A Dependency Hypothesis is a defeasible claim that specified relata stand in a dependency relation under stated conditions.
A dependency hypothesis does not determine in advance what kind of dependency obtains. The proposed relation may be causal, constitutive, enabling, constraint-based, structural, spatial, temporal, historical, conceptual, or mixed. The relation type must therefore be stated rather than inferred from the fact that one phenomenon is said to depend upon another.
Within APS, pathway arrows represent dependency hypotheses. The arrow itself is not evidence and does not establish causation, necessity, chronology, or organisational priority. Each claim must be unpacked sufficiently to identify the relevant relata, relation type, material or argumentative basis, evidential support and limits, alternatives, scope, and possible failure conditions.
Where these elements are explicitly specified for assessment, the more precise expression typed dependency hypothesis may be used.
Dependency hypotheses may be retained, qualified, retyped, rejected, redescribed, judged redundant, or assigned principled no gain. Their scientific standing therefore depends upon the evidence and argument supporting the particular relation claimed, together with comparison against the strongest relevant established explanation.