Abstract

In this paper I present and critically discuss Simon Evnine’s account of hylomorphically complex objects (as presented in his 2016 book Making Objects and EventsEVNINE, S. Making Objects and Events, Oxford: OUP , 2016a.). On the one hand, I object to the account he gives of how artifacts (which are for him the paradigmatic cases of hylomorphically complex objects) allegedly acquire their existence and identity conditions. I elaborate on two problems I see for this account: first, that it seems unable to explain our knowledge of the kinds to which artifacts belong; second, that it cannot offer a plausible solution to the grounding problem for coincident objects. I also object to the way in which he tries to adapt the sort of account he gave for artifacts to the case of organisms (in my view this fails because both cases are dissimilar at crucial points), and finally I also object to his attempt to extend that account, in a fictional way, to the case of natural non-organic objects (as I try to show, both his arguments to the effect that there are no such objects, and his positive fictionalist proposal to account for our talk about them, are flawed).

Keywords:
Object; Matter; Form; Intention; Concept; Artifact

1. INTRODUCTION

What makes it the case that certain material objects, say, the xs, compose a different object, y, that is, a whole of which those xs are parts? And in what circumstances could it be the case that such a y may at some time be constituted by the xs, but at some other time by the (possibly partially distinct) zs? These are some of the questions that Simon Evnine tries to answer in his recent book Making Objects and Events (2016EVNINE, S. Making Objects and Events, Oxford: OUP , 2016a.). And in giving answers to these questions, of course, the book offers in fact much more: it argues for a general view concerning material (and abstract) objects (and events). In what follows I will present and criticize some of the main tenets of his proposal, focusing in particular on some problems for his account of material objects.

Of course, there is already a long history of answers to both of the questions mentioned at the outset. The first question has been given much attention recently after Peter van Inwagen presented and discussed it at length in his groundbreaking monograph Material Beings (1990)VAN INWAGEN, P. Material Beings, Ithaca, NY: Cornell UP, 1990.. The second question has been around for even longer, as the continuing discussion of some classical puzzles, such as the one concerning the ship of Theseus, shows. Now, the simplest (and at least for some time also the most popular) answers to these questions have been those that follow from some extreme views regarding composition: so, according to Nihilism, for instance, composition never takes place, as a result of which the material world would be inhabited by (mereological) simples arranged in different ways - and, given that there are no composites, our second problem doesn’t even arise; on the other hand, according to Universalism, composition always takes place: for any set of objects, there is a further object they compose. As a result the world contains simples but also all sorts of ordinary and “extraordinary” combinations; according to its simplest version, besides, for any xs to compose y is for them to be identical to y, so that, in connection with our second question, nothing but the xs could ever constitute y.

These are nice answers, mainly because they are simple, but they are really non-starters as attempts at reconstructing our usual, common-sense notion of composition, according to which there are chairs (and not just simples “arranged chair-wise”), which may besides remain the same through changes in their parts (what Evnine very aptly describes as their “having a metabolism”), and according to which there is nothing composed of my chair and my cat, even when the cat is sitting on (or, for that matter, in any way bounded to) the chair.

Now, giving an adequate reconstruction of this notion is no easy task, and most of the attempts that have tried to account for it in terms of the obtaining of one single, or a limited number of, general wordly relation(s) among the xs, as sufficient for them to compose something, have failed, as there’s no shortage of counterexamples to such views: consider for instance the simplest proposal discussed by van Inwagen, according to which some xs compose a new object y iff the xs are in contact with one another; our example above of my cat sitting on a chair would already be a counterexample: according to the proposed account they should compose something (as they are in contact with one another), but, intuitively, they don’t.

The difficulties in finding a neat account of our received notions of composition and of persistence through change have prompted different attitudes among philosophers working in this area: some went on trying to come up with direct answers to the questions, in the spirit of van Inwagen’s proposals;¹ 1 Carmichael (2015), for instance, presents a “series-style” answer of this sort - where a ‘series-style answer’ is one which appeals to different relations among the xs for different sorts of composed y. some have adopted one or another of the simpler answers mentioned above (nihilism, universalism, etc.) in an explicitly revisionary spirit; and, finally, some have argued in favour of incorporating in the account a reference to something like a kind or a concept, that is, to something flexible and subtle enough so as to make room for different, sometimes apparently arbitrary, conditions for the existence and persistence of different sorts of objects.² 2 The difficulties just noted appear, independently, in connection with both questions concerning material objects mentioned at the outset. For instance, Fine’s proposal of understanding ordinary material objects in terms of “variable embodiments” makes use of something akin to a form to explain an object’s metabolism, but it would be just a more complex form of universalism as regards the problem of composition - giving rise to even more “extraordinary objects” than universalism itself. As making a taxonomy of the different views is not my main purpose here, I will be mostly concerned only with the broader distinctions and will ignore these subtler ones.

Now, this idea could be (and has been) implemented in different ways. One of them, expressing a broadly realist stance, has been to take kinds (or something else to the same effect: functions, properties, K-paths) as constituents of the objects themselves, where kinds may be understood either in a sparse and heavy-weight manner (Koslicki 2008KOSLICKI, K. The Structure of Objects, Oxford: OUP , 2008, Korman 2015KORMAN, D. Objects. Nothing out of the Ordinary, Oxford: OUP , 2015.), which would allow a fuller validation of our ordinary conception, or else as rather abstract parts of the objects (or of formal devices representing them) (as in Fine 1999FINE, K. “Things and their Parts”, Midwest Studies in Philosophy 23: 61-74, 1999., 2008_____ “Coincidence and Form”, Aristotelian Society Supplementary Volume 82: 101-118, 2008., Sattig 2015SATTIG, T. The Double Lives of Objects, Oxford: OUP , 2015., and Bennett’s 2004BENNETT, K. “Spatio-temporal Coincidence and the Grounding Problem”, Philosophical Studies 118: 339-271, 2004. plenistic proposal), which may validate only some aspects of such a common-sense view, giving up on others. The other way of implementing this general strategy, expressing what we could describe as a broadly “conceptualist” approach, consists in supposing that it is some attitudes on the part of human beings that determine under which conditions objects of different kinds exist (i.e., when their constituents satisfy some conditions encoded in our concepts) and persist (i.e., when their changes take place in ways that are acceptable according to the relevant concept). There are, again, different ways in which this broadly conceptualist strategy may be (and has been) implemented. The concepts involved in setting boundaries to the objects could be understood, either just as parts of a conceptual scheme, expressing (so called) application and co-application conditions, without being involved in the process through which the objects come to exist (as in the minimalist, easy ontology approach of Thomasson 2001THOMASSON, A. “Ontological Minimalism”, American Philosophical Quarterly 38: 319-331, 2001., 2007_____ Ordinary Objects, Oxford: OUP, 2007, 2015_____ Ontology Made Easy, Oxford: OUP, 2015), or they might be thought of as more substantially involved in such a process - for instance, as in Evnine’s own approach, according to which concepts are involved in the producer’s intentions (there are some other approaches belonging to this broad kind, differing in some of the details; cf. for instance Einheuser 2011EINHEUSER, I. “Towards a Conceptualist Solution of the Grounding Problem”, Nous 45: 300-14, 2011. or Sutton 2012SUTTON, C. “Colocated Objects, Tally-Ho: A Solution to the Grounding Problem”, Mind 121: 703-730, 2012, and our discussion below).³ 3 I’m painting here with a wide brush, and not trying to offer a comprehensive taxonomy of the views, but just to give an idea of the different directions in which they may be developed. There are views, such as that presented in Goswick 2018, which involve elements characteristic of both of the classes just distinguished: for her, for instance, an ordinary object is the sum of an abstract kind and some matter (so that the view is committed to independently existing kinds, which the objects have as parts), but according to this view the sum exists only in so far as somebody responds to the matter as if it was a thing of the relevant sort (and that’s the conceptualist aspect of the view). I hope the brief discussion above is nonetheless useful.

