The Other Control Dilemma: Recognition Lag and the Answerable Governance of Capability in Layered Technological Systems

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

In layered technological systems the binding constraint on progress migrates between layers as the system advances, and institutions — acting on representations of the system rather than on the system itself — can continue to allocate resources against an obsolete model of where progress is constrained. This recognition lag is consequential, and where the correctness of a recognition is revealed only after resources are committed, it poses an accountability problem that the governance of innovation, oriented to controlling a technology’s consequences, is poorly equipped to answer. This paper argues that capability recognition nonetheless admits an assessable standard of answerability that most objects of innovation governance do not. It decomposes the lag into four separable events — the migration, its earliest perception within the institution, its confirmability under a reasonable evidential standard, and the allocative response — and argues that, conditional on a specified system objective and feasible intervention set, the migration admits a retrospectively reconstructable referent, in practice often only interval-identifiable. Against that referent the quality of a recognition process can be assessed, so that accountability attaches to whether an institution was organised to detect, integrate and act upon evidence of migration, rather than to whether it predicted correctly, and tightens as evidence matures rather than being imposed with hindsight. Framed as a second dilemma of control alongside Collingridge’s — the apparatus that masters the current constraint is what blinds an institution to its migration — the account is developed across semiconductors, artificial-intelligence compute and quantum computing, and yields governance implications for firms, investors and public innovation agencies: a recognition architecture held broader than operational boundaries, and a recognition audit keyed to process and to evidential maturity.