Moti Mizrahi, in his 2019 “The Scientism Debate: A Battle for the Soul of Philosophy?” argues that the perceived threat of scientism to philosophy hinges upon a confluence of two distinct, but overlapping, phenomena. The first of these phenomena is that (some) philosophers find scientism institutionally threatening due to the ostensibly corrosive effect that the natural and social sciences have upon the future of philosophy in universities … [please read below the rest of the article].
Wittingslow, Ryan Mitchell. 2020. “Some Thoughts on the Relationship Between Scientism and Empirical Methods in Philosophy.” Social Epistemology Review and Reply Collective 9 (5): 64-67. https://wp.me/p1Bfg0-52Z.
The PDF of the article gives specific page numbers.
- Bishop, Robert C. 2019. “Scientism or Interdisciplinarity?” Social Epistemology Review and Reply Collective 8 (12): 46-49.
- Mizrahi, Moti. 2019. “The Scientism Debate: A Battle for the Soul of Philosophy?” Social Epistemology Review and Reply Collective 8 (9): 1-13.
- Mizrahi, Moti. 2017. “What’s So Bad About Scientism?” Social Epistemology 31 (4): 351-367.
Moti Mizrahi, in his 2019 “The Scientism Debate: A Battle for the Soul of Philosophy?” argues that the perceived threat of scientism to philosophy hinges upon a confluence of two distinct, but overlapping, phenomena. The first of these phenomena is that (some) philosophers find scientism institutionally threatening due to the ostensibly corrosive effect that the natural and social sciences have upon the future of philosophy in universities. The second is that (some) philosophers find scientism existentially threatening because introducing experimental methods into philosophy threatens the status of philosophy as an inherently a priori discipline (Mizrahi 2019, 2). It is the second of these hypotheses that I will be discussing here.
“Scientism,” Mizrahi writes, “is perceived as a threat to the sort of research that philosophers typically do because it advocates the use of empirical methods of observation, experimentation, and the like” (1). Mizrahi claims that philosophers suspicious of scientism are suspicious because empirical methods pose an invidious influence on the foundation of philosophy as an intellectual practice. This thesis is, however, simplistic. As I will outline in this comment, there is nothing that necessarily binds empirical methods and scientism in the way that Mizrahi supposes; no guarantee that “opponents of scientism want to resist the application of scientific ideas, methods, and practices in new domains” (10). Instead, I argue, it is entirely coherent to possess a principled suspicion of scientism whilst still being generous towards the use of empirical methods in philosophy.
In pursuit of that claim, I will first sketch one possible objection to scientism as Mizrahi defines it. This sketch is not intended to be authoritative (there is a great deal to be said about Mizrahi’s conception of scientism, and precious little time in which to explore it), but I hope this will suffice to make clear at least one non-trivial objection to scientism. Second, I will explain how this possible objection to scientism is entirely compatible with the use of empirical methods in philosophy.
Strong and Weak Scientism
In a 2017 paper, Mizrahi defines two kinds of scientism: ‘strong’ and ‘weak’. The former, he writes, is governed by the claim that “Of all the knowledge we have, scientific knowledge is the only ‘real knowledge’”. This position, as Mizrahi observes, has a number of problems. Most significantly, the unbridled, bristling positivism of strong scientism prima facie precludes the possibility of having meaningful knowledge in any non-scientific domain. Weak scientism, meanwhile, is more ecumenical, while still preserving the rarified status of scientific knowledge: “Of all the knowledge we have, scientific knowledge is the best knowledge” (Mizrahi 2017, 353-354). It is this position that Mizrahi endorses. But what exactly is intended by the claim that scientific knowledge is the ‘best’ knowledge? By this, Mizrahi means that scientific knowledge is both quantitatively and qualitatively superior to non-scientific knowledge. Both of these assumptions are problematic.