I largely agree with Evnine’s choices at the different decision points here: I think we should indeed reject the extreme, simple views on composition and persistence because of their revisionary consequences, and I also agree that we need to appeal to concepts and people’s intentions if we are to provide a reasonable conservative account of the objects there are and of their persistence conditions. In what follows, therefore, I won’t spend much time discussing (of even trying to ground my own acceptance of) these views of Evnine with which I agree, but will be mainly concerned with some of his more specific views concerning the roles our concepts should play in an account of ordinary objects that I find problematic. Before criticizing his views, anyway, I must first present them.

2. EVNINE’S VIEW

I have presented the problem to which Evnine tries to provide a solution in terms of two different questions: one of them is van Inwagen’s SCQ, namely, the problem of explaining under what circumstances some objects compose a further object; the other is the problem presented by the objects’ having a “metabolism”, that is, that of explaining how it is that such composite objects may remain the same through changes in their matter. In his own words:

He offers an “hylomorphic” account of material objects in response to both problems, where ‘hylomorphism’ is a general label he uses to characterize views according to which some objects are not identical to their matter. He endorses a rather peculiar variety of hylomorphism, which he describes as an “amorphic” one. This may sound incoherent on the face of it, but, of course, it is not: the idea is that what guarantees that the hylomorphic object is not identical to its matter is not a form, understood as a further constituent of the object alongside the matter, but simply the fact that the object has been produced (that is, that what was to be its matter has been arranged) so as to fulfil the descriptive conditions associated with a concept, fulfillment that makes it into an object of a specific kind. This particular take on hylomorphic composition is closely associated with two further ideas: on the one hand, with Evnine’s claim that artifacts provide the paradigmatic example of hylomorphic composite objects, as only in connection with artifacts is it the case that a concept, namely, the sortal concept in the mind of the producer, guides their production; on the other hand, with the further claim that hylomorphic objects are what he calls “ideal” objects, namely, that an idea is required (that is, in the mind of the producer) for a particular object to exist, and to belong to a specific kind.

The paradigmatic status of artifacts in Evnine’s view determines the structure of the defence of amorphic hylomorphism in his book, which takes place in two steps. He first defends his view for the case of artifacts, explaining how it would be appropriate to see them as having an hylomorphic structure and, as a result, a peculiar essence that makes them distinct from their matter. Then, in later steps, he tries to extend this account to other kinds of (alleged) material objects, such as organisms and natural non-organic objects (things like molecules and mountains). In the last chapter of the book, moreover, he makes a further extension of his view, by also providing an hylomorphic account of actions, understood as artifactual events; I think this is a very interesting aspect of the view presented in the book, and one I’m particularly sympathetic to; but, precisely because of this, I won’t discuss it in what follows, focusing mainly on what he has to say on material objects.

Now, let’s first consider with some detail the account given of artifacts. How precisely does an artifact come to exist? That is, what has to happen for some matter to become the matter of a new thing that belongs to a definite kind of human-made things and has fairly determinate persistence and individuation conditions? Evnine’s main idea here is that the production of an artifact requires two elements: on the one hand, what he calls ‘labour’, namely, the physical work of, say, cutting, arranging, glueing (etc.) the different chunks of matter that will eventually constitute the matter of the artifact;⁴ 4 Notice that ‘matter’ is used in two different senses in this clause: as equivalent to ‘something of a material nature’, in the first occurrence, and as ‘what something is made of’ in the second. The word will continue to appear in these two senses in what follows; I hope context will be enough to make clear which of them is meant in each case. and, on the other hand, some mental state in the maker, namely, the intention of producing a new thing of a certain specific kind. For analytic purposes we may further divide this state into an attitude (the intention) and the object of the attitude, which involves a conceptual aspect, which determines the specific kind to which the object to be produced will belong. The following quote adequately summarizes the view:

This basic scheme also allows us to explain how an object could change its parts over time while remaining the same. What’s required for this is that the relevant concept should determine the sort of metabolism the object is to have - which changes in the object’s parts are, and which are not, compatible with the object remaining the same. He then develops a set of notions to describe more precisely the changes in matter and how they allow for the object to belong to the same kind, even when the creative intentions were directed at a (possibly) different amount of matter. The most important among these notions is that of being a replacement, which is the relation that holds between quantities of matter just in case we get one from the other by a chain of successive losses or additions of parts. It then follows that, for an hylomorphic object to be constituted at some moment by some matter it must have been made with the relevant intentions by a producer either out of that same matter, or out of matter that has been replaced by its current matter.⁵ 5 Evnine characterizes two different senses of being made out of, a state and an event sense; I hope the distinction between ‘is constituted by’ (for the state sense) and ‘has been made out of’ (for the event sense, in this case for a past event) in the main text makes that distinction sufficiently clear - avoiding the introduction of subscripts.

Now, for the extension of this basic account to the other classes of objects.

As regards organisms, Evnine’s proposal is based on the claim that their generation involves elements analogous to those involved in the production of artifacts. This similarity is then assumed to guarantee similar results as those obtained in the case of artifacts - namely, that the process of generation determines the kinds to which the newly created organisms belong, and consequently their essences and identity conditions. If we try to spell out the details, we find that in the case of an animal, for instance, its generation will typically involve the fecundation of an egg by a sperm. Evnine suggests considering the sperm and egg here as performing the role of the agents in a process of production, and also as providing the matter (namely, their own matter) out of which the new organism is to be produced. On the other hand, the functions of the different parts of an organism correspond, in the case of organisms, to the teleological element provided, in the case of artifacts, by the intention of the producer.

