Objectives:
The article examines the concept of secret science as a distinct epistemic regime that emerges when secrecy is not merely an external administrative limitation imposed on scientific activity, but a constitutive and structural condition of knowledge production itself. The main objective is to conceptualise secrecy as an organising principle of scientific practice and to distinguish secret science from both conventional academic science and applied intelligence work. To achieve this, the article combines conceptual analysis with a historically grounded case study.

Methods:
The study employs qualitative methods drawn from science and technology studies and the history of science, including interpretive analysis of published literature, declassified sources, and selected private archival materials. Particular attention is given to reading archival silences and fragmentary documentation along the grain of secrecy, treating absence and restriction as analytically meaningful rather than as simple evidentiary gaps.

Results:
The analysis identifies four necessary conditions of secret science: epistemic sealing, method under constraint, structural secrecy, and existential necessity. It demonstrates that the Franco-Polish cryptologic collaboration of the 1930s fulfils all four conditions and therefore constitutes a fully articulated instance of this epistemic regime rather than a transitional or auxiliary form of scientific activity.

Conclusions:
The findings support the reclassification of early cryptology as part of the history of science, not solely as an object of intelligence studies. More broadly, the article shows that under conditions where openness is epistemically impossible, secrecy can function as a productive and coherent foundation for scientific knowledge-making.