Experimentation and concepts in Gilbert’s De magnete

Laura Georgescu

Ghent University

Abstract: 
This paper argues that, on Gilbert’s account, concepts in De magnete (1600) are articulated through experimentation, as a result of the systematic nature of Gilbert’s experimental practice. Experimental practice is integral to his articulation of certain concepts, and is not just an instrument for gathering “facts” or “proofs”. My claim is that what Gilbert takes these concepts (e.g., ‘electricks’ (electrostatic effects) or ‘magnetic coition’ (his version of magnetic attraction)) to refer to depends on the experimental practice (and not that the concepts are derived, or discovered via the experiments). My position here is non-trivial: previous analyses of Gilbert have mapped his concepts in relation to his theoretical commitments (his single-element matter theory and/or his commitment to geomagnetism). When Gilbert’s experimentation has been addressed in the scholarship, genealogical concerns have taken center stage (e.g., Zilsel, 1941, 2003; Henry, 2001). So far, questions about what types of experiments Gilbert appealed to, their epistemic aims, and the types of results obtained have rarely been addressed. Gilbert’s experimentation practices have been interpreted as theory-driven in a strong sense: his experimental program has been taken to be nothing more than an illustration of a preset theoretical agenda (Pumfrey, 2002). There is evidence for this interpretation: a large number of experiments use a spherical loadstone, the “terrella” (“little Earth”), as a model of the Earth itself; Gilbert projected his laboratory results onto the Earth itself; and so on (see Gilbert, 1958, esp. Book 3). However, it is precisely this insistence on opposing the experimental program to the theoretical agenda that obscures Gilbert’s conception of what experimentation does for knowledge production. On my interpretation, the burden of proof of experimentation is shifted away from the classical questions of how theories prove or justify. Instead, my concern is with what experimentation produces in the first place. Here, I argue that conceptual articulation is one such product. One of De magnete’s principal aims was the elimination of past errors and misconceptions about magnetism and magnetic phenomena. Consequently, Gilbert calls for a new domain to be instituted: the new “physiologia” of magnetic matter. In the process, however, a new conceptual apparatus for adequately dealing with both past and new magnetic effects was needed. Gilbert articulated this apparatus experimentally, as a result of the systematic character of his experiments. I argue for this by analyzing Gilbert’s definition of magnetic coition as the mutual action of magnetic bodies, whose strength depends on their relative positions and masses. Because his experiments come in groups, rather than as individual experiments in isolation, the articulation of concepts such as magnetic coition is prompted by the cumulative results of such experiments. The concept of “magnetic coition”—what it describes, what role it plays in the larger theoretical agenda, etc.—is context-sensitive for the group of experiments that brought it about in the first place.