Is science the only or best source of knowledge, justification, and rational belief? If you answer in the affirmative, you are affected by an epistemic disease known as scientism. As a scientist myself, I think that’s a problem.
My friend and collaborator Maarten Boudry has recently mounted yet another defense of scientism, so I feel compelled to articulate yet another offense here. Maarten and I actually co-edited a book presenting a range of opinions on this topic, and now he has written positively about and elaborated upon a recent paper published in Metaphilosophy by Johan Hietanen and collaborators.
Maarten begins by pointing out that “scientism” is both a term of abuse and a badge of honor. I am in the group of people who think it an abuse, while Maarten, Alex Rosenberg, and Don Ross, among others, wear it proudly on their sleeves. The situation is complicated by the fact that a number of purveyors of pseudoscience (creationists, for instance) often hurl the epithet of “scientism!” to perfectly reasonable statements made by scientists and philosophers of science. But this misuse shouldn’t distract us from whether there really is a scientistic attitude out there, and if so, whether it is problematic. After all, there are people who call themselves climate change “skeptics,” but we are not about to abandon that honorable and useful word just because they abuse it.
Much of what follows is a discussion about the meaning of words, but in case someone is tempted to dismiss it as “just semantics,” I will remind them that we literally could not understand each other if we didn’t get our semantics straightened out. So, let’s begin.
In fact, it is precisely an attempt to clear up the semantics that motivated, in part, the paper by Hietanen et al. They propose four related meanings of “scientism,” depending on how narrow or broad a conception of science one wishes to deploy, and how strong or weak the scientistic claim is taken to be. Here is the result (boldface terms are crucial):
(I) Narrow conception of science / weak scientistic claim: The natural sciences are the best sources of knowledge, justification, rational belief, and the like.
(II) Narrow conception of science / strong scientistic claim: The natural sciences are the only sources of knowledge, justification, rational belief, and the like.
(III) Broad conception of science / weak scientistic claim: The sciences are the best sources of knowledge, justification, rational belief, and the like.
(IV) Broad conception of science / strong scientistic claim: The sciences are the only sources of knowledge, justification, rational belief, and the like.
Maarten thinks that the only indefensible position is (II) above, and he quickly adds that nobody really supports it. Then again, he also admits that someone does, in fact, support it: the above mentioned Alex Rosenberg. I think Maarten is forgetting a few other people, for instance physicist and philosophy “skeptic” Lawrence Krauss. At any rate, how many people support what position is an empirical claim that cannot be settled without adequate sociological studies.
Both Maarten and Hietanen et al. conclude that the other three positions, however, are perfectly reasonable. I’m going to argue that none of the above positions about scientism are philosophically sound.
Much of Maarten’s conclusion hinges on his rejection of what he calls the “foundationalist fallacy,” which he claims critics of scientism commit whenever they point out that science itself relies on certain assumptions — like that there is a real, physical world, or lawful regularities, or cause and effect — which are themselves not amenable to scientific investigation. Maarten rejects foundationalism, relying instead on the well-known metaphor of knowledge as a web of interconnected and mutually supported notions, a la W.V.O. Quine.
It is interesting that Maarten is deploying the “you’ve committed a logical fallacy!” move, considering that the two of us co-authored a paper (with Fabio Paglieri) arguing that often alleged informal fallacies are not fallacious at all, but rather a reasonable heuristic or first step to approach a complex problem. In this case, the critics of scientism are simply asking a valid question: are some of the assumptions necessary for a scientific practice themselves amenable to scientific investigation? The fact that if they turn out not to be one can then ask the further question “where does that assumption come from?” is not an objection. Sure, if one keeps pushing foundationalism one risks ending up in an infinite regress (“turtles all the way down”), but that doesn’t mean that one should take a look at the first few turtles at least.
Also, the debate between foundationalism and coherentism (the sort of notion Quine was putting forth) is far from being settled, unlike what Maarten seems to imply. Moreover, and I don’t wish to come across as pedantic (but I remind you: this is a discussion about semantics), Quine’s metaphor of the web refers to belief, not knowledge. And as we are all aware of, setting aside Gettier problems, knowledge is justified true belief. So we’d really need to talk about all four of those crucial philosophical terms: knowledge, justification, truth, and belief. Let me just add that, pace Quine, and with due acknowledgment of the fact that metaphors of course have limitations, webs don’t suspend themselves in midair. They still need some type of anchor or support, although not necessarily a foundation.
