The Kahneman-Tversky, Gigerenzer, and Stanovich traditions have been read as cataloging biases. They have been cataloging probes. Each canonical paradigm was constructed by an experimenter to elicit a specific judgment. The data each paradigm produces probes one or more dimensions of the four-dimensional judgment process. The four dimensions are recognition, contextualization, resolution, and action. They are ordered by dependency. Each dimension is a necessary condition for the next. The six pairwise reversals of the ordering produce six structural classes (4-3, 3-2, 2-1, 4-2, 3-1, 4-1). The probes' class signatures fall into these classes. The cycle runs continuously. The recursion joint propagates each cycle's commitment to the next cycle's recognition. Class signatures index within-cycle failures on the recursion-delivered content. The "bias" the paradigm produces is the gap between the participant's recursion-accumulated default at the probed dimension and the experimenter's normative target. The structural account classifies probes, names what each probe targets, and reads the participant's response as the operational output of the recursion-accumulated default content at the targeted dimension. The canonical paradigms distribute across three categories. Clean single-class paradigms occupy four dependency-violation classes. Heterogeneous multi-class paradigms contain multiple class instances at the participant level. One separability demonstration isolates two positions as distinct operational stages. A fifth dependency-violation class is joint-occluded under the recursion-joint specification (the joint's normal operation loads the new cycle's contextualization with the prior cycle's pragmatic output before the recognition-to-contextualization step can fire on equal footing) and unrepresented by canonical paradigms. The joint-occluded label is distinct from the structurally empty cell at the 4-2 relation, where the edge does not exist in the architecture's edge set at all. The class signatures sort as follows. The 3-1 signature catches probes where the resolution operates on a recognition admission that does not include the targeted referent (Linda, availability, representativeness). The 3-2 signature catches probes where the resolution operates on a contextualization frame whose recursion-accumulated content is not the frame the situation called for (cab problem, anchoring, Asian disease, IKEA, endowment, hindsight, fundamental attribution error, base-rate neglect, confirmation bias, planning fallacy, in-group bias, status quo, false memory). The 4-3 signature catches probes where action operates without resolution having delivered against the currently relevant content (belief perseverance, sunk-cost). The 4-1 signature catches probes where the participant's action commits forward content the participant's recognition did not admit (confabulation, where orbitofrontal disruption breaks the recognition admission at the substrate level). The 2-1 signature is the within-cycle failure of the recognition-to-contextualization dependency. It is joint-occluded under the recursion-joint specification (the joint's grip on the frame occludes the within-cycle isolation the class requires) and is not represented by a canonical paradigm in the current catalog. Devine's three-study sequence is a separability demonstration of the contextualization / resolution distinction rather than a single class. The bystander and bandwagon effects are multi-signature paradigms whose data isolates multiple subtypes (3-1, 3-2, and 4-3 instances at the participant level) under the experimenter's flattened label. The catalog hands three openings forward. The first is a falsifiability prediction. Probes targeting the same dependency should produce correlated responses across participants on the same content domain. Probes targeting different dependencies should not, regardless of the named-bias label. For heterogeneous multi-class paradigms the prediction operates at the subtype level, with each subtype correlating with paradigms targeting the same dependency rather than the paradigm-as-a-whole correlating with other paradigms. The second is a construction problem at the 4-2 dependency, empirically unrepresented and structurally empty by the architecture's closure. The third is a construction problem at the 2-1 dependency. The within-cycle case requires a design that isolates the experimental cycle's recognition admission against the recursion-delivered frame. The correlation prediction's operationalization is bound by the architecture. The content-matching criterion is the recursion-accumulated frame the experimental cycle activates at contextualization (Stewart, 2026g). The dependency-versus-content separation is forced by the dimensions' invariance across participants (Stewart, 2026a, 2026g). The 3-2 block's structural homogeneity is forced by the rank-one convergence at resolution (Stewart, 2026a, 2026e). The 4-1 class's testability on healthy samples follows from graded admission-threshold defaults at recognition. Multi-class subtype resolution follows from the cycle's two-direction readability at resolution. The operationalization removes the analyst's content-matching and subtype-resolution latitude. The prediction is bound at the architecture rather than at the analyst's discretion. The philosophical companion (Stewart, 2026w) develops the architectural argument against the bias catalog. The structural companion (Stewart, 2026ay) develops the four-dimensional architecture and the six-class taxonomy at the cognitive scale this paper applies. **Keywords:** experimental probes, dimensional judgment, four-operation ordering, cognitive bias paradigms, default setting, structural typology, probe-to-dimension assignment, heuristics-and-biases tradition
Arthur Stewart (Sat,) studied this question.