This paper proposes to recover the topic of the philosophy of scientific method from its late nineteenth-century roots. The subject matter of scientific method sprouted from key inferential ingredients identified by Charles Peirce. In this paper, the historical path is traversed from the viewpoint of contemporary Cognitive Structural Realism (CSR). Peirce’s semiotic theory of methods and practices of scientific inquiry prefigured CSR’s reliance on embodied informational structures and experimentation upon forms of relations that model generic scientific domains. Three results are shown to follow from this convocation: (i) a naturalization of Peirce’s interconnected abductive, deductive and inductive stages of the logic of science, here characterized de novo in terms of CSR. (ii) a perspective to scientific modeling that incorporates processes of abstraction and generalization as originated from Peirce’s logic of science, and (iii) diagrammatic reasoning as a pivotal method in analyzing scientific reasoning in experimental practices.

This paper discusses the American scientist and philosopher Charles S. Peirce's (1839–1914) classification of the sciences from the contemporary perspective of interdisciplinary studies. Three theses are defended: (1) Studies on interdisciplinarity pertain to the intermediate class of Peirce's classification of all science, the sciences of review (retrospective science), ranking below the sciences of discovery (heuretic sciences) and above practical science (the arts). (2) Scientific research methods adopted by interdisciplinary inquiries are cross-categorial. Making them converge to an increasing extent with the sciences of discovery, especially the methodeutic of normative logic, is one of the future challenges for studies on interdisciplinarity. (3) The overall structure of Peirce's classification, were it to be applied in today's situation, would not, in any major respect, be radically different from what it was designed to reflect a hundred years ago, in spite of the virtually exponential creation and production of new domains and the massive increase in investment in research and scientific publication. Accordingly, charges that the sciences of discovery are becoming ever more fragmented are not new.