Finally, as regards natural non-organic objects (namely, things such as rocks, planets, atoms, the solar system, etc.), Evnine resorts to a fictionalist account: the idea is that our talk about them should be interpreted as a sort of pretense, which does not commit us to their existence. There is a role reserved here, anyway, for the hylomorphic account that has been offered for the paradigmatic case of artifacts, given that, according to the proposal, the fiction that would guide our talk of, say, mountains, in a situation in which there are only some rocks arranged “mountain-wise”, is one to the effect that the rocks have been so arranged by a demiurge with the intention of producing a mountain.

3. SOME WORRIES CONCERNING EVNINE’S VIEWS.

I think Evnine’s proposal faces important problems: his arguments, both for the claim that the conditions he spells out for the production of artifacts are sufficient to account for their existence and their having determinate essences and identity conditions, and for the further claim that this account could be extended to explain similar features of organisms and (fictionally) of natural non-organic objects are not, in my view, particularly persuasive. I begin by making first some remarks on these applications of his views on artifacts to other sorts of objects, and then go on to consider some problems I find with his treatment of the paradigmatic case of artifacts.

3.1 The coming into being of organisms

Let us grant for the moment that Evnine’s account of artifacts is correct. Recall that the main idea in this account is that an artifact comes to exist, and to have its peculiar essence and identity conditions, because it has been produced by an agent that does the relevant labour with the intention of making an object of a certain kind. A further aspect of the proposal that will be relevant for the discussion that follows is that when they are made in this way artifacts are also assigned some functions, some of them idiosyncratic (that is, pertaining to specific individuals), some kind-dependent (that is, possessed as a consequence of belonging to a given kind). As we already mentioned, the idea Evnine explores and defends in Chapter 6 of the book, devoted to organisms, is that the production of artifacts and the generation of organisms involve elements that work in both cases in such analogous ways as to make it reasonable to expect similar results: given that in both of them we find some initial matter, an act of making with its agents, and functions associated with the kind, and that these circumstances have been supposed to explain the existence and identity conditions of artifacts, then we could also expect them to be sufficient to explain the existence and identity conditions of organisms.

Let’s go now more into the details. This is how Evnine describes the agents and the act of making involved in the generation of organisms:

As regards matter, on the other hand, the proposal is that the sperm and the egg involved in an act of fertilization contribute their own matter as the matter of the newly produced organism. This is not really analogous to, but rather contrasts with, what happens in the case of (objectual) artifacts, in which the matter is typically external to the agent. Anyway, even though this amounts to a disanalogy, I think nothing important hinges on it.

Now, as the author’s own scare quotes in the previous passage seem to indicate, the idea that a sperm or an egg might be taken as agents in an act of creation seems certainly odd (he admits this in remarks on p. 177). We should remember here that, according to Evnine’s analysis of artifactual production, two elements are required to bring into existence, and provide definite identity conditions to, the object produced: (a) a sortal concept, a representation of the kind of object to be produced, which also specifies the object’s existence and identity conditions; and (b) the intention to produce such an object, a feature of the creative act necessary for the agent to effectively impose the identity conditions encoded in the concept to some piece of matter.

Now, it is not so easy to identify elements corresponding to both of these aspects in the act of fertilization “performed” by sperm and egg. The details of Evnine’s account seem to me far from clear here, but his proposal seems to consist in understanding (a) in terms of the notion of function, and (b) in terms of the notion of selection, as these have been elaborated in etiological accounts of functions and teleological explanations, according to some philosophical interpretations of the theory of evolution. Now, the idea of selection, understood in evolutionary terms, is perhaps suitable enough to perform a role similar to that of intention; indeed, both notions seem to belong to a family that include also ideas like preference or choice (which may be understood also metaphorically).⁶ 6 The case of conventions may be seen as an interesting intermediate case between cases of intentional choice and cases of natural biological selection, which may perhaps strengthen the idea that they all belong to a single, broader class. Indeed, we may take conventional practices as practices that we conserve (and so “are selected”, where this is very similar to “are chosen”) because it serves to solve coordination problems in a community (cf. Lewis 1969; I found useful here a remark in García-Carpintero 2016, p. 101). Again, as we saw in the case of matter, the parallel is not that close: indeed, while in the artifactual case the intention is located in (the mind of) the producer, the mechanism of selection is not located in the agents of the act of fertilization (but, presumably, in some aspects of the past history of the trait selected); but, again, this seems to be inessential. What is more problematic, though, is the issue of what might play, in the case of the generation of organisms, the role that concepts play in the production of artifacts. This is because the notion of a function, the one Evnine relies on here, seems to behave very differently in the two cases.

The first point to notice here is that kinds and functions are assigned very different roles in connection with artifacts and organisms. In the case of artifacts, Evnine defends a view according to which “[m]embership in the given kind is not explained by an object’s having the associated function; [...] [r]ather, membership in the kind explains the having of the function” (p. 122). This is connected to the idea that the concept of a kind is richer, that is, more definite and informative, than the notion of having a function - for instance, there may be different kinds (stool, chair, armchair; glass, goblet, cup, mug) associated with what seems essentially the same function.⁷ 7 Of course, there is an issue as to how to individuate functions - one could have a very “fragile” conception, according to which very small differences regarding the specific way in which a function (in a broader sense) is fulfilled makes them different. In any case, it is interesting to note that Evnine himself endorses a coarse-grained conception of functions, according to which this would not be the case. The notion of kind would then seem to be prior, in this case, to the notion of function. On the other hand, there seems to be a reverse priority in the case of organisms: given that what is selected is a function, the notion of a kind should then be accounted for on the basis of the functions (or purposes) of certain traits, with functions identified on the basis of the traits giving rise to some specific effects (which may nonetheless not be present in some cases).⁸ 8 It should be noted that, following Millikan, Evnine focuses on a notion of function in so far as it is equivalent to that of a purpose, not to that of functioning as (cf. Millikan 1989, p. 20).

Now, the real problem here is not so much that there is this difference in explanatory priority but, rather, that it is not clear whether it would be possible to explain that something belongs to a kind on the basis of its having a certain function. According to the etiological explanation, the functions of traits are identified with the effects those traits produced in the past, and which partly explain their having been selected and reproduced. Functions are accounted for, therefore, on the basis of certain causal patterns.⁹ 9 In this explanation Evnine mainly follows the account of Neander 1991, which in turn exhibits strong affinities to that given in Millikan 1989. But it should be clear that the terms connected in such patterns are unable to explain what kinds the objects involved in them belong to. This is so because, in the first place, what causal patterns connect are sufficiently natural properties of objects (assuming, for instance, a standard account of causal laws such as Armstrong’s (1983ARMSTRONG, D. What is a Law of Nature?, Cambridge: CUP, 1983.)). But now, the properties involved in causal necessitations (such as having a certain mass, having a certain charge, etc.) are typically not kinds, that is, they are not associated with determinate identity and persistence conditions. And so it is to be expected that functions, if they are to be accounted for in terms of some past effects, and therefore in terms of non-kind properties, cannot by themselves determine what kinds the things having those functions belong to. This latter point fits nicely with the already mentioned circumstance that a kind is, typically, much more specific than a function, so that, even though a trait should be selected because of its having a certain function, its having that function would not settle what kind the trait (or the thing it characterizes) belongs to. But, of course, it is only kinds, not functions or other non-kind properties, that are associated with identity and persistence conditions, and with modal properties.

