A Program for historically informed Philosophy of Science

This point illustrates a striking advantage of the... position: it characterizes the general presuppositions of the methods by which perceptual experience is scientifically studied. These methods presuppose independent individuations of sensations and of the physical correlates of these sensations. This is in sharp contrast with physicalist analyses of color, which rest upon particular theses of perceptual science, and are thus contingent on the outcome of scientific disputes (Sect. 4.2). This advantage becomes especially clear when we consider secondary qualities which are more poorly understood than color. Whereas a physicalist analysis of olfactory qualities must wait for a more mature theory of the physical correlates of smell, the...analysis can already provide a general answer to the epistemic question of the the relationship between smells as experienced and smells in the world. Furthermore, it can explain the practices of olfactory research, which we observe struggling to characterize (i) the structure of our experience of smell; (ii) the structure of the physical correlates of smell; and (iii) the process of calibration which explains the shifting correspondence between these two across different contexts (Sect. 5.2)--Alistair M.C. Isaac (2014) "Structural Realism for Secondary Qualities" Erkentnisse (emphases added; I also removed his use of "structural realism;" see the long footnote* below*)

Last week, in Edinburgh, I caught up with Alistair Isaac (who did an obituary of Suppes here a few years ago). We got talking about research, and Alistair described me the work he was doing on the history of the methods used in science of color perception. I think Alistair has latched on to an important justification for doing HPSy work in the philosophy of science that can contribute to scientific self-understanding and even the kind of reflection that may  improve scientific practice. (Of course, one can also be interested in philosophy of science and/or the history of science for other, intrinsic or philosophical reasons.) What follows is Alistair's insight, but I was primed to see it due to very fascinating work by two PhD students: one is Laura Georgescu, who works on the history of magnetism (and meta-philosophical reflections on the philosophy of scientific practice); Laura has articulated claims that in a different idiom that get at similar insights (but her metaphysics is very different from Alistair's). The other, Femke Truijens, is working on evidential practices in psychotherapy research.

Before I get to explaining the insight I'll attribute to Alistair, I need to do a bit of stage-setting (and I'll ignore his first-order claim defending a form of structural realism in secondary qualities). First, one reason to study the history of a particular scientific practice is that the underlying methods used in many sciences often remain remarkably robust (despite technological, computing, and mathematical innovations) through long periods of time (sometimes this is even necessary to ensure backward compatibility of evidence). And when methods/practices are standardized, debates over the nature and justification over these methods disappear (or are suppressed). Often those debates occur only at the origin of the practice. The original debates are useful to study because they often include an analysis of what can be expected from the method and the nature of evidence it can or cannot provide. Often technically sophisticated methods have to be explained by an innovator or her disciple to a discipline that does not fully understand these methods. [Sometimes the methods are adopted from other fields by people trained in those field who then moved to other areas (a migration from physics, say, to economics or from statistics to biology, etc.)] From a philosopher's perspective the original debates often tend to be more interesting and sophisticated than text-book explanations at a later date. {This paragraph uses 'often' often in order to signal that there are lots of exceptions to these generalizations!}

To be sure, one need not do history to engage in immanent critique. There are good examples of philosophers of science that have scrutinized existing scientific practice in some contested (IQ research, gender-biased knowledge, etc.) area and have offered compelling evidence that exposed the methodological limitations of some 'scientific' practice(s).

Second, Alistair's analysis is rooted in considerations of measurement. For Alistair, "a measurement procedure establishes a structural correspondence between two spaces," that is "the measured space" (for example, in thermometers, "the space of possible mean molecular motions" ["the only property of this space which is represented in the measuring space, however, is the relational property that mean molecular motions can be linearly ordered"] and "the measuring space" the space of possible real numbers). This is not the place to discuss how such a correspondence is established, but in it, calibration of the instrument, categorization of scales, background physical theory,** and differential response to physical processes all play an important role. All of these involve interesting conceptual and philosophical issues. What matters here is that "the three basic components of measurement (measuring space, measured space, calibration process) shape the scientific investigation of the perception of secondary qualities." And this last point has an important implication.

Okay, so much for stage-setting. In his paper, Alistair carefully shows that the components of measurement and the structural correspondence they give rise to also provides a kind of constraint on what can and cannot be known through that practice. While, in our conversation, Alistair and I quickly started talking about how there is a kind transcendental argument lurking here in how a measurement practice is a condition of possibility of certain kind of structural knowledge, it is worth noting that there is nothing especially profound here: while measures can (and do) get decoupled from the theories, practices, and even technologies that give rise to them they are limited purpose instruments. If I stand on my scale, one can learn a lot about me but not (I think) the color of my eyes.

Let me wrap up, with the larger point. A lot of bread and butter practice in philosophy of science takes the scientific 'output' (to be found in a textbook or a journal article) as a given (as it were 'input') for further philosophical analysis of the science, or the application of the science as evidence for a philosophical theory, or development of better scientific or philosophical theory. (This is, in part, a remnant from a lack of epistemic interest in the context of discovery.) What Alistair shows is that the methodological practices within the science that produce this output has implications for the kind of ontological claims one can or cannot make. (This previous point tends to be obscured in discussions about explanation or evidential strength.)

