* this description has been adapted from our original funding application *

In this project, we aim to investigate the nature of metaphysical explanation. This central notion in metaphysics is underexplored in its own right. By considering metaphysical explanation by analogy with more familiar forms of explanation, we hope both to elucidate the notion, to justify its continued use in metaphysics and elsewhere, and to shed new light on questions about fundamentality, grounding, truthmaking, and explanation generally.

Metaphysical explanation is typically understood as explanation that accounts for the existence and/or nature of something in terms of something else on which it constitutively depends or, to use by now standard terminology, in which it is grounded. As an example of explanations of this kind, consider accounts of the nature and existence of so-called concrete particulars in terms of the existence of bundles of abstract tropes. Or consider accounts according to which properties exist and are what they are because of or in virtue of the existence of some (structured) sets of concrete particulars. These are two examples from the field of metaphysics, but we believe metaphysical explanation has even broader application. To see this, consider debates;

  1. in the philosophy of language over whether and how the meaning of a sentence is grounded in its social or biological functions;
  2. in the philosophy of mind over the intimate relationship between consciousness and brain activity;
  3. in ethics over whether a thing’s moral properties depend on its natural properties;
  4. in social and political philosophy, over whether the existence and nature of gender, race, and class ought to be understood with reference to a certain group’s collective intentions and behavior towards certain people, and;
  5. (to take just one example from outside philosophy) reflect on the now familiar conception of the sciences as divided up into ‘levels of scale’, with the more macroscopic levels (like those described by geology or biology) grounded in the more microscopic levels (like those described by chemistry or physics).

The concern in all of these debates is that of giving an explanation. These are certainly not causal explanations, for they do not involve essential reference to causal relations. Rather, these are explanations which account for the existence and/or nature of something in terms of that on which it constitutively depends. If these are explanations, therefore, they are best characterized as metaphysical explanations. Making sense of metaphysical explanation will hence not only put its theoretical role in metaphysics on much firmer footing; it will also help legitimize its use in countless other disciplinary contexts.

Although the nature of metaphysical explanation per se is surprisingly little discussed, its relation to grounding and related notions is all the more debated. According to recent orthodoxy, this relation is intimate: grounding is an explanatory relation. In fact, that grounding is explanatory is often said to be what sets it apart from purely modal notions of dependence and determination, such as, e.g., supervenience. A fact that is said to provide the proponent of grounding with a reason for adding it to the metaphysician’s ‘toolkit’. Relatedly, grounding relations are said to ‘inherit’ their properties from the properties of metaphysical explanation. It is because grounding relations are explanatory that they are e.g., asymmetric, irreflexive, transitive, well-founded, and/or nonmonotonic. And where there is disagreement over which of these properties grounding has, it is because there is disagreement over whether or not the relevant property is characteristic of (metaphysical) explanation. But to say that grounding and metaphysical explanation are intimately connected is not yet to give an account of how they are related. Nor does saying that grounding relations ‘inherit’ the properties of metaphysical explanation tell us what those properties are. Here the debate is underdeveloped. Metaphysical explanation, we are told, relates to grounding relations either by characterizing or by tracking them. We are rarely offered more detail, but this is very far from an innocent disjunction. If saying of grounding relations that they are explanatory is to characterize them, then at least some of the things true of grounding ought to be true of metaphysical explanation as well. In particular, if grounding relations are objective constituents of mind-independent reality (a common assumption), then whatever characterizes them ought to be objective and mind-independent as well. But where does this leave us with regards to the epistemic virtues normally taken to be essential to explanation? What about the role of understanding and other subjective and context-sensitive factors?

One might claim that these are questions that need not concern the student of metaphysical explanation precisely because metaphysical explanation is so very different from ‘ordinary’ explanation. Coming to grips with its true nature, one might argue, will require innovation rather than emulation. A theory of the nature of metaphysical explanation, then, is a theory describing an entirely new kind of explanation, a kind of explanation that has little more in common with explanation of a more familiar variety than that it falls under the same umbrella-term. One worry with going down this road is that we end up with something too far removed from the genuine article to deserve the title ‘explanation’. Another worry is that it is now no longer clear on what grounds we justify claims we make about (the nature of) metaphysical explanation. This is because what intuitions we have about explanation are intuitions derived from our experience of explanations of a more ‘ordinary’ kind.

The alternative is to try to develop an account according to which metaphysical explanation tracks but does not literally characterize grounding relations. On this view, although the holding of worldly grounding relations is necessary for there to be metaphysical explanation, it is not sufficient. This means that room can be made for the epistemic virtues to play a role. And this means that we no longer have to consider metaphysical explanation as radically different from explanation of a more familiar kind. Rather than innovate, we can therefore emulate. And one pre-existing account of explanation seems especially suited to provide a model: causal explanation is normally understood as a kind of explanation in which some underlying worldly relation – causation – plays a role, but where for there to be an explanation, more than the holding of this relation is required. If metaphysical explanation tracks but does not characterize underlying grounding relations, the same seems to be true of it.