Besides, I’d like to mention a further problem with Evnine’s proposal, one which would affect his view even if we accepted that functions could in general perform the role he expects from them. This is connected with another difference we can find between the cases of artifacts and organisms, namely, that while functions are usually ascribed to artifacts as wholes, in the case of organisms they are rather assigned to some of their traits or parts, and not (at least in ordinary cases) to whole organisms. This has consequences for the role that functions might play in an explanation of the existence and identity criteria of the objects characterized by them - and, in particular, in the explanation of what makes the different parts constitute a unified object. So, while Evnine remarks that “the unity of the artifact is imposed ‘from above’”, namely, because the function of the whole artifact makes the different parts belong to a single whole, in so far as their functions contribute to the fulfilment of that of the whole artifact, he also notices that “the unity of an organism is emergent ‘from below’” (p. 180), in so far as an organism is understood as a whole made up of interconnected organs and organic parts “associated with more or less tightly conjoined co-evolved sets of functions” (p. 181). So, for instance, an animal gets its unity from the fact that it involves, say, among other things, a heart and quantity of blood, the function of each of which is connected to that of the other, and then it also involves some other parts similarly interrelated, so that the organism is a sum of tightly functionally interconnected parts and traits.

Now, such a difference seems relevant for the issue of whether functions, so understood, could play for organisms the role in individuation that concepts play for artifacts. Indeed, the alleged analogy seems to break down here. The main problem is that there seems to be no definite answer to the question of whether some organs or organic tissues with interrelated functions do or do not belong to a single organism. So, for instance, there are functional connections between the abdomen and the antennae of a particular ant, but there are just as tight connections between the ant as a whole and some other ants in the same ant nest: they communicate, for instance, using some pheromones, the cognitive state of one of them therefore modifying the behaviour of the other. Indeed, in so far as just functional interdependence is concerned, it seems to me that the relation between the ants is not so different from the relation between, say, different organs in a single mammal, when how one of them works depends on information that some other organ sends through hormones along the bloodstream. Certainly, there are some differences between the two cases: in one case the spatial connection is tighter than in the other. But this has nothing to do with functions, which, according to Evnine’s official proposal, should by themselves be sufficient to account for the identity and individuation conditions of objects. Something which, if our remarks here are correct, they are not.

What lesson should we draw from the foregoing remarks? It seems to me that the difficulties we have just mentioned point to a single, deeper problem, namely, that in the case of organisms there seems to be nothing corresponding to the “ideal” element so important in Evnine’s account of the individuation of artifacts. Recall that this “ideal” element, provided by the concept guiding the labour of the producer, was introduced to provide definite boundaries (both local, temporal and modal) to artifactual objects. But there is nothing able to perform this task in the case of the production of organisms: we are told here about traits and their effects, functions or purposes, but nothing there, as already mentioned, seems able to provide the objects produced with individuation and identity conditions.

Now, Evnine seems to be aware of some of the problems we have just mentioned. In connection to the second one, he says:

He takes this to imply that “there are thus likely to be quite dramatic cases of vagueness as to the boundaries of organisms on the view advocated here” (p. 182). But in my view this amounts to an understatement of the problems the view faces: what happens in this case with organisms is not just an example of the general fact that “all complex entities [...] are subject to vagueness” (p. 181). The latter seems to be what happens when, say, we have a definite idea of what a mountain is, but also acknowledge that it is indeterminate where it begins or ends. But in cases such as those we just described, it seems that the indeterminacy is of a much more radical sort: it seems we just don’t have a clue what the kinds themselves could be - not just where the boundaries of some of their tokens would be.

It would seem, then, that, if the remarks in this subsection are on the right track, Evnine’s account of how artifacts get their identity conditions cannot really be extended in a satisfactory way to organisms.

3.2 The existence of natural, non-organic objects

Evnine also intends his account of the existence and identity conditions of artifacts to extend, in a way, to natural non-organic objects, such as planets, stones and molecules. I say “in a way” because he will not defend the view that such objects exist as a consequence of their production being in some way similar to that of artifacts (as he argued in the case of organisms), but just that we may be nonetheless justified in talking about them as we do - that is, as if they did exist. Accordingly, he deploys, in the chapter devoted to this class of objects, two lines of argument: on the one hand, he argues against the existence of such natural non-organic objects; on the other hand, he argues in favour of taking a fictionalist attitude towards our discourse about these entities, an argument that will involve a fictional use of the explanation given for artifacts.

3.2.1 Against the existence of natural non-organic objects

The argument against the existence of natural non-organic objects is developed, in turn, in two stages. First, some considerations are offered in order to show that such objects don’t exist, and then some arguments are given against views that do admit them.

On the first issue, Evnine is surprisingly brief: he first claims that there only seem to be two ways in which an object distinct from its matter may come into existence: either by being made with some specific creative intentions, as in the case of artifacts; or by being an organism made up of organs with interrelated functions as a result of natural selection. But, given that nothing of that sort seems to be going on in the (alleged) coming into being of planets or molecules, we are forced to conclude that nothing new, nothing distinct from their matter, comes to exist when that matter is arranged in the relevant way - either planet-wise, or molecule-wise, etc.

Now, it certainly seems to be the case that natural non-organic objects do not exhibit these features that would explain, according to Evnine’s view, their existence as objects in their own right. But, first of all, it is not clear why we should admit that these are the only ways in which composite material objects may come to exist. On the other hand, it should anyway be stressed that a view that denies the existence of things like molecules and mountains is committed to a strongly revisionary claim: not only do we usually accept that such objects exist, but it is even reasonable to suppose that they behave like hylomorphic compounds. Of course, the charge of being revisionary is no knock-down argument against a view. Nonetheless, the fact that our intuitions seem to conflict with it should certainly be counted as a drawback. One might suggest that a possible move for Evnine here would be to say that his position just expresses that, in his view, intuitions of a rather general nature concerning the conditions for composition should rank higher, when doing metaphysics, than our case by case common-sense views about which objects there are. But even if he had a good argument for holding this (which in any case is not provided), it seems to me this is not a position Evnine would be comfortable with: after all, compatibility with common-sense views does seem to rank high for him, as both his reasons for rejecting universalism (namely, its commitment to “extraordinary” objects; cf. p. 196), and his very effort to provide a view of the nature of material objects that would allow us to include among them artifacts and organisms seem to show. To the extent, then, that compliance with common-sense is taken as an important desideratum for accounts of material objects, the fact that his view implies the rejection of mountains and molecules should then be considered as counting against it (though, of course, this may be counterbalanced by virtues in other areas).

