世 光 · 世 应 实 验 室
AAG 2027 · PAPER
Attributive Ambiguity: How Anticipated Deniability Deforms State Disaster Capacity
Zhijun He · The Whitney and Betty MacMillan Center for International and Area Studies, Yale University
Looking for a session
For session organizers
This paper is not yet assigned to a session, and I would like it to be. If you are organizing a session it would fit, you need only my abstract code — 53271 — to add me. Write to zhijun.he@yale.edu and I will confirm the same day.
The session organizing deadline is 5 November, with an open add period from 6 to 12 November during which organizers do not need my permission to add the abstract. After that, unassigned paper abstracts convert to posters.
Sessions this would fit
- Critical disaster studies; the politics of disaster attribution and "natural" disaster framing.
- Political geography of blame, accountability and the state; policy narratives and causal-story analysis.
- Climate adaptation governance and adaptation finance; the political economy of resilience.
- Himalayan environmental change; cryospheric hazards and hydropower.
- Quantitative or mixed-method approaches to state capacity, and critiques of indicator-based capacity assessment.
I am equally glad to be a discussant or to join a panel where the paper is a poor fit but the argument is useful. If you think the paper belongs in a session that is not on this list, I would rather hear that than not.
Abstract
As submitted and accepted
Abstract 53271 · Paper
A single disaster can be truthfully described as both a climatic extreme and a failure of governance. This paper theorizes dual attributability, which I call attributive ambiguity, as an ex-ante force in resource allocation rather than post-disaster rhetorical competition. Critical disaster studies have long established that disasters are never natural; I argue that states treat this constructedness as an asset to be drawn down in advance. Because decision-makers anticipate that losses can later be settled into a "natural disaster" frame, the political cost of suppressing risk assessment and protective redundancy approaches zero. States therefore invest in visible, credit-claimable response capacity while neglecting prevention capacity that is invisible and legible only in failure.
Climate change intensifies this mechanism. It raises risk in real terms and hands every governance failure an alibi; non-stationarity makes this worse, since "once-in-decades" language presupposes a climatic baseline that no longer exists. The long-run consequence is not general decay of state capacity but deformed specialization: response strengthens as prevention lags. This is an equilibrium of the legitimation circuit, not a malfunction.
Building on Stone's causal stories and blame-avoidance theory, and engaging Pielke's critique of disaster attribution, I assess the concept through "settlement traces": attributive wording in official texts, the scope of accountability, reconstruction standards, and compensation channels. Two Himalayan cases ground the analysis: Chamoli (2021) and the South Lhonak glacial lake outburst (2023). I close with falsification and scope conditions: Derna's 2023 dam collapse, where the natural-disaster frame failed to hold, probes the mechanism's limits.
What the paper is arguing
The move, in three steps
1 · Attribution is treated as an asset, not a risk
Critical disaster studies established decades ago that no disaster is natural. The usual next step is to read post-disaster framing as rhetorical struggle — who succeeds in making the event look like weather. This paper moves the mechanism earlier. If decision-makers can anticipate that losses will later settle into a natural-disaster frame, then that frame is available to them before the event, as something to draw down. It changes what they build, not only what they say afterwards.
2 · The consequence is specialization, not decay
The prediction is not that state capacity erodes. It is that capacity deforms: response strengthens because it is visible and creditable; prevention lags because its success is an absence and its failure is deniable. A state can therefore look increasingly capable on every conventional measure while becoming less able to stop the next event. I call this deformed specialization, and I read it as an equilibrium of the legitimation circuit rather than as dysfunction to be corrected by better management.
3 · Climate change is not only a hazard multiplier
It raises risk in real terms and, at the same time, supplies an alibi for every governance failure that follows. Non-stationarity sharpens this: the language of "once-in-decades" events presupposes a stable baseline that no longer exists, which makes the natural-disaster frame cheaper to invoke and harder to contest.
Settlement traces
How the concept is assessed
An anticipated attribution leaves no direct record. What it leaves is a settlement: the point at which an event's cause is fixed in official practice. The paper reads four traces of that settlement, each independently documentable:
- Attributive wording in official texts. How the cause is named in the documents that carry legal and budgetary weight, and how that wording moves over time.
- The scope of accountability. Who is investigated, at what level, and where the inquiry stops.
- Reconstruction standards. Whether rebuilding restores the prior standard or raises it — the clearest signal of whether the event was accepted as a governance failure.
- Compensation channels. Whether payment flows as disaster relief or as liability, which are different admissions.
The four are chosen so that they can disagree. A case in which official wording says "natural" while reconstruction standards rise is telling us something specific, and the framework has to be able to record that rather than average it away.
Three cases, three jobs
What each case is doing in the argument
Chamoli, Uttarakhand · 2021
Establishing the mechanism
A rock-ice avalanche in the upper Rishiganga catchment generated a debris flow that destroyed hydropower infrastructure downstream. The event sits squarely in the ambiguous zone: it can be read as a cryospheric extreme, and it can be read as a question about siting and about what was known before construction. The case is where the settlement traces are read first, because both readings remained live long enough to be observable.
South Lhonak Lake, Sikkim · 2023
Testing the "known risk" prediction
A glacial lake outburst flood from a lake whose hazard had been documented in the scientific literature for years before it occurred. This is the harder test. If the mechanism is real, prior documentation should not have translated into proportionate prevention, because prevention here was invisible and its absence deniable. If instead prior documentation did drive protective investment, the mechanism is weaker than claimed.
Derna, Libya · 2023
Scope condition and falsification probe
Two dams failed during a storm and the natural-disaster frame did not hold: maintenance deficits were already on record, and accountability did not settle into weather. The case is in the paper precisely because it is where the mechanism appears to fail, and it asks what conditions have to obtain for anticipated deniability to work at all — state coherence, documentary trail, the presence of an actor with standing to contest the frame.
The three are ordered deliberately: one to establish, one to test the sharpest prediction, one to find the edge. A framework that only works on the case it was built from is a description, not a theory.
Why this is not only an academic argument
What follows from it
If the mechanism holds, then any instrument that measures a state's disaster capacity as a single quantity will systematically read response capacity as adaptation capacity — and will do so in the direction that suits the state being assessed. That is not a marginal measurement problem. It is the basis on which climate adaptation finance decides where money can be absorbed.
The operational half of this argument, including the prior art we got wrong and the indicators we are testing, is set out separately: the method. The wider intellectual position it comes from is here.
Session organizers, and anyone who thinks this is wrong
Zhijun He — zhijun.he@yale.edu · Abstract code 53271
Research programme: seesaltorg.org/zhijun/research · Full thesis: After the Catastrophe: Rwanda's Transformation
The abstract above is reproduced as accepted. Meeting details from the AAG: 2027 Annual Meeting, New York City, 8–12 February 2027, theme "Grounded Relations and Desirable Futures".
播 种 光 明 ， 收 获 永 恒