Ostensieve

Ostensieve is a contemporary conceptual framework used in philosophy of science, media theory, and AI interpretability to describe a mode of making hidden mechanisms visible through guided demonstration and selective revelation. The neologism blends ostension—showing or pointing to something directly—with sieve, indicating filtration of information to expose core structures.

In practice, Ostensieve involves three ideas: ostensive demonstration, where the process is shown through concrete examples or stepwise elucidation; sieve-like reduction, which strips away noise to reveal essential components or patterns; and reflexivity, where the observer's role is acknowledged and accounted for.

Origins and usage: The term emerged in early 21st-century discussions about transparency and explainability, particularly in digital technologies and data science. It has been applied to AI interpretability methods that present models’ inner logic in a staged and accessible way, and to journalism and governance practices aiming to reveal causal chains behind complex events.

Applications and examples: In AI, Ostensieve methods include visualization of feature attributions via progressive disclosure; in science communication, it refers to presenting experimental workflows as a sequenced, interpretable narrative; in policy, it denotes public-facing disclosures of data pipelines.

Criticism: Critics argue that Ostensieve can oversimplify complexity, mislead by cherry-picking demonstrations, or neglect emergent properties that are invisible in stepwise demonstrations.

See also: interpretability, transparency, explainable AI, data visualization.