A recent strand of Carnegie Endowment analyses by George Perkovich and others suggests that South Asian crisis behaviour under the nuclear shadow is more stable than widely assumed.
Much of what has long been interpreted as nuclear signalling is recast as rhetorical noise, with the absence of leadership ultimatums taken to mean that escalation risks are overstated. The implications of these analyses are incorrect and consequential: that controlled conventional operations may be feasible under the nuclear overhang and that thresholds remain distant. This reading risks reinforcing the belief that escalation can be controlled on favourable terms, an assumption that South Asia’s crisis record does not support.
This interpretation is analytically elegant but strategically incomplete and dangerous, as it risks converting restraint into a permissive space for escalation. More importantly, this line of analysis is not without policy consequences. By discounting ambiguity and indirect signalling, it lowers the perceived cost of escalation and risks informing policy choices that treat deterrence as a permissive space rather than a constraint. It rests on a narrow conception of credible signalling that privileges observable indicators and declaratory clarity over doctrinal intent, perception and crisis behaviour. In doing so, it mistakes restraint for absence and ambiguity for incoherence.
Pakistan-India crises – Kargil (1999), the 2001–02 standoff, Mumbai (2008), 2016, Pulwama-Balakot (2019) and the 2025 crisis – tell a different story. Large-scale war has been avoided not because nuclear risk is marginal, but because deterrence operates through calibrated ambiguity, incremental signalling, and disciplined management. The problem is not the absence of nuclear threats, but their misrecognition.
At the centre of this instability lies the unresolved Kashmir dispute, which continues to generate crises in a nuclearised environment of compressed timelines and weak crisis communication. South Asia remains a nuclear flashpoint not simply because of capabilities, but because unresolved disputes are increasingly shaped by a Hindutva-driven strategic outlook that seeks to normalise coercion. Stability since 1998 reflects not structural equilibrium, but active management under asymmetry, including repeated demonstration of restraint where escalation could have been expanded.
The Carnegie framework usefully disciplines a discourse often prone to exaggeration by distinguishing gestures from explicit threats. Yet its insistence on visible preparations and top-level attribution discounts subtler mechanisms through which deterrence operates. In South Asia, where readiness patterns are evolving and timelines are compressed, actions presented as limited can quickly be interpreted as preparatory, narrowing the margin for calibrated signalling and increasing the risk of miscalculation.
Deterrence is fundamentally psychological. As Thomas Schelling observed, the power to hurt is most effective when held in reserve. Credibility derives from the adversary’s belief in possible consequences, not from constant display. Nuclear powers routinely rely on uncertainty, indirect communication and calibrated signalling. South Asia is no exception; indeed, it reflects these dynamics in one of their most compressed and demanding forms.
A further danger lies in the weakness of direct diplomatic and military communication during crises. In nuclear settings, sustained bilateral contact is a stabilising necessity, not a political concession. Where such channels are downgraded or avoided, misperception deepens, signalling becomes harder to interpret and reliance on outside intervention becomes more likely.
In 2001–02, missile tests and mobilisations conveyed resolve without overt nuclear ultimatums. In 2019 and 2025, Pakistan’s calibrated conventional responses helped re-establish deterrence balance without immediate wider escalation. These episodes did not demonstrate a safe, conventional space under the nuclear shadow; they showed how quickly even limited action can trigger perilous escalation pressures. The absence of explicit threats in lower-intensity episodes reflected prudent assessment, not deterrence failure. Much signalling targets not just the adversary but external actors whose intervention often ends crises.
The historical dependence on outside, often US, intervention to arrest escalation reflects not regional resilience but the inadequacy of indigenous crisis-management mechanisms. Reclassifying these patterns as ‘noise’ obscures their cumulative power. Pakistan’s crisis behaviour reflects resilience combined with restraint. Pakistan’s past restraint should not be misread as evidence that escalation pressures are indefinitely manageable. No deterrence system can rely indefinitely on one-sided caution in the face of repeated escalation probes.
The bluff narrative’s real danger lies in underestimation. When calibrated signals are dismissed as noise, the incentive structure shifts; thresholds appear distant, risks appear manageable and limited war begins to look feasible. This is precisely how miscalculation becomes embedded in doctrine, encouraging boundary-testing and normalising the illusion of ‘limited war space’ under the nuclear overhang.
Yet South Asia’s geography, frequent crises, domestic pressures, and accelerating technologies (such as canisterisation of missiles and precision strikes) compress escalation dynamics. Actions intended as controlled quickly trigger counter-responses that narrow manoeuvre margins. The stability-instability paradox was never a licence for safe, limited conflict. If calibrated signals are dismissed as noise, the perceived cost of risk-taking falls. In a nuclearised environment, such risk-taking behaviour and illusions are not stabilising; they are pathways to catastrophic miscalculation.
The nuclear bluff is not a property of South Asian deterrence. It is a misreading of it – one that risks reshaping strategic behaviour in dangerous ways. If restraint is interpreted as weakness, and ambiguity as absence, the incentive structure shifts toward risk-taking. In a region where escalation compresses rapidly, such miscalculations will not remain contained.
Ambiguity is not a flaw but a familiar feature of deterrence practice. It preserves uncertainty, sustains strategic flexibility and complicates confident assumptions about controllable escalation. Major powers rely on such ambiguity routinely. Treating it as sophistication in some cases but manipulation in South Asia reflects an analytical double standard that distorts both understanding and policy.
Restraint in crises reflects reciprocal recognition, not unilateral illusion. Stability since 1998 is real but conditional – managed through credible capabilities, signalling, institutional control and caution under uncertainty.
Deterrence has held not because risks are low, but because decision-makers have demonstrated sustained awareness of these risks. In future crises shaped by tighter operational integration, faster decision cycles and emerging technologies, the space for diplomatic off-ramps will narrow significantly. Crisis termination may not keep pace with escalation. In such an environment, escalation is likely to outpace diplomacy and the expectation of timely external intervention or controlled termination may prove illusory.
South Asia remains a nuclear flashpoint not because deterrence is weak, but because it operates under structural stress – a Hindutva-driven coercive outlook, technological acceleration, unresolved disputes and compressed timelines. Attempts to test deterrence on the assumption that escalation can be controlled will not preserve stability; they will narrow signalling space, deepen crisis instability, and increase the risk of miscalculation. In such an environment, assumptions of escalation dominance may outpace the crisis-management mechanisms needed to arrest escalation before it becomes far harder to contain.
Deterrence has held because consequences remain credible and uncertainty remains intact. To treat that uncertainty as a bluff is to invite its collapse and, in a compressed, nuclearised environment, such a miscalculation will not remain constrained.
The writer is a former brigadier with expertise in deterrence, arms control and disarmament affairs.