At any rate, Maarten dismisses the very idea that science requires background assumptions. He says that those are really “working hypotheses,” that “are being tested as we go along.” Really? How exactly are scientists testing, rather than taking for granted, that there is a real physical world out there? What sort of experiments are being done on the constancy of the laws of physics? What does it mean to “test” causality? These are not working hypotheses, these are bona fide axioms of scientific practice. And that’s okay, because we all know that we have to start somewhere, in science just as in logic and mathematics (see Godel’s incompleteness theorems).
Which brings me to my major beef with supporters of scientism. Look again at the four possibilities above, outlined in the paper by Hietanen et al.: they take for granted certain specific meanings of the words “science,” “knowledge,” “justification,” and “rational belief.” But there is a large and vibrant philosophical literature on all of these, and moreover, that literature contains not a single set of systematic observations or experiments, the distinguishing characteristics of science.
Let’s begin with “science.” This isn’t a natural category, unlike the impression you may get if you listened to Rosenberg and Krauss. It is a time-dependent cultural construct (Helen Longino has written very insightfully about this). Nowadays, “science” is a complex cultural activity that includes granting agencies, university faculty, industry researchers, Ph.D. programs, textbooks, peer review, and so forth. By this definition, Galileo wasn’t doing science, though he was doing something that resembles it and certainly eventually led to it. Was Aristotle doing science? That depends. Are modern historians and philosophers doing science? Not really, though they may sometimes use some of the tools of science, or interact with scientists.
What about “knowledge”? Well, forgive me for the presumption, but I can confidently say that I am knowledgeable about the literature in philosophy of science, as well as about how to make a variety of martinis (my favorite being dirty, with vodka). Am I doing science? Not in any sense I would have recognized when I was a practicing biologist.
Let’s move on to “justification.” That word has a panoply of meanings and applications, many of which have nothing whatsoever to do with science. Am I justified in yelling at my neighbor if he makes what I think amounts to too much noise while I’m trying to write a blog post? In part that depends on empirical information (we can measure the noise level), but in greater part it depends on my values (I don’t think yelling at people is, generally speaking, a good idea), as well as by the rules of the building both my neighbor and I happen to inhabit. These latter criteria are in no way “scientific.”
Finally, “rational belief.” There are two concepts at play here, actually, that of rationality and that of belief. But even if we treat them in unison, we still need to articulate what criteria we are going to use in order to assess the rationality of a belief. Shall we deploy formal logic? An instrumental notion of rationality? A pragmatic one? Do I really need to point out that such a discussion is not going to be settled by systematic observations and experiments?
The point I’m trying to make is that a classic scientistic move — which Maarten deploys with gusto — is to more or less implicitly expand the definition of science, or knowledge, so much that one looks like a fool if one rejects scientism. But one could also reasonably wonder what justifies (!) such expansion, other than an a priori ideological position (after all, scientism does end with an “-ism”), bent on colonizing other fields of inquiry or scholarship, not to mention monopolize public discourse.
Maarten tells us that “plumbing is not a science, but it is continuous with science, because it makes use of similar methods (observation and logical inference) and is connected with scientific knowledge, for example about fluid dynamics.” By that standard, my ability to successfully navigate the New York City subway system is also “continuous” with science. As well as my knack for martini making. You see how silly this whole thing quickly becomes? If a word means everything then it doesn’t mean anything.
Maarten lays down his cards when he claims that “there are no other ways of knowing apart from the sciences (broadly construed, including history and the humanities).” Yet he actually needs to include not just history and the humanities, but also logic and mathematics. And economics, and psychology, and sociology. And political “science.” What about library “science”? If we make our container as broad as the entirety of academic disciplines then by definition there will be no other ways of knowing outside of science. That, however, has now become an entirely trivial claim.
Moreover, what has always baffled me about both scientists and philosophers who wear the scientistic badge with pride is that they don’t seem to be familiar with the actual on-the-ground differences between fields. Have you ever read a scientific paper (say, in physics or biology) next to a paper in philosophy or literary criticism? I have. And I have written a good number of papers in both biology and philosophy. They are nothing like each other. There is very little common ground in style, methods, and subject matters. Except for the fact that human beings are trying to use their best understanding of facts and their best reasoning to write such papers. Are we now going to expand the definition of science to the point that it is that field of inquiry that deals with facts (however understood) and reasoning (however understood)? If we do, there is going to be very little that falls outside of that definition. Which makes the definition itself utterly useless.
Even Maarten, though, apparently has limits, and they are more strict than those bounding Hietanen et al. He rejects their notion of “epistemic opportunism,” according to which science does not follow any methodology, but rather picks up whatever tools happen to “work” (a position very similar to the methodological anarchism famously expounded by one of the philosophers most criticized by scientists: Paul Feyerabend).