As already mentioned, Evnine tries to strengthen his case by also arguing against some of the views that admit natural non-organic objects, in particular those he characterizes as universalist, conceptualist and minimalist views. I won’t spend much time here on his treatment of universalism, both because some of the points he makes (for instance, the one mentioned in the previous paragraph, that the view allows us to accept ordinary objects only at the cost of burdening us with unwelcome extraordinary objects) have already been extensively discussed in the literature, and also because I mostly agree with what he says in this connection. But I do find what he says about conceptualism and minimalism as deserving further discussion.

A first point to note in this connection concerns the very characterization of, and distinction between, these two approaches, which have been sometimes confused in the literature. What Evnine calls ‘conceptualism’ is a sort of view like the one Karen Bennett discusses in her paper on the “grounding problem” for coincident objects (Bennett 2004BENNETT, K. “Spatio-temporal Coincidence and the Grounding Problem”, Philosophical Studies 118: 339-271, 2004.).¹⁰ 10 The grounding problem (which I discuss more extensively in the last section of this paper) challenges the different views on material objects (in particular those allowing for coincident objects) to provide an account of how the kind and other sortal properties of objects are grounded (hence the name) in other more basic properties. On the other hand, the view presented in the quote from Bennett’s paper is just one of two different views that she considers as possible ways of cashing out the idea that the sortal properties of objects could be grounded in the subjects’ attitude towards them. The other one she discusses very briefly, as it is clearly a non-starter. There Bennett considers a solution to this problem based on a view according to which

As it is characterized here (and elsewhere in her article), this proposal seems rather underspecified, given that what she says could be understood in different ways. Evnine takes this view as one according to which objects are in general produced as artifacts are supposed to be according to his own proposal - with the only difference that, according to conceptualism, the labour side of the process just consists in the agent’s taking things as falling under some concepts, or, as he also puts it, in merely “casting a glance” on them. In his own words:

According to conceptualism, then, something would count as an object of a certain definite kind if there is (at least) a minimal worldly elaboration of some matter, where this minimal elaboration consists in an agent (or perhaps also in a community of agents) considering that matter as satisfying some conditions encoded in a concept.

I think Evnine’s main objection to this way of understanding conceptualism in fact undermines it. It is the objection, a rather usual one against suspected anti-realist views, based on the observation that, intuitively, there could have been (and we think that there have been) mountains and planets and molecules well before there were human beings (or other concept possessors) around to bring them about - in the minimal sense in which this takes place according to the view. The problem with this line of response, though, is that it seems ineffective against other possible ways of cashing out the notion that it is “our conceptualizing activity” which “call[s] things into existence”, ways that do seem to me to give rise to rather promising accounts of material objects. I think in particular of views according to which the conceptualization is not understood as a “worldly” activity of making, like the making of an artifact, but rather as a process by which we as subjects provide some sort of order to our experience - process which could be understood either in a Kantian, in a response-dependent, or even in some other way, perhaps incorporating social or institutional dimensions (cf. Einheuser 2011EINHEUSER, I. “Towards a Conceptualist Solution of the Grounding Problem”, Nous 45: 300-14, 2011., Kriegel 2008KRIEGEL, U. “Composition as a Secondary Quality”, Pacific Philosophical Quarterly 89: 359-83, 2008, 2012_____ “Kantian Monism”, Philosophical Papers 41: 23-56, 2012 and to some extent also Goswick 2018GOSWICK, D. “The Hard Question for Hylomorphism”, Metaphysics 1: 52-62, 2018 for proposals of this general sort). To the extent that these are indeed reasonable views, then we could have alternative ways of accounting for objects’ kind-membership in terms of our conceptualizing activity that would be immune to Evnine’s considerations - that they are so immune may be shown with an argument similar to the one I present in connection with minimalism in the next paragraph.¹¹ 11 I think positions of this general kind are indeed able solve this problem (see the next paragraph in the text), even though some of their proponents do not think so: for instance, Goswick (2018, p 58) thinks her view falls victim to the argument that appeals to the fact that, intuitively, there were, say, dinosaurs, before there were human beings around to respond in appropriate ways to their matter. I don’t see why it would be problematic to think that dinosaurs are constituted by our present responses to past chunks of matter. In any case, a fuller discussion of this issue falls outside the scope of the present paper.

As just said, then, minimalism, the other view Evnine discusses, also escapes this objection. This is a view, also known as the “easy ontology” approach, that has been defended by Amie Thomasson for almost two decades now, and which also appeals to concepts (hence the usual confusion), but in a very different way - here concepts aren’t involved in making the objects, but just in recognizing them (cf. Thomasson 2001THOMASSON, A. “Ontological Minimalism”, American Philosophical Quarterly 38: 319-331, 2001., 2007_____ Ordinary Objects, Oxford: OUP, 2007, 2015_____ Ontology Made Easy, Oxford: OUP, 2015). This is how this is supposed to work: we have concepts of objects belonging to some kinds, which are associated with what she calls “application” and “co-application” conditions, so that the fulfilment of the application conditions in a certain region (of space-time) implies that an object of the relevant kind exists there, while the fulfilment of the relevant application conditions in a pair of such regions, and of the relevant co-application conditions by the pair itself, implies that a single object of the relevant kind is present in both locations.

Now, consider a mountain (allegedly) existing in the past, at a moment at which there were no mammals yet. Can the minimalist (or, for that matter, a Kantian or response-dependent conceptualist) say that there was a mountain then and there? Of course she can: if that space-time region in the past contained some rocks arranged mountain-wise, then it follows from the rules that guide the use of our concepts (or from the way in which they organize our experience) that there was a mountain there. Concepts enter here not as part of a worldly activity of making, but as part of our way of understanding (or organizing) what is given to us in experience (no matter how indirectly, as in this case).

Evnine is right, then, in not using this objection against minimalism. But the objection he does use against this view seems to me ineffective. It’s as follows: it seems reasonable to assume that one should not use the concept of a kind of thing when stating the conditions under which objects of that kind exist. But now, he goes on, when the minimalist states the co-application conditions for a given term, she violates that requirement, given that she states these conditions, say, for a term such as ‘mountain’, by stating what it means to say that something ‘is the same mountain as’ something else. I cannot see that this is a fatal objection to minimalism, though: it is true that informally we could describe in that way what we are doing when we are giving the co-application conditions of ‘mountain’; but, so far as I can see, it is not really necessary to describe them in this way: one could (and, indeed, should) describe those conditions in more basic terms, terms mentioning some of their parts and their spatio-temporal connections, without involving the notion of mountain - and I really can’t see why that couldn’t be done.¹² 12 The point I’m making here is (intended to be) similar to one Searle has made when trying to show that characterizing something to be money just in case it is “thought of, or regarded as, or believed to be money” is not indeed a circular characterization - namely, because ‘money’ in the latter phrase is “just a placeholder for the linguistic articulation of all [the social] practices” that makes something be money (cf. Searle 1995, p. 52). Similarly, then, ‘being the same mountain’ could be a placeholder of a more definite characterization in terms not involving ‘mountain’. So that, if this is all he has to say against minimalism, it doesn’t seem to me that he has given any strong reason to reject this sort of view. So that, so far, two different conceptions of material objects that accept natural non-organic objects have not been convincingly shown to be wrong.¹³ 13 Evnine criticizes with some more detail Thomasson’s view in his 2016b; she responds to his arguments in her 2015, pp 221-9 (although the focus in that discussion is a bit different).

