tactical reprisals would still serve the bargaining function by demonstrating a willingness to “up the ante” and to continue doing so until the other side agreed to settle the war…Reprisals against forces, especially tactical forces, allow us to demonstrate this possible intent at minimum provocation and at minimum initial damage to our own economy and population.” — Glenn Snyder, Deterrence and Defense (1961).

“The curtailing of our taste for unequivocal victory is one of the prices we pay to keep the physical violence, and thus the costs and penalties, from going beyond the level of the tolerable.” — Bernard Brodie, Strategy in the Missile Age (1959).

“[Limited war is] part of a general ‘strategy of conflict’ in which adversaries would bargain with each other through the medium of graduated military responses, within the boundaries of contrived mutual restraints, in order to achieve a negotiated settlement short of mutual destruction.” — Robert Osgood, Limited War Revisited (1979).

Two of the best thinkers in the U.S. strategic firmament, Robert Osgood (Limited War: The Challenge to American Strategy, 1958) and Bernard Brodie (Escalation and the Nuclear Option, 1966), tried to make the case for limited wars below and even across the nuclear threshold. Both were wise enough to have had second thoughts. The conduct of a limited war between nuclear-armed adversaries calls to mind Oliver Hardy’s complaint to Stan Laurel: “another fine mess you’ve gotten me into.” Getting out of this mess presumed that, despite their differences, adversaries could remain on the same page – a questionable assumption if they have stumbled into war or have engaged in combat because of irreconcilable differences.

In Limited War Revisited (1979), Professor Osgood acknowledged the weakness of his earlier analysis: “One trouble with all strategies of local war in Europe is that the Soviet Union has shown virtually no inclination to be a partner to them.” While U.S. strategists were hypothesizing rungs for the escalation control ladder, it turned out that the Soviet General Staff was planning for a blitzkrieg accompanied by a large number of nuclear detonations across Europe. This cautionary note now applies to India and Pakistan.