Explanationism in metaphysics
Explanations and explanatory inferences abound in metaphysics; sometimes it is hard to see how metaphysics could even be done if not by comparing potential explanations. I will not question the credentials of all such explanatory reasoning tout court, but rather focus on explanationists who emphasize the continuity between metaphysics and science. Here are some exemplars of this more specific target, in alphabetical order.
Armstrong (1983, 1997) famously argues for his account of natural laws via inference to the best explanation. Armstrong also explicitly likens his argument to explanatory inferences to theoretical entities in science. Even induction by reference to laws “becomes a particular case of the inference to explanatory (‘theoretical’) entities,” and “the law, a relation between universals, is a theoretical entity, postulation of which explains the observed phenomena and predicts further observations” (1983, 104). Armstrong ultimately says precious little about the nature and justification of this kind of inference in general; he just maintains that the main explanatory advantage of his laws-as-universals view is unification, the relevance of which Armstrong supports by reference to the unificationist theories of scientific explanation.
We avail ourselves in [metaphysics] of whatever apparently good cannons of explanation we possess or can develop. The basic insight here seems to be involved in the concept of a good explanation: that it should genuinely unify, and that it should be genuinely informative. (Armstrong 1983, 105)
The premise required here is that, as argued by Michael Friedman [and Philip Kitcher], in the natural sciences what, before anything else, counts as a good explanation is something that unifies the phenomena. (Armstrong 1997, 235)
We can continue the list of like-minded explanationists with Bigelow, Colyvan, Dorato, Ellis, and so forth. All these prominent philosophers—mostly from
Australasia, as it happens—have advocated a conception of scientific realism according to which inference to the best explanation can support not only realism about electrons and quarks and the like, but also metaphysical views about, for example, mathematical abstracta, possibilia, laws of nature, the nature of properties and probabilities, and mereology. In the case of Ellis (2009), for instance, the explanationist methodology results in “metaphysics of scientific realism" the ontological count of which includes properties, powers, causes, events, propensities, dispositions, and spatiotemporal and numerical relations, all of which are argued for via inference to the best explanation—a methodology unifying metaphysics and science.
Psillos (2005), an explanationist in philosophy of science, sympathizes with Ellis’s explanationism by acknowledging that Ellis’s project “rests on the only workable criterion of reality . : something is real if its positing plays an indispensable role in the explanation of well-founded phenomena" (398). As Psillos notes, there is a close connection between this “explanatory criterion of reality," also found in Sellars ( 1991), and the well-known indispensability argument for mathematical Platonism. The latter, as championed by its leading contemporary advocates, is again directly associated with explanationism in science.
Swoyer (2008) presents an argument for ontological commitment to mathematical and other abstracta that is quite unlike the indispensability argument. But it is also a forthright instance of explanationism, with a clear emphasis on methodological similarity with science.
The chief philosophical benefit claimed for ... abstract entities is that they ... help explain otherwise puzzling phenomena. . [IBE] is not some arcane concoction of meta-physicians. We often infer that something exists on the grounds that its existence would explain something that would otherwise be puzzling. . Such inferences also seem common in science. (2008, 16)
Swoyer reflects on the methodology of metaphysics, and for him explanationism represents the answer to “how ontology might be possible" (Swoyer 1999). The idea that explanationism and its methodological affinity with science provide an answer to methodological scruples about metaphysics has also been recently defended by Paul (2012), among others.
Explanationism is popular in meta-ethics as well. Here an explicit connection to explanationism in science is drawn by Sturgeon (2006), for example. Examples proliferate; clearly explanationism, with a nod to science, flourishes in metaphysics and beyond. Every philosopher above has their critics, of course, but no wholesale assessment has been given of their common methodological denominator: a widespread justificatory appeal to a similarity between metaphysical theorizing and the explanatory practices of science. There are obvious questions to be asked here: What kind of similarity is this? What kind of justification does it support? What are the limits of such justification of metaphysical views ?
I will next characterize in more detail two ways in which philosophers have attempted to capitalize on the similarity between naturalistic metaphysics and science, setting the scene for a more critical discussion to take place in sections 4 and 5.
-  2 See, e.g., Bigelow 2010; Bigelow and Pargetter 1990; Colyvan 2006; Dorato 2012; and Ellis 2009.
-  According to Ellis, the test of a good metaphysical hypothesis is twofold: “it must be consistent with the knownfacts, and be part of a unifying account of reality that explains the overall structure of what we are able toobserve" (2009, 123).
-  See, for example, Colyvan (2006, 229): “I will take the indispensability argument to be an argument that putspressure on the marriage of scientific realism and nominalism. It does this because the style of argument [viz.IBE] is one which scientific realists already endorse."
-  See Paul (2012, 22): “This is a central part of my thesis: if we accept inference to the best explanation in ordinaryreasoning and in scientific theorizing, we should accept it in metaphysical theorizing.”
-  See Sturgeon (2006, 243): The justification of many of our beliefs, not just perceptual beliefs, derives from their explanatory role.Thus, for example, the justification of many scientific beliefs is said to lie in their contributing to goodexplanations of observed evidence ... and beliefs about others psychological states may explain whatwe observe of their behaviour. . [A]lmost everyone agrees that explanatory coherence ... is an epis-temic virtue in a set of beliefs, contributing other things equal to its being well justified. And this isthought to be especially so when there is explanatory integration across significantly different categories: beliefs about the past explaining evidence in the present, or beliefs about unobservables explaining what we observe, for example. Now, if evaluative beliefs play a plausible role in explaining otherfacts we have reason to believe in, facts that are not themselves evaluative, then they will be candidatesfor justification of a similar sort, by their integrative explanatory role.
-  In the sphere of epistemology, Pargetter (1984) advocates explanationism as a response to skepticisms aboutother minds by viewing inferences to other minds as being relevantly similar to arguments to the best explanation in science. McLaughlin (2010) makes comparable use of inference to the best explanation in philosophy ofmind, and Biggs (2011) runs a similar line of thought in relation to modal epistemology.