3.2.2 Fictionalism

As regards Evnine’s positive fictionalist proposal for our talk about natural non-organic objects, its main idea is that we could still go on talking as if there were mountains and molecules in so far as we take such talk as governed by a the following fiction F:

Now, I have some misgivings concerning this proposal. The main problem is that it is not at all clear to me just what role F is supposed to play in the proposed fictionalist interpretation of our talk about natural non-organic objects - and, in particular, whether any real contribution is made by the fictional story about the making of artifacts. Evnine says that F, or its conventional acceptance, is what underlies, or governs, our fictional talk about natural non-organic objects. But it is not clear to me what “underlying fiction” is supposed to mean here; in particular, if it was taken to mean what Yablo, from which he borrows the term, understands by it, then it doesn’t seem to me that F could indeed play that role. This is because Yablo takes this notion to refer, not so much to a sort of narrative (as F is), but rather to something like a function or rule, understood in close analogy to language rules, which act rather “as medium and not message” (2001YABLO, S. “Go Figure. A Path through Fictionalism”, Midwest Studies in Philosophy 25: 72-102, 2001., p. 76). So, just as the rules of English make it the case that, say, ‘Snow is white’ is true under certain circumstances (namely, just in case snow is white), a fiction (such as, in his example, the number fiction) makes it the case that, say, ‘The number of apostles is twelve’ is fictional (i.e., acceptable in a game of make-believe) just if there are twelve apostles. Let me quote with some more detail Yablo on “governing fictions”:

Notice that here the “governing fiction” is just a rule that makes it fictional that some entity exists when some conditions obtain. But it remains silent on what further properties these objects possess, or on why or how they did come to exist, or even on why it would be reasonable to talk about them (which is the role apparently given to F). Something similar happens in the example Evnine discusses to illustrate his views on fiction, and which he seems to take as paradigmatic, namely, the one discussed by Walton (1990WALTON, K. Mimesis as Make-Believe, Cambridge (MA): Harvard UP , 1990, pp. 21-4) of some children who, in a game of make-believe, talk of some tree stumps as if they were bears. In such a context, a sentence like ‘there are three bears’ would then have a literal content, that there are three bears, and a distinct real content, that there are three stumps, that would make it appropriate to assert it (and would make it fictional, or “true in the fiction”). But it’s interesting to notice here that the children’s pretense is clearly not in need of any justification: the children just act as if the stumps were bears, but they do not necessarily do so on the basis of a narrative that would explain, justify, or otherwise make sense of, their doing so: the game just consists in taking one thing as if it was another. It seems to me to follow from what we have just said that a natural way to apply the notion of a governing fiction to the case of natural non-organic objects would only involve having some sort of correlation to the effect that, whenever there are, say, some rocks arranged mountain-wise, it is fictional that there is a mountain. But no reference seems to be required to its being thought of as an artifact, or to its having been created in one way or another - briefly: we need no rationale for engaging in the fiction. So that F, in so far as it pretends to offer such a rationale, would not be playing the role of a guiding fiction, in Yablo’s terms.

Now, taking into account that F is a narrative, and that it is offered as a guide for what is appropriate to take as fictional, one might think that the story F could alternatively work here more as a prop, just as a book does in typical literary fictions, and the stumps did in the children’s game. To take an example in which a story (a series of inscriptions) works as a prop, the text of The Adventures of Huckleberry Finn makes it fictional (or true according to the fiction) that Huck met Jim, which is also the literal content of a statement according to which Huck met Jim (which is accepted only fictionally), while its real content (accepted without qualification) would be just that the text written by Mark Twain states that Huck met Jim. It isn’t clear, though, that this is the role that Evnine has in mind for F: props are typically material objects, ones which have a definite role in the practice and which participants are usually aware of as having that role; while, even though F is indeed mentioned as being “conventionally accepted” when giving rise to the fiction of natural non-organic objects, it doesn’t really seem to work as a prop. On the contrary, it seems that the rocks arranged mountain-wise are that which performs a role similar to that of the tree stumps in the children’s game; and while we could not rule out that the rocks plus the story are what jointly plays such a role, which would then make up something like a combined or “hybrid” prop, that seems to be a departure from orthodoxy that would certainly deserve special mention and justification - which are not provided. Besides, F, as presented by Evnine, is rather the content of a story than a story by itself - its textual peculiarities do not seem to be relevant for its role in the proposal.

This last comment suggests one further possible role for F in this context, namely, as the content of the simulation. To clarify this: when we engage in a game of make-believe, we engage in a pretense guided by some real objects or facts (the props), on the basis of which we pretend something else is the case (although there is of course a systematic connection between real and fictional content, as we just saw).¹⁴ 14 Cf. Yablo 2001, pp. 76 ff. for stressing that a guiding fiction has the nature of a function, and Thomasson 2015, ch. 5, for stressing the importance of their being a significant difference between the two terms of the relation. So, in the case of natural non-organic objects, we find some rocks arranged mountain-wise, and pretend that there is a mountain, something made by a demiurgic creator with some definite intentions (namely, that of making a mountain) which would then (mytically) explain their being so organized. That is, just as in the stump-bear game the stump is fictionally taken as a bear, in this case the rocks arranged mountain-wise are taken as an artifactual mountain. Now, what we have just said already points to what’s problematic with this proposal: it’s simply no part of what we think about mountains, of what we are trying to account for in a fictionalist account, that mountains are artifacts produced by some sort of demiurge. This seems to show that the proposal would not be adequate if presented as an hermeneutic fictionalist account of our discourse.¹⁵ 15 I rely here on a distinction usually made between hermeneutic fictionalism (where the fictionalist account is presented as an interpretation of our actual thought and talk about some kinds of entities) and revolutionary fictionalism (where the fictionalist story is proposed as a way of making sense of a discourse that would be problematic, from an ontological point of view, if taken at face value). But it would be equally problematic to suppose that Evnine’s proposal is being advanced as a revolutionary fictionalist one: he assumes that for his proposal to be adequate we should be able to say, for instance, that “There are mountains, and they are formed not by the activity of a demiurgic creator but by the entirely natural and unguided collisions of tectonic plates” (p. 204), and he tries to make sense of this by appealing to Yablo’s notion of Relative Reflexive Fictionalism. But this will not do. What that sort of fictionalism allows us to do is to refer to the fictional objects not only to say something about the real world (as we say something about rocks when we apparently refer to mountains), but also to speak of the mountains themselves, either in a disengaged manner (for instance, to say “in the ontology room”, that there are no such things), or in an engaged manner, to describe them as they appear in the fiction. But in none of these attitudes would it be appropriate to say that mountains are the product of natural forces (not in the disengaged manner, because in this case what’s appropriate to say is that there are no mountains; nor in the engaged manner, because according to F they are created by a demiurge). I’d rather say that in such a discourse what we really have is, again, a way of saying things about the real world (namely, that rocks came to be arranged as they are by the causes there mentioned). But the story of the demiurgic creation plays no role in this sort of discourse.