Maarten responds, correctly, in my view (italics are mine): “Suppose that tomorrow we discover that reading tea leaves really worked, furnishing reliable predictions about future events. Scientists would definitely pay attention, and perhaps they would eventually be forced to incorporate tea leaves in their toolkit. But of course that would be a radical departure from science as we know it, and would lead to a complete overhaul of our scientific worldview. It would be lame to then say: ‘Scientism is still correct because reading tea leaves is now part of science!'”
Right, but how do we judge, exactly, just how “radical” a departure from science we have to have before it’s no longer science? And is that a scientific question? Ultimately, science is a Wittgensteinian family resemblance concept, and the only way to establish what the term means is by looking at the language game we all play in that respect. As such, I would suggest that most people do not use “science” in a sense that includes plumbing, or subway navigation, or martini making. Indeed, not even logic and mathematics neatly fit under the science umbrella (contra what explicitly stated by Krauss), because these latter two fields do not need any observation or experiment in order to arrive at their conclusions, and because logicians and mathematicians tend to deploy a coherence theory of truth, while scientists lean toward a correspondence theory.
Given all the above, I want to make a modest positive proposal, and suggest a minimalist definition of scientism that we could all get on board with (well, except creationists):
The natural sciences are the major sources of knowledge, justification, rational belief, and the like about the natural world.
Massimo Pigliucci
Massimo Pigliucci is the K.D. Irani Professor of Philosophy at the City College of New York. His books include How to Be a Stoic: Using Ancient Philosophy to Live a Modern Life (Basic Books), A Handbook for New Stoics: How to Thrive in a World Out of Your Control (The Experiment, with Gregory Lopez), and How to Live a Good Life (Vintage, co-edited with Skye Cleary and Dan Kaufman). Check out more from Massimo at massimopigliucci.wordpress.com.
Cosmological measurements of the fine structure constant.
The detection of a particle without and anti particle would indicate a problem with the causal structure of QFT.
I’m glad you wrote this in response to Maarten Boudry’s article. I just read it yesterday, so it’s a well-timed read for me. You raise good points here – but I’m still left disappointed by what I (think I) see in philosophy’s response to scientism itself and to the corresponding defense of science: its theoretical/conceptual arguments tend to ignore the practical facts of how science (or Science, as I think it’s better capture science in its current institutionalized mode) actually gets done, how practicing scientists tend to think about their studies and findings, and how “consumers” of scientific knowledge (… many of whom likely have no idea what went on in the study, what design the actual research took or how it came to have that particular design, or how to think about the limits that are or are not or should be applied in trying to extrapolate that knowledge from the particular to the general). What seems more important to me in considering scientistm isn’t its theoretical weakness but its “meaning” and consequence in the real world. So, for example, take clinical medical doctors: while not bench scientists (in this example anyway), the certainly consider their medical practice to be scientific (and, just as importantly, so does everyone else). And take people who end up with some vague un-diagnosis like MUPS (Medically Unexplained Physical Symptoms), which probably shepherds them over to the psychiatry department. The problem is that Mary feels pain/has headaches/has started to have sudden weakness in her legs and falls down a lot/is having lots of diarrhea or a bloated belly/etc, and that medicine – which, to repeat the primary point here “Science!” – has not identified the concrete causal mechanism for whatever symptoms Mary has. (Hence Mary’s dispatch over to the psychiatry department.) The doctor’s basic premise is a scientistic one: having done all the “right” tests and followed the “best practices” diagnostic algorithm and so on, no “cause” showed itself, and so the doctor invested in “Science!” concludes that there is no cause. But this is “Science!” and not science: it might be that some other test would show a cause, or that the cause cannot be identified because “Science!” has not yet summoned the epistemic resources to identify and name it, or that the doctor is not especially competent in the interpretation of test results, or X or Y or Z or a hundred other reasons. But Mary *knows something about herself and her living in the world, and one of the things she knows is that something is going wrong in her body-self. **This is exactly the kind of thing that gets lost in the scientistic vision of the world as lived and practiced; the doctor isn’t thinking about foundationalist fallacies and so on, he’s thinking about the algorithmic vision of the world constrained “just so” such that whatever practices the scientist does are (1) scientific and (2) limitless in their explanatory power. Approaches to science from STS or the social studies of science do a better job of getting at these sorts of problems (but then they don’t usually name scientism as the object of their critique, and that makes it hard to make those critiques speak with critiques from philosophy).
I’d be really interested to hear (read) how you think those sorts of responses to scientism (i.e., focused on practice rather than on theory; maybe mimicking the kinds of analysis that come from STS) could (or couldn’t) be compatible and integratable with a philosophy-based approached scientism and its defenders and its critics. In any case, thanks for this good read.