With this in mind, we will approach the question of what is the nature of metaphysical explanation via several mutually interrelated investigations.

Research Questions

(RQ1) How does metaphysical explanation relate to grounding relations? We intend to argue that, to be a genuine kind of explanation, metaphysical explanation must leave room for the usual epistemic virtues, and in particular for understanding, to play a role. This rules out a view according to which metaphysical explanation relates to grounding relations purely by characterizing them. Instead, metaphysical explanation relates to grounding relations by tracking them. But what does this mean, more precisely? We will answer this question by tackling a number of more specific research questions, including:

  1. What is ‘tracking’? If metaphysical explanation explains by tracking grounding relations, we need to know more about this relationship. On one problematic view (cf. Kim, 1994), tracking is portrayed as a kind of grounding (as in: if a tracks b, b grounds a). But this account runs the risk of circularity, and is in any case highly uninformative. For, arguably, countless propositions that are clearly not metaphysical explanations are grounded in worldly states of affairs (if Truthmaker Maximalism is accepted, all true propositions are). We will therefore propose an alternative account.
  2. Does accepting the ‘tracking’ view mean having to treat grounding itself as (a kind of) causation? According to some very recent research (Schaffer 2016, A. Wilson, forthcoming), grounding and causation have more in common than their both being ‘tracked’ by explanation, namely their being relations of the same kind. Could this idea help throw some light on the nature of metaphysical explanation? In dealing with this question, we will contribute to this recent debate both constructively and critically.
  3. What – if any – is the role of laws in metaphysical explanation? Typically, causal explanation not only involves citing the presence of some underlying causal relation, but also subsuming this particular state of affairs under a law of nature. If metaphysical explanation is modeled on causal explanation, is the same true of it? If so, what kind of law can take the place of a law of nature? A law of metaphysics? Laws of metaphysics seem to be rather different from laws of nature, and in ways that might seem perplexing. We will investigate and formulate views on the nature and function of metaphysical laws generally, as well as on their role in metaphysical explanation.

(RQ2) What makes a relation ‘explanation-apt’? Even if we intend to argue that metaphysical explanation must leave room for epistemic virtues to play a role, and even if this means that metaphysical explanation cannot be understood purely in terms of the (nature of the) grounding relations it tracks, the nature of those relations will nevertheless play a role in metaphysical explanation. So what is it about relations such as grounding (and causation) that make them, but not some of their close cousins, ‘explanation-apt’? We could not answer that grounding is ‘explanation-apt’ because it is explanatory. We will therefore look for a more substantive and noncircular answer to this question.

(RQ3) What are the properties of metaphysical explanation? We will assess the claim that grounding relations ‘inherit’ their properties from the properties of metaphysical explanation, among other things to see how likely it is given that metaphysical explanation tracks, but does not characterize, grounding. According to the orthodoxy, grounding relations ‘order’ levels of reality from bottom to top, and thus are asymmetric, irreflexive, and transitive. We will investigate whether these are properties reasonably attributed also to metaphysical explanation. Among other things, this involves taking a stand on whether or not transitivity is compatible with a view on the nature of metaphysical explanation according to which it essentially involves understanding (which most likely isn’t transitively transferred). It will involve deciding whether taking explanation to be irreflexive means regarding as false the idea that everything must have an explanation (assuming that some things are fundamental and hence not grounded in something further). And it will involve answering questions to do with the presumed asymmetry of metaphysical explanation, including that of whether a symmetric explanation must be considered circular, and whether, if it must, circular explanations must be considered essentially flawed.

(RQ4) What role do the epistemic virtues play in metaphysical explanation? One problem for the view that metaphysical explanation characterizes grounding relations was that this seemed to leave little room for the usual epistemic virtues – and in particular for understanding – to play a role in explanation. But why think that epistemic virtues must play a role in explanation? Which virtues play such roles, and how? Reflecting on these and related questions will, apart from informing us about the issue at hand, also help further our understanding of other kinds of explanation – such as scientific explanation – for which the role of the epistemic virtues, and in particular the role of understanding, arguably remains something of a mystery.

(RQ5) Must explanation come to an end? If explanation must come to an end, does t his mean that there must be a fundamental level of ungrounded entities? If everything must have an explanation, does this mean that everything that exists is grounded in something further? These are questions about exactly how closely tied metaphysical explanations are to the grounding relations they track. The common view is that this relationship is pretty close: if explanation must come to an end, there must be a fundamental level of ungrounded entities. If everything must have an explanation, grounding must go on indefinitely. But why ought this to be accepted? In dealing with this and related questions, we will contribute new perspectives to an already lively debate on fundamentality, infinite regress, and the principle of sufficient reason.