Toward a paradox-oriented paradigm of knowing: the form theory of observation and quality in empirical sciences.

We develop a form-theoretical account of knowing in empirically oriented sciences and argue for a paradox-oriented understanding of epistemic quality. Drawing on Spencer-Brown and Luhmann, we start from the idea that observation always operates by drawing a distinction and indicating one side. What appears as an object or "phenomenon" is thus the marked side of a form; its unmarked side persists as horizon. This implies an origin paradox of knowing: any knowing system can only operate by means of distinctions whose own contingency cannot be observed in the very act of use. We reconstruct how this paradox is functionally invisibilized and proliferates within science understood as a social system operating under the guiding distinction true/not-true. Programmes (methods, designs, standards), organizations and memory structures (journals, universities, reviews) distribute the paradox across three dimensions of scientific communication: in the factual dimension, as object formats, evidential hierarchies and growing residual categories; in the social dimension, as roles, communities and responsibility attributions; and in the temporal dimension, as replication cycles, escalation programmes and crisis semantics. This stabilizes an implicit quality paradigm that equates quality with coherence, consensus and cumulative stability. As an alternative, we propose a paradox-oriented quality concept: quality in this perspective lies in how a system addresses the constitutive paradox of its own observation without blocking its autopoiesis. We conceptualize quality as a calibrated rhythm of invisibilizing and briefly exposing the form of observation, increasing both robustness and the range of next possible operations. While developed with reference to empirical science, the argument generalizes to any knowing system.