This is going to be my last report from the 2014 meetings of the Eastern Division of the American Philosophical Association. Obviously, it has been a rather idiosyncratic set of choices, reflecting my own interests, not necessarily the (much) broader scope of the meeting. Now if I could only convince some of my other colleagues to publish similar commentaries — including here at Scientia Salon — covering their specific interests in science and philosophy…
At any rate, this session was chaired by Otávio Bueno (University of Miami), and the speakers were Bueno himself and Jody Azzouni (Tufts University). Indeed, this was a strange, but welcome, session, with only one paper, entitled “Reasons to Think That Atoms and Subatomic Particles Are Real vs When to Suspend Judgment about the Reality of Atoms and Subatomic Particles,” co-authored and kind of co-presented by a scientific realist (Azzouni) and a scientific anti-realist (Bueno). (The presentation was mostly by Bueno, occasionally humorously interrupted, or substantively supplemented, by commentary from Azzouni. It really was fun!)
So, consider a new scientific theory coming along and positing the existence of some unobservable entities. The standard realist argument is that if those objects do significant work within the theory, and the theory has compelling empirical evidence in its favor, then one should think of said objects as actually existing.
Bueno then introduced a distinction between thin and thick epistemic access: the first one involves only theoretical reasons for thinking of an entity as real; the second one requires some type of detection device or methodology to gather at the least indirect empirical evidence for the existence of the entity.
Of course, perceptual experiences are one kind, but not the only kind, of source of thick epistemic access. Think of transmission electron microscopes, for instance. What convinces us that an object — say a newly discovered sub-cellular organelle — is a real biological structure and not an artifact of preparation is that we can replicate the observation under a variety of conditions and with a variety of instruments, as in fact happened with the discovery of, for instance, ribosomes .
This is interesting, and I get the value of separating thick and thin epistemic access. However, it also seems to me that Bueno has significantly extended the categories of observables compared to the standard anti-realist position, which would confine them pretty much to macroscopic objects accessible by standard human senses. And in fact, Bueno did acknowledge that adopting a van Fraassen’s type of strict empiricism  about unobservables makes it difficult to make sense of actual scientific practice (e.g., in the 21st century we really should consider viruses to be observables, even though of course they cannot be seen with the naked eye).
Again, this strikes me as fundamentally right, but it is definitely moving anti-realism much closer to realism; then again, on the other side of the divide, Ladyman and Ross-style structural realism  has been moving realism closer to antirealism, so there.
Given the above, does the empiricist antirealist still have a line to draw between observables and unobservables? Bueno thinks so, and Azzouni, predictably, disagrees.
The example proffered by Bueno is that of the scanning tunneling microscope : it generates a constant electrical current which is then used to scan a given object up and down. This doesn’t generate an image per se, only information about the object based on how its surface interferes with the electrical current. This information is then processed by computers to generate an image for human consumption. People these days claim that with this instrument we can “see” atoms, but this — according to the anti-realist — is definitely not the case, because the microscope does not actually provide thick epistemic access to the objects of interest. This is evident from the fact, for instance, that the raw data can be represented in a variety of ways, generating very different images. (The “image” accompanying this article is of silicon atoms on the surface of a crystal of silicon carbide, obtained by a scanning tunneling microscope.)
(Notice that the anti-realist here is obviously not saying that atoms don’t exist, only that an empiricist should suspend judgment about the ontology of the entity under study, because the microscope, as sophisticated as it is, does not actually provide anything like images, so that we cannot properly talk about “observables.” Also notice that the anti-realist is perfectly aware of the fact that the distinction between observables and unobservables pertains to human epistemic access to the world, and it doesn’t cut the world at its joints, so to speak. But both realism and anti-realism are epistemic positions in philosophy of science, not metaphysical ones.)
A rather colorful way to think about the situation is that some instruments generate “public hallucinations,” analogous to rainbows: we can all see rainbows, and they have objective existence in the sense that we can describe them and even measure some of their characteristics, but that doesn’t mean there is an actual object out there that corresponds to the visual concept of “rainbow.”
Another interesting case concerns macroscopic objects, like the quest for detecting extrasolar planets. The standard method involves the application of a combination of Newtonian physics and relativistic corrections to infer the existence of an extrasolar planet from the wobbling of a given star. The technique has undergone very fast and very impressive refinements over the past few years, leading to the discovery of more and more, smaller and smaller extrasolar planets . The thick epistemic access here is to the behavior of the star, not the characteristics of the planet (about which we have instead thin epistemic access).
The basic point is that thinking in terms of thick vs thin epistemic access gives the philosopher a tool to wade through cases where there is ontological disagreement stemming from the positing of unobservables in scientific theories, until (and if) they are eventually resolved empirically (e.g., ether: doesn’t exist; genes: do exist and they are made of nucleic acids). This leads to a more reasonable form of empiricism — anti-realism, which does justice to actual scientific practice while still retaining caution when it comes to ontological claims (van Fraassen’s famous goal of avoiding “inflationary” metaphysics in philosophy of science). I am certainly looking forward to seeing the published version of Bueno and Azzouni’s paper!
Massimo Pigliucci is a biologist and philosopher at the City University of New York. His main interests are in the philosophy of science and pseudoscience. He is the editor-in-chief of Scientia Salon, and his latest book (co-edited with Maarten Boudry) is Philosophy of Pseudoscience: Reconsidering the Demarcation Problem (Chicago Press).
 Ribosome, Wiki entry.
 Constructive Empiricism, by B. Monton and C. Mohler, Stanford Encyclopedia of Philosophy, 2012.
 Structural Realism, by J. Ladyman, 2014.