I take it to follow from the above discussion, then, on the one hand, that it is not at all clear how exactly Evnine’s fictionalist proposal should be understood. Indeed, we were unable to assign a proper function to the “guiding fiction” F, and especially to the mention of creative intentions of a demiurge there - as we saw, that story is not required, either to justify our fictional practice, as a prop, or as part of the fictive content of the pretence. On the other hand, Evnine’s arguments against minimalism and (some versions of) conceptualism do not seem convincing. So that, if what was said here is correct, minimalism, some conceptualisms and a non-demiurgic fictionalism seem all in a better position than the specific fictionalist proposal advanced in the book to make sense of our talk of natural non-organic objects.

3.3 Artifacts and creative intentions.

In the last two sections I discussed Evnine’s attempts to extend to organisms and natural non-organic objects the account of hylomorphic complex objects he gives for the paradigmatic case of artifacts, and found both of them flawed. It’s now time to consider the account he gives of the central case of artifacts.

The main question that motivates his treatment of artifacts is: what explains that there are objects that are distinct from (i.e., not identical with) their own matter? As we saw, his answer to this question is rather idiosyncratic: what makes an object (specifically, an artifact) distinct from its matter is not the presence of a further component, a form, as standard hylomorphic views would have it, but the matter’s having been worked on by someone with the intention that it should become an object of a certain definite kind. In his words: “Artifacts come into being, are made, by someone’s working with certain intentions on some material that becomes the artifact’s matter” (p. 67). Discussing a case in which a child builds a sand castle, he elaborates further as follows:

As we saw, this so called “imposition of form” involves two different aspects: on the one hand, the agent’s intention of imposing a form on some matter and, on the other hand, the concept of a form or kind, which determines the kind to which the object created is to belong.

The account seems reasonable, and it is certainly plausible that something very much like what is described in the passage above must take place whenever an artifact is made. But what exactly does this talk of imposing the producer’s mind, or a concept, on matter, mean? How should the metaphor be spelled out? To what extent does what he there describes differ from its just being the case that the matter comes to satisfy (perhaps as a result of its being worked on intentionally) the application conditions associated with the concept of a thing of that kind? It’s difficult for me to find any such difference, and so it’s not very clear in my view what gain in explanatory power we could get from this suggestion that the maker imposes, or impresses, a conceptual element on matter. Let me spell out why I think so.

The first worry I want to mention is epistemological, and is connected with the seemingly reasonable desideratum that one’s metaphysical views concerning a certain area of reality should be consistent with an explanation of how it would be possible for us to gain knowledge of that area of reality (if, indeed, that metaphysical account is to be an account of precisely that section of reality we claim to be knowing). I think this requirement involves two more specific ones: on the one hand, that the account given of the nature of the objects involved should be such as to make them accessible (either directly or indirectly) from the point of view of people allegedly knowing something about them; on the other hand, that the epistemic practices that we actually take to provide reliable knowledge of such objects might really be sensitive to what happens with them.¹⁶ 16 This is what Peacocke has called the “integration challenge” for any area of reality (Peacocke 1999), a specific instance of which had been first discussed by Benacerraf for the case of mathematical objects (cf. Benacerraf 1973)

Now, in connection with the latter aspect, it seems that we are reliable enough in tracking what kinds the objects we encounter in experience belong to. What seems to guide us in this activity is our sense of whether what we find in experience satisfies or not the conditions, encoded in our concepts, required for belonging to a definite kind (i.e, the concepts’ application conditions). Of course, such conditions may include having been made intentionally. For instance, it is part of our understanding of what a statue is that it should have been made with certain specific intentions (for instance, that it should be intended to be a work of art). And this should be no problem, as we are also pretty good at recognizing when something has been done or not with certain intentions. On the other hand, it seems that nothing else in the original act by which the agent makes the object, in particular, none of her more specific intentions, including those concerning the kind to which the object made is to belong, plays any role in our classification. Let me elaborate a bit on this.

Suppose then, for instance, that an artist is trying to make not just a regular statue, but a very peculiar one, one that would be destroyed if it lost a very specific shape that it has - for instance, suppose she is trying to make something like van Inwagen’s gollyswoggle. Now, when we see the artifact that she has made, we would naturally take it to be just a sculpture; whatever else the artist has tried to accomplish when “impressing her intentions” on the piece of matter, this is something we are unable to recover - indeed, Evnine seems to assume as much, when he mentions the need for making explicit an “artist’s sanction”, namely, an explicit statement of the conditions under which the work would cease to exist, if the work of art being made does not belong to socially recognized kinds. Or, to take another, possibly more controversial, example: suppose that a glass artisan intends to make just a fancy sort of glass, but that, unbeknownst to her, what she eventually makes corresponds to a more specific type also recognized in (parts of) our community - say, a margarita glass. If the glass the agent made satisfies the conditions associated with ‘margarita glass’, then, it might be argued, it will count as such, even if that is not what the agent intended (I take this example to be more controversial because it is not clear to me whether intending something to belong to a very specific kind is necessary for the object to belong to that specific kind of thing (even assuming that it is necessary that she should intend it to belong the more generic one it also belongs to); it certainly seems insufficient, as the previous example showed).

In general, what these examples show, in my view, is that what guides us in the determination of the kind to which an object belongs is not so much the mental attitude of the agent when she produced it (because, of course, in many cases we are not in a position to know much about it), but the kinds that are socially recognized and whose conditions the object appears to satisfy. So, in so far as that which determines the kind to which an object belongs should be something accessible to us (and this concerns the first constraint mentioned above), and something that our actual classificatory practices should be responsive to, even indirectly (our second constraint), then it would seem that the intentions of the original producers should not be assigned any role here. On the other hand, what does seem to play a relevant role, and conform to the other requirements stated, are the social conventions specifying the conditions that make something belong to a specific kind. So that kind-membership seems to be much more a matter of objects complying with certain (socially recognized) requirements than with some individual’s intentions.¹⁷ 17 There seems to be an interesting parallel between the failure of the original intentions to fix the kind to which an artifact belongs, and some counterexamples (for instance, Evans’ (1973) example of swapped babies) that aim to expose similar failures of the original acts of baptism in securing the later reference of proper names, for some simple versions of the causal-historical account of the reference of such terms (as first suggested in Kripke 1980). Unfortunately, I don’t have space to elaborate on this parallel here.

