Networks in contemporary philosophy of science: tracking the history of a theme between metaphor and structure

Valter Alnis Bezerra

Universidade de São Paulo

Abstract: 
In this work, an attempt is made at reconstructing and interpreting a certain part of the history of a theme in contemporary philosophy of science, through a temporal window of approximately 120 years — namely, the theme of networks. The idea of a network received quite a few different formulations in the work of authors such as Duhem, Feigl, Hempel, Neurath, Quine, Thagard and the structuralist metatheory of Sneed, Moulines and others. The purpose of this work is to chart the unfoldment of this theme as it is formulated successively by means of different concepts by these authors. The idea of a network is viewed here as a unifying trait amongst a certain class of philosophical images of science. The idea of a network seems highly appealing as it allows one to express the notion that scientific knowledge is somehow interconnected; also, it seems to suggest tools for philosophical analysis that appear somehow richer and more flexible than other available constructs. It is suggested here that one possible manner of understanding the various concepts of network — taken as different particular formulations of a general theme — is to describe these in terms of a sort of conceptual space with two dimensions. One of them is associated with the degree in which a given concept has a metaphorical character, and possesses heuristic and suggestive power. Another component is related to the degree to which this concept is capable of receiving a precise structural formulation. We aim to show that the various particular concepts of network exhibited different proportions of the “metaphorical” and “structural” components. (This manner of interpreting philosophical concepts is not unrelated to Holton’s thematic model of concepts in science.) Thus, we can identify several distinct stages in the development of the theme of networks in contemporary philosophy of science. (The order of presentation followed in this selective and interpretive exercise in the history of philosophy of science is not strictly chronological; it takes place according to the order of reasons, so to speak.) First, the idea of a network is posited by some authors (Duhem, Neurath) as a metaphor or suggestive image of great heuristic value. Then, in some authors (Feigl, Hempel), it is taken as a notion susceptible in principle to a precise implementation — but, as it turns out, it still has an unavoidably implicit and virtual character. Then the idea of a network is recast as a concept with a quite explicit implementation (Thagard), but still lacking an epistemological rationale — a gap that some other authors have tried to fill (Bonjour and other coherentists). Then, finally, it emerges as a concept with a very precise structural implementation (Moulines-Sneed) and philosophical explanatory power, able to model the structure, the kinematics and the dynamics of scientific knowledge.