5XÉçÇø

Crisis research is retrospective by design. Protest waves, institutional breakdowns, and conflict onsets are coded once they become publicly visible, and explanation works backwards from the moment of rupture. The period before rupture is left underspecified, treated as background rather than as a phase with dynamics of its own. The difficulty is not only empirical. There is little agreement on what a pre-rupture phase consists of, or what evidence could establish that something was building.
This panel treats that as a problem of concepts and methods rather than of forecasting, and invites papers on any part of it.
A first set of questions is theoretical. Is there a pre-rupture phase that can be specified in advance, or is its appearance an artefact of hindsight? Where private preferences stay concealed until the moment of cascade, observable early movement may be unavailable in principle. Where strain accumulates slowly and crosses thresholds, it may be traceable, but only under a different account of causation and timing.
A second set concerns what a signal looks like. Some approaches read the statistical behaviour of a series without regard to its content, watching for rising variance or slower recovery from disturbance. Others measure substantive conditions: institutional predictability, belonging, the horizons people hold for their own futures. Digital traces, textual corpora, administrative records and survey series each carry different lags, silences and biases.
A third set is reflexive, and cuts in two directions. Retrospective designs invite reading signals backwards from known outcomes, since indicators are sought where rupture is known to have occurred, and rarely tested against the quiet years that resemble them. Rupture is also rare, leaving false positive rates severe for any plausible indicator. Beyond measurement, claims to anticipate crisis do political work of their own, constituting some populations as objects of monitoring and settling in advance what will count as risk.
We welcome conceptual, formal, computational and interpretive contributions, including papers arguing that early detection is unattainable and explaining why. Case-based and comparative work is equally suitable. The aim is to sharpen retrospective method, not to furnish an instrument for preemptive intervention.