According to Mizrahi, scientific knowledge is quantitatively superior because a) more scientific peer-reviewed papers are written than non-scientific peer-reviewed papers and b) scientific papers have, on average, a higher research impact than papers from other disciplines (358). The problem with this approach is that it uncritically equates the success of an enterprise with the popularity of that enterprise. Observing, as Mizrahi does, that research in the humanities only comprises 1-3% of global research output tells us nothing about why scientific research is better, unless we also accompany that figure with a clear description of the institutional, cultural, and economic structures in place that incentivise certain kinds of research over others, to speak nothing of how that research is structured, disseminated, and utilised by members of the academic community and the world at large. It also tells us nothing about the systemic problems that may or may not feature within given scientific research communities, such as the replication crises metastasizing in certain corners of psychology and medicine. Without these crucial details in place, it’s very hard to see how scientific research is ‘best’ simply because it produces more papers.
Meanwhile, Mizrahi claims that scientific knowledge is qualitatively superior because scientific knowledge is better at inferring to the best explanation than non-scientific knowledge: a best explanation, in Mizrahi’s view, having the features of being unified, coherent, simple, and testable (360; also see Mizrahi 2012, 134). This, I think, is true, assuming that one important caveat in place. Scientific knowledge is indeed better at inferring to the most unified, coherent, simple, and testable explanation than non-scientific knowledge—but only in domains where unified, coherent, simple, and testable explanations constitute best explanations. Unfortunately, Mizrahi’s account is not sensitive to this fact; he thinks that all explanations seek to do basically the same thing, and consequently does not discriminate between the pragmatics of different kinds of explanations in different disciplines.
Kinds of Explanation
An example may be helpful. Mizrahi claims that inferring to the most unified, coherent, simple, and testable explanation is present in literary analysis, via the question: “What makes a good story?” In pursuit of this, he argues that what is called ‘archetypal’ or ‘myth’ criticism—an approach that emerges from the works of Carl Jung, Northrop Frye, Maud Bodkin, and Claude Lévi-Strauss, among many others—best explains the enduring appeal of certain literary works in a way that is unified, coherent, simple, and testable (361). There is at least one problem with this assumption—even ignoring the fact that archetypal criticism categorically fails Mizrahi’s own requirements for what constitutes a best explanation.
Mizrahi’s account is premised upon a fundamental misunderstanding of what literature scholars do and why they do it. Scholars who analyse literature are not in the business of explaining why people like to read certain things: they don’t ask of themselves “What makes a good story?” First and foremost, this is a rather boring question. There is, after all, no great mystery why we prefer some stories over others: people like plots that are exciting, characters that are believable, writing that is evocative, and so on. Moreover, as far as ‘scientific’ explanations go, it’s both far more parsimonious and far more predictively rich than those analyses offered by archetypal criticism: if you write, in good English (or whatever), an exciting story peopled with believable characters, there’s a pretty good chance that people will like it.
Because those sorts of questions have boring explanations, literary scholars tend not to ask them. Indeed, there is little inherent value for literary theorists in explanations of literature that are unified, coherent, simple, and testable. Instead, for a literary theorist to infer to the best explanation she should infer to the most interesting explanation, given the boundary conditions imposed by the text. An explanation might be interesting because it is unexpected; because it offers some hidden insight into our circumstances; because it speaks to our status as creatures both embodied in space and embedded in systems of representation. The literal truth or falsity—let alone the unifiedness, coherence, simplicity, or testability—of these kinds of analyses is almost accidental: a very different beast to scientific explanation indeed.
Furthermore, this observation is generalizable: what constitutes a good explanation in one domain may not be a good explanation in another. While Mizrahi’s good-making criteria for the ‘bestness’ of an explanation are broadly appropriate for both the natural and social sciences and much everyday reasoning, they are entirely inappropriate measures of ‘bestness’ in domains, such as in much of the humanities, that are governed by different explanatory norms. This, in short, is one reason why philosophers are justified in being suspicious of scientism, as Mizrahi defines it. Scientism, so sure of the universalizability of its own internal norms, assumes a) that all disciplines have the same standard of explanation as the natural and social sciences and then b) judges those disciplines according to those standards of explanation. It is little wonder, then, that Mizrahi thinks that scientific knowledge is the best kind of knowledge.