There is no special historical or philosophical skill that allows philosophers to reflect on the ontological and (epistemically) modal significance of  methodological practices. The tools for this reflection come from (say) measurement theory and an understanding of the scientific concepts. But unless there is a Kuhnian crisis, or a field in which competing methodological schools flourish, (normal) scientists have little reason to spend a lot of time reflecting on what tools, which have proven their worth in scientific practice, can and cannot show. (There is also little incentive for them to do so.) HPS-style philosophers can provide such a reflection. One potential pay-off of an inquiry into what can be known by particular methods is that philosophers can help scientists and the public they serve better understand the reach of their claims and, be motivated, to explore alternative methods with a different kind of reach. In fact, I bet that quite a few philosophers of science, including HPSy types, already do what I am gesturing at in this post (and I would be curious to learn about methodological statements along these lines).

 

 

*Alistair defends a version of structural realism (but not one of the types familiar, I think, from debates in physics and its history). He sums up his own view as follows:

I have argued for structural realism, the claim that the structure of our possible experiences corresponds to the structure of possible ways the world can be. Since this structural correspondence between experience and the world is calibrated differently across different contexts, however, we cannot directly identify particular experiences with particular properties in the world. We cannot identify red with a particular surface reflectance property, warmth with a particular temperature range, pungency with a particular molecular shape, etc. This is why this realism is structural: it is not committed to a metaphysical reduction of the properties of the world as experienced to the properties of the world as it is.
Nevertheless, structural realism is still realism since the preservation of relations between properties across the correspondence between experience and the world ensures that property attributions are in general veridical (semantic realism) and demonstrate knowledge (epistemic realism), so long as they are assessed against a contextually established calibration baseline. The dissociation of ontic from epistemic and semantic realism is motivated by an analogy with measurement. Quantities in the world are not themselves numbers, yet we can use numbers to represent them once we establish a structural correspondence between the real line and possible values of a quantity through an act of calibration. The three basic components of measurement (measuring space, measured space, calibration process) shape the scientific investigation of the perception of secondary qualities. To demonstrate this, we’ve surveyed the examples of warmth, color, pitch, and odor. Consequently, structural realism is the epistemological analysis of the status of secondary qualities most strongly supported by scientific practice." [Emphasis in original]

**I would emphasize that Alistair is working with theory-mediated measurement. Not sure he would.

Comments

[Alistair Isaac]

Eric, thank you very much for the interest in my work, and for this kind discussion of it.

Just a couple points of elaboration to further the conversation:

I take the basic idea of interest to be this: (i) the (experimental / statistical / etc.) methods of a science may presuppose substantive philosophical positions; (ii) insofar as these methods are robust across changes in high-level theory, the presuppositions that support them suggest more robust philosophical implications of the science at issue than the prima facie claims of high-level theory; and (iii) these presuppositions may only be uncovered through historical work, and are often only explicitly discussed when the method is first introduced.

I initially became interested in this idea not so much as a suggestion for how philosophy of science should proceed, but for how science should inform philosophy in general, especially areas of philosophy other than philosophy of science (eg philosophy of mind or metaphysics). This is not to say there aren’t implications for philosophy of science as well, but I think Eric is right that there is already a burgeoning HPS literature that focuses on methods and their implications. My own ideas here have been heavily influenced by several figures in HPS, especial Patrick Suppes, Hasok Chang, and George Smith. The last of which has worked closely on the question of “backward compatibility of evidence,” see for instance his “Closing the Loop” in Biener and Schliesser (eds) Newton and Empiricism.

In the paper Eric discusses, there is really a two-fold appeal to measurement. On the one hand, I take the psychophysical methods for measuring perceptual qualities to be the appropriate target for extracting implications for the philosophy of perception (rather than current theories in perception science). On the other, I draw an analogy between measurement and its key features (basically as articulated in the “representational theory of measurement”) and the key features of perception. This analogy is meant to help articulate the first-order claim that “epistemic structural realism” is the appropriate position within the perceptual realism debate implied by psychophysical methods. Eric is exactly right, however, to dissociate this first-order conclusion from the second-order claim that it is scientific methods we should be examining as philosophers trying to extract substantive conclusions from science, rather than high-level theory. I would love nothing more than to find an interlocutor who agrees on the second-order point, but disagrees about its first-order implications.

[While I also agree with Eric that the analogy between “structural realism for secondary qualities” as I articulate it and structural realism about, say, physics, is loose at best, I do think these views share a common historical origin in 19th century post-Kantian psychology. I’m currently engaged in an ongoing project to trace these origins, especially a strand of influence from Joh. Müller, through Helmholtz, to Hertz.]

Finally, returning to the question of whether there is an interesting methodological project here for HPS — I think one place to see it is in the recent surge of interest in scientific measurement from an HPS perspective. Measurement is the practice by which we connect theory to world and (even when theory-mediated) may be the source of evidence robust across theory change. The intuition that the methods of measurement and data analysis are a potential source for deep insights on scientific progress and the epistemology of evidence, seem to me implicit endorsements of the view. Like Eric, I would love to hear of any related or explicit methodological pronouncements to this effect.

[David Duffy]

"...methods of measurement and data analysis are a potential source for deep insights on scientific progress and the epistemology of evidence..."

Perhaps not exactly what you were thinking of, but the kind of measurement model (SEM) I am most used to:

https://www.researchgate.net/profile/Rogier_Kievit2/publication/225188983_Mind_the_Gap_A_Psychometric_Approach_to_the_Reduction_Problem/links/09e41500d1a48e1497000000.pdf