I’ve argued these points before. However I would be curious to how you respond to:
1. The difference between naturalism and scientism as they seem to get misused as being the same.
2. Your assuming rational belief about the natural world means truth or so it seems. How do you respond to scientific anti-realism and the problem of emphircal versus real truth?
3. Surprised there was no mention of Hume or Nelson Goodman’s problem or new problem of induction as its a big problem. The difference between deductive and inductive statements is a important difference most people don’t grasp or understand.
Good article through.
One thing I often find omitted in these discussions, which I would like you, Massimo, to expand on, is how many things in our lives through which we both learn and generate new knowledge are rather art than science. In fact, one of the most famous books on electronics, a subject that heavily relies on physics, is called ‘The Art of Electronics’ (by Horowitz and Hill) for good reasons. Even the very techniques we use when performing experiments and being inventive with making new engineering designs or scientific theories can rather be described as art than some organized and coherent principles stemming from a scientific theory.
Regarding assumptions that science makes, I think we need to consider which assumptions are strictly necessary for doing science.
‘How exactly are scientists testing, rather than taking for granted, that there is a real physical world out there?’
Even this assumption is not necessary. If an objective external reality exists then our science will describe it to some extent; if it does not exist, then the concept of science describing an external world is vacuous; but they can still do science knowing that scientific ideas would only apply to the external world to the extent that it exists.
Here are a few other things that are commonly thought of as being assumptions:
causality,
the existence of physical law,
and the constancy of physical law.
These need not be assumptions. Causality in the theory of relativity was never an assumption. The idea developed from investigating the non invariance of Maxwell equations under Galilean transformations and the action-at-a-distance nature of Newtonian gravity.
We need not assume the existence of physical law, nor its constancy. What would happen if we did not make this assumption? Nothing; we would still be doing science trying to understand the principles of nature to the extent that they exist.
People often strive to do doing things even if they don’t have a guarantee that it is even possible.
To the extent that we understand the words ‘quantum’ and ‘gravity’ in current physics, there is no reason to assume that a quantum theory of gravity exists. The does not prevent our looking for one.
I agree with all the claims made here yet I still find myself agreeing with meaning IV of scientism. I think the word “best” is doing a lot of heavy lifting here. To me science isn’t the best source in that it is always preferable to other areas for every question, rather, it’s the best on net in that it produces reliable conclusions to many important questions.
I suppose to deny this claim you must accept that either (1) there is a better source, (2) that all sources are equally good, or (3) that you can’t rank sources of knowledge, justification, rational belief, and the like and I just find myself less inclined to accept any of these statements than to accept scientism.
I struggle to understand which one of these three the author would favor. I guess probably 3? It seems he would say that science is good for some questions and poor for others. And I do think that questions of whether something is “best” rather than “best for” are often totally incoherent.
About the “that often alleged informal fallacies are not fallacious at all, but rather a reasonable heuristic or first step to approach a complex problem” part, I fully agree with it, not just that, most of the time, those alleged informal fallacies, but it can also be applied to formal fallacies and logical fallacies as a whole, are not fallacious at all, they are actually relative in most of the cases, they are often a reasonsable philosophical or humanist first step or first steps to approach complex problems or even social problems, and, eventually, leading into a more complex discussion and argument about them, and also, alleged informal fallacies can also bring positive things and have positive things in social, political and cultural issues as well, and as well as work as a humanist force sometimes, without mention that an informal fallacy (but it also works for all fallacies at all) are only fallacious or non fallacious according to its context, that’s the why they are relative, one can say an informal fallacy, and it actually be fully valid if it has data and arguments in favor of it, they often are valid for social and cultural issues, and even for deal with complex issues, like the ones of “not all” and “all problems are relative (Fallacy of Relative Privation)”, they aren’t fallacious at all if you’re using them in a context that would bring good results or avoid bad things to happen, like “not all germans are nazis” or “your problems nowadays are not that bad as your problems from 6 months ago”, they are not fallacious at all if the context they are insidered are not fallacious, the first one about someone who believes all Germans are nazis and the second one someone who wants to give up in life, I can even give an example of “not all humans are evil” or “not all muslims are terrorists” in case of arguing with xenophobes and racists. And even the “No true Scotsman” can not be fallacious at all if it is insidered in a context about nationalism and patriotism, like fighting in a war of national liberation or even about being proud of your country in something. So, you need to analyse their context before you take any conclusions about that, all fallacies are relative, at least they often are relative. So yes, the use of alleged informal fallacies is valid for philosophical and humanist purposes and even for initial approaches for complex problems or even avoid bad things to happen.
Thanks for the informative read.I am curious to know whether the domain of the natural world fully satisfies the requirements of the domain of science,as we know.