So, does this principled suspicion of scientism necessarily translate into hostility towards the use of empirical methods in philosophy? Absolutely not. As Robert Bishop (2019) rightly points out, introducing scientific methods into philosophy is an example of interdisciplinarity, not scientism. While scientism does indeed privilege experimental methods, among other methods that produce explanations that are unified, coherent, simple, and testable, there is nothing in experimental methods themselves that implies or entails scientism. To be interdisciplinary is to be methodologically casuistic, using the best tools at your disposal for the problem in front of you. It is for this reason that until very recently (and as a number of commentators have observed in this publication), experimental methods were a key part of the methodological toolbelt of a philosopher: there are some explanations that, given certain specific contexts, are more useful when they are unified, coherent, simple, and testable.
So, what of how empirical methods feature in this purported battle for the soul of philosophy? Although I’m hesitant to make too final a statement on this point, my intuition is that this fear (such as it is) ties into philosophers’ conception of philosophy as a basically normative discipline. Although there are clear exceptions, a great many of the foundational questions in philosophy are about establishing evaluative standards of one kind or another: “Why should I be good?”, “Why should I believe this rather than that?”, “When is something beautiful?”, and so on. Furthermore, and as part of this enterprise, it can be helpful for philosophers to employ empirical methods when establishing these evaluative standards, as Mizrahi correctly observes (2019, 11).
That said, because philosophy is basically normative there are some hard limits on the kinds of philosophical questions that empirical methods can answer. While empirical methods may be helpful in establishing the relevant facts of the matter, the actual philosophical process of establishing evaluative standards—the how and why of why certain concepts can and/or should hang together—necessarily exceeds the explanatory ambit of those methods. This is, naturally, because philosophy has different explanatory standards than the natural and social sciences: standards that differ from Mizrahi’s standards of unifiedness, coherence, simplicity, or testability. Given my defence of preserving distinct explanatory standards between disciplines, this is a category error that we would do well to avoid.
Contact details: Ryan Mitchell Wittingslow, University of Groningen, firstname.lastname@example.org
Bishop, Robert C. 2019. “Scientism or Interdisciplinarity?” Social Epistemology Review and Reply Collective 8 (12): 46-49
Mizrahi, Moti. 2019. “The Scientism Debate: A Battle for the Soul of Philosophy?” Social Epistemology Review and Reply Collective 8 (9): 1-13.
Mizrahi, Moti. 2017. “What’s So Bad About Scientism?” Social Epistemology 31 (4): 351-367
Mizrahi, Moti. 2012. “Why the Ultimate Argument for Scientific Realism Ultimately Fails.” Studies in History and Philosophy of Science 43 (1): 132–138.
 Archetypal criticism fails Mizrahi’s own requirements for what constitutes a best explanation because it is neither unified, coherent, or simple, by his own definitions:
1. First, archetypal criticism attempts to collapse the countless exigencies of human myth and culture into a small number of archetypes drawn from Western psychoanalysis. This heavy-handed reductionism is clearly incompatible with the spirit of Mizrahi’s ‘unification’: “choose the explanation that explains the most and leaves the least unexplained things”.
2. Second, archetypal criticism is grounded in the assumption that all humans share certain pre-cultural mythopoetic structures. There is no evidence for this claim. Consequently, archetypal criticism is not coherent because it is not “consistent with background knowledge”.
3. Third and finally, positing the existence of these pre-cultural mythopoetic structures in the absence of evidence explicitly violates his principle that we should “choose the least complicated explanation” (Mizrahi 2012, 134).
Even the criterion of testability doesn’t come easy. Archetypal criticism is only testable insofar as we can use it as a framework to make sense of a given piece of literature, in much the same way that we can use any other analytical method. However, there is no principled way by which we can or should prefer one given method over another; no explicit criteria of falsifiability or measurable capacity for predictive power that guarantees the efficacy of one method over another.
Categories: Critical Replies