Overview
A philosophical paper that reveals the epistemological foundations of Deming's later management thinking. Published in 1975, it builds on the foundational distinction established in on-the-distinction-between-enumerative-and-analytic-studies twenty-two years earlier, extending the argument into the domain of decision-making and action.
Key Arguments
Deming argues that statistical inference alone cannot provide a basis for action — one needs a theory, a prediction about the future, which statistical analysis can inform but never replace. A probability statement about past data describes what was; it says nothing about what will be unless accompanied by a theory that connects past conditions to future ones. This is not a minor caveat but a fundamental limitation of statistical reasoning that most practitioners overlook.
The paper connects Deming's statistical work to C.I. Lewis's pragmatic epistemology — the philosophical position that knowledge is inherently predictive and that the meaning of any concept lies in its implications for future experience. Lewis's influence on Deming is often noted but rarely examined in detail; this paper is where the connection is most explicit. The argument anticipates the theory-of-knowledge component of the system-of-profound-knowledge: knowledge requires prediction, prediction requires theory, and theory must be tested through the pdsa-cycle-plan-do-study-act.
Significance
The paper is the intellectual bridge between Deming's statistical career and his management philosophy. The claim that action requires theory beyond what data alone can provide is the epistemological foundation for Deming's insistence that "there is no knowledge without theory" and his critique of management-by-the-numbers. Managers who rely on quarterly reports and performance metrics without understanding the processes that generated them are committing exactly the error this paper identifies: treating analytic problems as if they were enumerative.
Source
Free full PDF available from the Deming Institute at the URL above. Also available from the ELFT NHS digital library. This paper is essential reading alongside on-the-distinction-between-enumerative-and-analytic-studies for understanding the epistemological core of Deming's thought. Link to education-and-early-statistical-career-1900-1927, walter-a-shewhart.