I want to conclude my discussion by considering a rather more specific problem, connected in this case with Evnine’s treatment of the so called “grounding problem” - namely, the problem of whether, and eventually how, sortal and other identity-related properties of material objects might be grounded on their more basic physical properties (following Karen Bennett (2004)BENNETT, K. “Spatio-temporal Coincidence and the Grounding Problem”, Philosophical Studies 118: 339-271, 2004., I’ll be calling these two classes of properties ‘sortalish’ and ‘non-sortalish’ properties, respectively.)¹⁸ 18 I have in mind the distinction between “non-sortalish” and “sortalish” properties, as Bennett (2004) characterizes them (the first are those that appear in scientific explanations, the latter are connected with the objects’ belonging to certain kinds). The distinction is presupposed in all the literature on coincident material objects and the “grounding problem” for their kind and modal properties (cf. inter aliaFine 2003 for a similar distinction and discussion). We could indeed distinguish two varieties of the problem: a general one, which arises for any view about material objects, and a more specific one that has been used in particular against views admitting coincident objects. The general version of the problem takes as its point of departure the (apparently reasonable) assumption that the sortalish properties of objects should somehow be grounded on their non-sortalish, basic physical ones, and then goes on to point out that it is difficult to see how that could be the case, challenging then any view on material objects to give a plausible account of how such grounding takes place. But there is a more specific form of the problem, which is particularly directed against views (such as Evnine’s) that admit coincident objects - such as, say, a piece of clay and the statue made from it. In this form, the problem consists in just drawing attention to the fact that the general grounding problem just mentioned seems devastating for such views, because in this case what would require an explanation is how the same basic, physical properties, which are shared by the allegedly coincident objects, could give rise to distinct sortal (and “sortalish”) properties, as the statue and the clay are in fact distinguished from each other by having different such properties (for instance, it is usually argued that the statue and the piece of clay from which it is made are numerically distinct because, say, while the former is a statue, and would not survive being squashed, the latter is a piece of clay, and would survive being squashed).

Evnine has then to provide a solution to this problem, at least to the more specific one, if his view is to be tenable (as Bennett points out, the more general problem is a problem for everybody, and so perhaps not legitimately wielded against specific views). One striking characteristic of his answer to this problem is that it offers very different explanations of how the statue and the lump of clay get their sortalish properties, something he justifies because, as he says, those two objects “belong to different ontological levels” (p. 82). According to his view, then, while the statue gets its sortalish properties “because it is made by the imposition of a concept that determines those sortalish properties” (p. 81), the lump of clay derives them, on the contrary, from its own non-sortalish properties. Now, this would be an adequate solution to the problem if it could be made to work, but the proposal seems to me to face important difficulties.

One first problem is that we never get an explanation of how the non-sortalish properties of the lump of clay (or of any other such quantity or lump) would be able to ground its sortalish ones - and, therefore, it just sidesteps the general grounding problem. But a deeper problem is that a solution such as the one sketched above seems inconsistent with some of the main motivations for an amorphic hylomorphism. Recall that this is a view about composite objects that gives a prominent place to the explanation of how they come to belong to definite kinds, something that is important because this is in turn what makes it possible to explain their having determinate identity and persistence conditions. Now, as we saw, this explanation of kind-membership makes reference to the fact that artifacts, in particular, are made with creative intentions with “ideal” contents (or something to the same effect), which are impressed on the matter and which specify, as a result, the kind to which the newly created object is to belong. It seems to follow from this explanation, as presented, that the obtaining of such circumstances is a necessary condition for material objects to belong to definite kinds. Now, Evnine also admits, on the other hand, that things like lumps of clay belong to definite kinds (namely, in this case, to the kind lump of clay), as a result of which they possess their own determinate identity and persistence conditions. He in fact says, for instance, that “[t]he kind lump of clay implies persistence conditions that are independent of shape (the same lump can endure drastic reshaping)” (pp. 81-2). But, at the same time, he denies that lumps of clay are hylomorphic compounds, and that their coming into being should be explained as that of artifacts: “[The lump of clay] itself does not belong to an artifactual kind and is not, on my view, a hylomorphically complex object at all” (p. 81). Now, these views strike me as inconsistent with one another, so that Evnine should give up at least one of them. Besides, they also seems to suggest that the conditions for kind-membership put forward by the theory are perhaps too demanding, as we seem disposed to describe things as belonging to kinds even when these conditions are not satisfied.

In any case, it is also interesting to note that Evnine’s general picture does not really force him to adopt such a view as regards pieces of matter. His reasons for doing so seem to include the apparent fact that such objects do not have a metabolism, as hylomorphic compounds do, and also that they have not been created with specific intentions, as artifacts are. But I’m not convinced of either of these claims. On the one hand, it seems that a lump of clay may be thought of as belonging to a definite kind, and therefore as an hylomorphically complex object with a characteristic metabolism: the fact that it belongs to a kind of object that could not survive through a change in its matter does not imply that it has no metabolism, but only that it has a peculiar, limiting case of metabolism, namely, one according to which the material parts have to remain the same (in the same way in which a constant function is still a function). On the other hand, it is not clear to me that lumps of clay, of the sort used in making clay statues, are not artifacts: indeed, making them seem to be involved in the process of making a statue, so that I don’t see why they should not be considered as artifacts.¹⁹ 19 C. Sutton, who defends positions somewhat similar to those of Evnine, considers that lumps are constituted in a similar way as statues, acquiring their identity conditions by way of the agent’s intentions (cf. her 2012, p. 712). If what I have just said is right, Evnine could then be in a position to give a parallel account of how both pieces of clay and statues get their sortalish properties, and avoid the specific problem I have just sketched.

4. CONCLUSIONS

I think the upshot of the above discussion is more or less as follows: on the one hand (and this has not been taken up in the discussion because I agree with Evnine on this point), there seem to be good reasons to suppose that an account of composition and kind-membership that respects (at least most of) our pre-theoretic intuitions concerning which objects there are and how they behave should give a role to human intentions and concepts, as they are directed to those objects. But, on the other hand, the above remarks seem to suggest that this reference to intentions and concepts is misplaced in Evnine’s account: they should not be located in the producer’s mind (most of the problems I found arise from this claim), but rather in our social conventions, as they specify the conditions for kind membership. That is, they should be located rather on the receptive side - that of the users and perceivers of the objects. These are claims that point to an alternative, conceptualist view of material objects that I cannot attempt to present here, but which I hope to be able to develop in future work.²⁰ 20 I’d like to thank Simon Evnine and Gonzalo Rodriguez-Pereyra for comments on an earlier version of this paper.

REFERENCES

Publication Dates

History