Nothing explains essence

Abstract

ABSTRACTI argue that there is significant pressure to adopt a domain-fixing conception of essence if one wants to vindicate the claim that essentialist facts are proper ends of metaphysical explanation. I then argue that the zero-ground account best accommodates a domain-fixing view. I then respond to a number of objections to the zero-ground account and show how they can be met. I conclude that there is good reason to give the zerogrounding view more serious attention than it has received.KEYWORDS: EssencegroundexplanationZero-Groundmodality Disclosure statementNo potential conflict of interest was reported by the author(s).Notes1 For discussion of fundamentality see Fine (Citation2001), Schaffer (Citation2010), Bennett (Citation2017) and Sider (Citation2011), among others.2 For discussion of zero-grounding see Fine (Citation2012), Litland (Citation2017), Donaldson (Citation2017), Muñoz (Citation2020), and more recently, Kappes (Citation2020a; Citation2020b; Citation2022). The notion of being generated from nothing may be stronger in the case of grounding than in corresponding case of theorems. Given that validity is monotonic, it follows that a theorem is also derivable from any set of premises. But because ground is not monotonic, it may turn out that a zero-grounded fact is only grounded in the zero-ground.3 I’m not the first to advance the zero-grounding view in relation to the explanatory distinctiveness of the essentialist facts. Miller (Citation2022) and Kappes (Citation2020a) both argue that this view is motivated by taking the essentialist facts to play the role of explanatory links. These sorts of considerations will not play a particularly significant role in the argument of this paper.4 I borrow this term from Raven (Citation2020).5 In particular that it is asymmetric, transitive, irreflexive. We thus obtain a partial ordering over the domain of facts. I note the standard assumptions here to help the reader gain a sense of what grounding is supposed to be, although nothing I say in what follows hangs on these assumptions being wholly vindicated across all instances of grounding. I also take partial ground to be definable in terms of full ground. P is a partial ground for Q just in case P on its own or together with some other facts fully grounds Q. See Fine (Citation2012, 50).6 See e.g., Mason’s (Citation1967) translation of the Leibniz-Arnauld Correspondence p. 13–17 for a rich expression of Leibniz’s view.7 One must, however, be careful here. On this framing, it appears that objects can have essences without existing which engages with some rather delicate issues about being, existence, and quantification, the discussion of which lie beyond the scope of the present paper. In particular, such a conception appears in tension with what Williamson calls the being constraint. Roughly, the being constraint says that having properties requires existence, and it corresponds model-theoretically to the domain constraint; that all predicate extensions be drawn from the domain of that world. See Williamson (Citation2013) § 4.1 for critical discussion.8 This case is borrowed from Kment (Citation2014) §6.2.1.9 It’s worth pausing here to point out that invoking essences in this way looks like it appeals to a principle often referred to in the literature as Essence Grounds Prejacent (EGP) s: □xP ≺ P. This principle looms large in the background of a significant portion or recent work by metaphysicians theorizing in terms of essence and ground. However, it has recently come under more scrutiny. See Kment (Citation2014) and Rosen (Citation2015) for discussion in favor of EGP, and Glazier (Citation2017) as well as Zylstra (Citation2019) for objections. Here no stand is taken on the truth of the generalization expressed by the principle. In so far as essence theorists take essences to be explanatorily distinctive in the way characterized above, they will recognize the truth of a large class of instances of EGP. All we are interested in presently is developing an account of essence that explains how it can play the explanatorily role sketched above as exhibited in the large class such instances.Despite no present stand being taken with respect to the generalization expressed in EGP, it is worth noting that were the generalization to be true, it would be germane to the argument developed over the next few sections. As Kappes (Citation2020a) has pointed out, there is significant pressure in light of EGP to be open to the zero-grounding account. Here is a terse reconstruction of Kappes’s argument. Given that essence iterates, essence facts are explained in further (iterated) essence facts (similar to how EGP would have it). But Kappes argues that it is a mistake to take the explaining essence fact to be a reason why the explained essence fact obtains. Instead, his proposal is that the explaining essence fact is a link of what he calls ‘empty-base explanation’ that generates the explained essence from nothing. This is analogous to the proposal being made in the present paper. Where Kappes takes the explanatory link that generates an essentialist fact from nothing to be a further essentialist fact, the present proposal suggests that the link is an instance of zero-grounding. Kappes (Citation2020a) anticipates this and later argues that if we are unsatisfied with his account of essentialist facts as explanatory links, we might be able to appeal zero-grounding grounding as a way to make sense of the intuition behind explanation by essential status. One way of taking the aim of this paper is to increase our confidence in Kappes (Citation2020a) claim.10 Fine (Citation2005) characterizes unworldliness in contrast to worldliness. For Fine, a truth/fact is worldly just in case it is made true by/grounded in the circumstances of the world. It is unworldly just in case it is true regardless of/is metaphysically independent of the circumstances of the world. This distinction is meant as a modal parallel to the distinction between being true at a time and being true timelessly.11 Raven has recently made it clear that he didn’t intend this to be a conception of essence per se. Rather it was meant to be a certain view as to what the grounds of certain essentialist facts are. It was meant to leave open whether other essentialist facts were not grounded in generative facts. Raven (personal correspondence)12 It is important to note, however, that they may be satisfactory stopping points in a different sense than the one presently being discussed. Glazier, partly out of dissatisfaction with ground-theoretic ways of understanding what it means to be an explanatory end, proposes essentialist explanation as a distinctive kind of explanation. On his view, essentialist facts can be ends of explanation if they admit of no further essentialist explanation. See Glazier (Citation2017). It may very well be the case that the proponent of the generative view can still maintain that essentialist facts are explanatorily distinctive in this special non-ground-theoretic sense. Thanks to an anonymous referee for pressing this point.13 Glazier (Citation2017)’s argument against the fundamentality of essentialist facts takes this line.14 This analogy is due to Fine (Citation2012) and Litland (Citation2017).15 Some have argued that grounding does not entail necessitation. See e.g., Leuenberger (Citation2014) and Skiles (Citation2015). Typical worries stem from trouble in telling a well-motivated story about what grounds certain kinds of universal generalizations. I find these kinds of worries generally unmoving. For a convincing line of reply see DeRosset (Citation2023, ch 8 § 3).16 This is typically cashed out in terms of modal recombination principles. Recombination is supposed to follow from the Humean principle of ‘no necessary connections between wholly distinct entities.’ See e.g., Lewis (Citation1999; Citation2001). There is an interesting question as to how to determine when two entities (in this case, facts) might fail to overlap. Dixon (Citation2016) proposes a characterization of overlap between facts in terms of what he calls groverlapping. If two facts are different, they groverlap just in case (1) either one partly grounds the other or (2) they share a common ground. Fundamental facts certainly won’t groverlap. One might on this basis attempt to directly vindicate a modal recombination principle for the fundamental facts. For discussion of this sort of argument see Wang (Citation2016).17 Here we rely on the grounding machine metaphor to cash out the notion of zero-grounding. Despite the power of the metaphor, it would be nice if something could be said about that of which it is a metaphor. A very attractive picture developed in Litland (Citation2017) ties the mechanisms of the grounding machine to explanatory arguments. In such a setting the notion of zero grounding is much less mysterious. A truth ϕ is zero-grounded if there is an explanatory argument from the empty collection of premises to the conclusion ϕ (Litland Citation2017, 298). We can think of explanatory arguments as composed from basic explanatory inferences; e.g., conjunction-introduction, disjunction introduction, the inference from a is F to a is G – where F is a determinate of the determinable G. (Litland Citation2017, 289).It is natural to see these explanatory inferences as importantly connected to the essences of certain items (e.g., conjunction, disjunction, etc.). Some attention has been focused on developing inference-based accounts of the nature of logical operations. As an illustration, on such a view we may define conjunction as the operation such that one can infer P ⋀ Q from P, Q. (see e.g., Fine 1994 and Correia Citation2012). Call the inference involved in this definition of conjunction an essentialist inference. We can exploit this essentialist inference to determine when certain truths involving ‘⋀’are zero-grounded – in particular, the truth that [P, Q < P ⋀ Q], since we will be able to construct an explanatory argument from the empty set of premises to [P, Q < P ⋀ Q]. See Litland (Citation2017, 302–304).One may attempt to extend this picture more generally and characterize the essences of other items in terms of essentialist inferences. Take our knife K1 composed of handle H1 and blade B1. On such a view, we might consider the following an explanatory argument involving K1: [H1 is joined with B1 .˙. K1 exists]. If this is an explanatory argument, it’s natural to see it as involving an essentialist inference that is related to the essence of K1. We might try to exploit this essentialist inference in a way analogous to the case of conjunction above to show that from no premises we can derive [if H1 is joined with B1 then K1 exists], which on this view would be a zero-grounded truth. Interestingly, this ‘zero-grounded’ truth seems to express exactly what we would want to regard as a constitutively essential truth about K1.This is the barest sketch of one way to cash out the grounding-machine metaphor for zero-grounding and how to assimilate Nothing Explains Essence into that picture. The devil, of course, is always in the details; details which I leave to future work.18 Muñoz (Citation2020) argues that certain facts are zero-grounded but only contingently so (in particular, negative existentials). This however relies on there being good reason to reject necessitation, which I do not think there is. See footnote [26]. For further productive discussion of the differential modal import of being fundamental/ungrounded and being zero-grounded, see De Rizzo (Citation2021).19 For discussion of the Principle of Sufficient Reason generally see Della Rocca (Citation2010) and Amijee (Citation2020).20 Raven (Citation2020) develops such a worry.21 That is, the zero-ground account can resolve the tension without having to motivate some restriction on the ‘all’ as it is featured in the typical expressions of metaphysical rationalism.22 It’s worth mentioning that some have appealed to the role of explanatory arguments in grounding to make sense of how facts like [p ⋀ p] and [p ∨ p] can have the same ground, yet different explanations. See e.g., DeRossett (Citation2013, 22–24). If that is ultimately the correct way of squaring away Dasgupta’s worry, we have reason to be optimistic about the zero-ground account, given the ways in which it seems to be able to accommodate such a role for explanatory arguments. See footnote [31].23 Many thanks to Jon Litland, Louis deRosset, Josh Dever, Rob Koons, Mike Raven, Derek Haderlie, Carlos Romero, as well as two anonymous referees at Inquiry for their constructive interventions in the project and for the ways in which their comments improved the quality of this manuscript.