Are Strong Scientific Realists Tempted by Cognitive Illusions?
When it comes to deep, postulatory theories I am a nonrealist. I deny that we have sufficient evidence and argument to conclude with confidence that even our most mature theories are true, or very nearly, that at best minor tweaking will be necessary.[57] But I am not a die-hard anti-realist.
For all I know, some theory might be true. However, we can never be sufficiently sure to say that we know it is true. Moreover, whether or not it is true simply does not matter for human action, either to further research or to applications. Or so I claim. Thus I remain agnostic, although I place the burden on the side of the strong realists to make their case,1In the future I shall assume the qualifying phrase without repeating it. including the case for why it matters.[58] Theirs is, after all, the stronger position and thus the one that most needs defending. So, if forced to say so, I do lean toward the anti-realism side of nonrealism. Prominent versions of allegedly strong realism are not even genuine realisms comparable to good-old-fashioned scientific realism.
As I am using the term, strong realism divides into two somewhat distinguishable components, strong epistemological or “epistemic” realism and strong ontological or “ontic” realism. A strong epistemic realist holds that we can, in some cases, know that a theoretical claim is true. Meanwhile, a strong ontic realist claims that we can know that the postulated entities and/or processes exist and that we understand what they are. That is, the entities are intelligible to the relevant experts, as a scientific replacement for the metaphysics of old, not merely a vague or ambiguous reference to whatever entities or processes may satisfy abstract equations (shades of Frank Ramsey). There are also cases in which we now have good observational or experimental access and control of entities once postulated.
In these cases, the theoretical realism is now justified, but it is shallow because it is no longer deeply theoretical. But even in some of these cases ontic realism fails. Do we really even understand what electrons are, let alone quarks? Without entering into further discussion of it, I note that structural realism gives up the “metaphysical” pretensions that strong realists attempt to keep.There are several good reasons, in my opinion, not to be a strong realist. In this section I focus on one of those reasons, or lines of reasoning, but only briefly. Details can be found in a companion paper (Nickles 2016) where I present a dozen or so cognitive illusions that appear to make strong realism more plausible than it is. We only seem to see scientific truth circling around us everywhere.
Many of the illusions derive from a failure to be sensitive to our place in a longer history, including future history. What I call the flat future or end-of-history illusion leads us to believe that the future will be relatively flat, uneventful, in relevant respects (cf. Quoidbach et al. 2013). This is understandable because we are concretely aware of past dynamic changes in our personal lives and in world historical development; but we do not have, and cannot have, such knowledge of the future. Our tendency is to project our present views and preferences forward, as if the transformative changes are now over. Applied to science, the temptation is to see the future of a currently successful field as similarly flat—as if the dynamical, innovative history of the field is now ended. On this view, future science will likely produce more applications and connections to other work, but the science a thousand years in the future will be basically the same as today. Nothing transformative will occur either by rapid revolution or by slow evolution. In short, such a field is over—sterile—as a research frontier.
The flat future is an illusion owing to our particular historical perspective.[59] It helps to remain nonrealist here by remembering that just about every attempt to predict long-term futures has been a ludicrous failure, including scientific and technological futures and even when ‘long term' is only 100 years.
Ditto for attempts to identify an “end of history.” Think of Hegel or Marx or Fukuyama.The maturity illusion has a similar basis. Why should we think that today’s best science is true when past scientists believed the same of their science—which we reject today as badly wrong? The answer some strong realists give is that today’s science is now mature, whereas theirs was not. After all, we now know of flaws in their theories, their instrumentation, their experimental design, their goals and standards, etc. But wait! What is to keep our distant successors from saying the same about us? Just because today’s most successful theoretical claims seem practically flawless to us does not mean that they really are. Realists who make the maturity move are violating what Mary Hesse once called “the principle of no [historical] privilege” (Hesse 1976, 264). They are treating maturity as an absolute, ahistorical threshold that we have crossed. Any claims that people make about maturity are historically relative and should be historically indexed, for it is only maturity as they characterize it at their particular historical moment.
We are like fish in water in that key parts of our intellectual environment are invisible to us. In some cases, they are probably not yet articulable by us. Therefore, we are unable fully to locate ourselves in historical development, obviously in future development but also in relation to the past—in the sense that future historians and culture commentators (who are also fallible, of course) will make insightful observations about us in relation to our past that we have failed to see— and that perhaps will only become articulable in a still undeveloped future. In that sense we cannot escape from history, even though we can identify progress in overcoming previous historical blindness. Or, to put the point more provocatively, we can overcome history incompletely and only to the extent that we become strong historicists, able to take into account deep and subtle historical perspectives.
What I call “the Copernican illusion” is the idea that, via an extended “Copernican” program, we can gradually identify every human perspectival element and eliminate it, so that, finally, our science will be completely objective, without an ounce of anthropomorphism (Nickles 2017). Our unavoidable historical perspectives already doom such a program. We are not able to stand outside history like gods. The same conclusion follows from the models-all-the-way-down point stressed by Paul Teller (e.g., 2001) and others. Step-by-step removal of approximations and simplifications until the perfect representation of the target system is realized is an increasingly unlikely scenario. The claim that our models will, quite generally in mature science, become as realist as you please is a very strong claim that has not received the degree of justification that it requires.
In my view strong realism reduces historical sensitivity in both directions— future as well as past. Most of the discussion has concerned the past, but insensitivity to the future leads to a tame conception of scientific frontiers, one that tends to replace genuine uncertainty (where even the probabilities and utilities are unknown and often undefined) with mere risks (as ‘uncertainty’ and ‘risk’ are defined in standard rational decision theory).
I hope that these examples convey the flavor of the many cognitive illusions presented in Nickles (2017).
2