Log in / Register
Home arrow History arrow Integrated History and Philosophy of Science: Problems, Perspectives, and Case Studies

Models, Theory, and Evidence: The View from the Life Sciences

Well-entrenched debates in the philosophy of science—concerning the relation between theory and observation, data and hypothesis, confirmation and evidence— often take on a new level of complexity when applied to the life sciences. Methods and approaches that have been found to be suitable for the physical sciences— though themselves by no means simple—often resist straightforward application to the life sciences. As Nicolas Rasmussen puts it: ‘This situation arises largely from the nonspecificity, incompleteness, and over simplicity of physical and chemical theory in its application to biological complexity: either theory is too general to be of real use, or when worked out in detail for a given problem, too uncertain and itself subject to revision.’ (Rasmussen 1993, 231) At the level of experimentation, improvements in the available instruments and techniques—such as the shift from light to electron microscopy—not only open up novel realms of inquiry, but also raise new questions about the calibration of instruments, commensurability of empirical findings, and concordance (or discordance) of evidence. As biologists began to study the molecular basis of life, the line between what Ian Hacking (1983) has characterized as the contrast between ‘representing’ and ‘intervening’ by necessity had to be frequently crossed. As Evelyn Fox Keller puts it: ‘Biologists could see more, but they could no longer watch’ (2002, 214).

In spite of these challenges, a growing body of philosophical literature explores how models, theories, and evidence come together in experimentation in the life sciences.[1] Such studies have often taken the form of detailed case studies ‘of scattered and usually quite recent episodes of experimentation’, sometimes at the expense of ‘diachronic studies that would reveal how [various general] guiding principles came to be applied, how the principles may have mutated in character over time, or how the principles may have changed in their mix or hierarchy relative to one another amongst the set deployed by experimenters in a field’ (Rasmussen 1997, 12). While there is certainly a risk that merely assembling more and more case studies might result in historians and philosophers of science no longer being able to see the forest for the trees, nonetheless a well-chosen case study, in our opinion, may not only provide diachronic depth, but may also cast a spotlight on philosophical problems and challenges. Indeed, the case we wish to discuss in the next section—of models of the cell membrane—aims to achieve just that, and it does so by looking in depth at how the interplay of models, theoretical hypotheses, and empirical findings (and their interpretations) has developed over time, sometimes in response to advances in experimental techniques and instrumentation, sometimes by fitting evidence to a preferred theoretical hypothesis concerning the structure of cell membranes.

In what follows, special emphasis will be placed on how various experimental findings acquired evidentiary significance regarding the various theoretical hypotheses that were proposed for the structure of the cell membrane. At an abstract level, it will be useful to distinguish, following Jacob Stegenga (2013), between signs of success and conditions of success in relation to evidence in biology. Whereas the former refers to formal features (e.g. in terms of probabilistic support) of statements of evidence once an empirical finding has been recognized as such (rather than, say, dismissed as background noise), the latter refers to features of methods (including prior to the consideration of any evidence generated by it) and of the evidence itself—e.g. recurring patterns, concordance from multiple streams of evidence, plausibility and so forth. Stegenga illustrates this point with an analogy: when we try to assess whether a particular wine is good, we might rely on the fact that it has been awarded 90 out of 100 points by an expert wine critic or, alternatively, we might appeal to features of the wine itself (its bouqet and taste, say) or to specific aspects of its method of production (e.g., where it was grown).[2] The first type of approach relies on identifying purely formal signs of success, whereas assessments of the second type engage with substantive conditions of success—considerations of the sort a wine critic might herself rely on in assigning a numerical score. Any approach—whether in oenology or the philosophy of science—that places signs of success at the heart of its project may, in this sense, be considered necessarily post hoc, since ‘one already must possess and have evaluated the evidence with the conditions of success in order to do business in the signs of success tradition’ (Stegenga 2013, 982).

Taking as one of his case studies Avery et al.’s successful identification, in the 1940s, of nucleic acids as the material basis of heredity, Stegenga argues—plausibly, in our judgment—that ‘characterization of evidence in terms of the conditions of success has the virtue of accurately describing those aspects of evidence which appear to matter to scientists’ (Stegenga 2013, 1002), not least the quality and relevance of the methods employed, and the believability, patterns, and concordance of the empirical evidence thus generated (ibid., 997). Shifting the emphasis in philosophical discussions of scientific evidence from its formal features to its diverse conditions of success has the further advantage of being able to account for the fact that, when dealing with new methods and technologies of producing data, ‘rather than simply assessing features of the method, scientists also assess features of the evidence itself’ (Stegenga 2013, 988).[3] Yet, as Jane Maienschein has argued, debates in science often concern ‘what is to count as evidence or how much certain evidence is to count for or against a given argument’ (Maienschein 2000: 122), and so we should be prepared to find that judgments concerning the evidentiary significance of empirical results will often themselves be driven by (potentially diverging) epistemological strategies.

  • [1] See e.g. (Creath and Maienschein 2000), (Rheinberger 1997), and refs. in fn. 1.; for a review ofthe philosophical literature on scientific models, see (Gelfert 2016).
  • [2] See (Stegenga 2013, 982).
  • [3] On this point, see also (Bechtel 2000).
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
Business & Finance
Computer Science
Language & Literature
Political science