SUTTON’S SOLUTION TO THE GROUNDING PROBLEM

Abstract I critically discuss Sutton’s 2012 attempt to solve the so-called “grounding problem” for coincident objects, namely, the difficulty of explaining how such objects, such as a statue and the lump of clay from which it was made, can have distinct kind and modal properties, even though they share the same proper parts and basic microphysical properties. Sutton bases her solution on an account of the extrinsic composition of the different sorts of objects involved in such cases - in particular, artefacts, organisms and persons. I show that the accounts she gives of their composition are flawed, and that her proposal therefore does not solve the grounding problem.


Introduction
In this paper I discuss the solution to the so-called "grounding problem" for coincident objects presented by C. S. Sutton in a recent paper (Sutton 2012), and I try to show that her attempted solution actually fails. 1 The interest of this exercise, though, seems to me to go beyond just showing the flaws of a particular proposal: given that she makes use of resources that seem promising for solving this problem, such as the reference to intentions or biological species, trying to expose some of the difficulties that arise in connection to their use seem to me to offer general lessons that future proposals should take into account.But first, in any case, let me present briefly the problem and Sutton's solution.
The grounding problem has been wielded against views that admit coincident objects, that is, numerically distinct objects that occupy the same spatial (or spatiotemporal) region -or, alternatively, that have (for some time) exactly the same proper parts -views defended by philosophers described as "pluralists", "colocationists" or "multithingers".The standard example of coincident objects, also used as a paradigm case by Sutton, is that of a statue (call it "Goliath") and the piece of clay ("Lumpl") from which it is made.Of course -and this is why the view is controversial -we could also, and perhaps more naturally, say that in this situation we have only one thing, which might then be described either as a statue or alternatively as a piece of clay -as "monists" or "one-thingers" think.Given that this more economic description is available to us, why should one prefer a pluralist account of the situation?Pluralists typically try to motivate their view with an argument based on Leibniz's Law, namely, that the statue and the lump must be two distinct things because they instantiate different properties: Goliath is (essentially) a statue, would not survive being squashed into a ball, but would survive losing a tiny part; while Lumpl, in contrast, is (essentially) a lump, would survive being squashed into a ball, and would not survive the loss of a tiny part.Of course, one may ask in this regardand this is precisely the challenge that the grounding problem poses to pluralists -whether it would indeed be possible for things such as Goliath and Lumpl, which seem to share all their proper parts, and all their basic, microphysical properties, to differ in the other properties just mentioned -properties which, connected as they are with the sorts of things they are, could be called their "sortalish" properties. 2 As can be seen, the problem of accounting for these properties is particularly pressing for the pluralist, but it is a problem for which any view of material objects should, in any case, provide an answer.In its most general form, then, the problem is whether "sortalish" properties, such as belonging to a certain kind or having certain persistence conditions, are grounded in "nonsortalish" ones -paradigmatically, microphysical properties -and if so, how.
2 I borrow the term "sortalish" from Bennett 2004.In a paper that has strongly influenced the recent debate, Karen Bennett distinguishes three distinct possible solutions to this problem, which she describes as the Foundationalist, Conceptualist and Primitivist solutions.According to Foundationalism, the sortalish properties of ordinary material objects are supposed to be grounded -in ways yet to be made clear -in some of the non-sortalish, either intrinsic or relational, properties of objects.Yet, as she convincingly argues, it is difficult to see how this could be possible, as there seems to be no way around the fact that coincident objects (or the parts that compose them at some level of decomposition) necessarily share both their intrinsic and relational non-sortalish properties. 3According to the alternative Conceptualist approach, on the other hand, the sortalish properties are supposed to be grounded, not in the basic objective properties of the coincident objects, but rather in our attitudes towards them.(This is a view that Bennett also rejects, on the basis of some elaborate arguments that I cannot discuss here.) 4Finally, according to the Primitivist approach, the sortalish properties of objects are not supposed to be derived from any other properties or relations of those objects, but should rather be understood as primitive features of them.(Let's add, just for the record, that even though this sort of response may seem initially 3 Some more recent attempts at providing foundationalist accounts of sortalish properties try to circumvent some of the problems discussed by Bennett (see, for instance, Saenz 2015 and Jago 2016).I think these views are still problematic, but I cannot discuss them here.
unattractive, Bennett ends up claiming that a version of this view would be the best available option.)I think that Bennett's taxonomy of possible solutions provides a reliable tool for assessing the different answers to the grounding problem, as it does indeed seem to exhaust all the basic available options.5If this is right, then one would expect Sutton's proposal to fall under one or another of these categories.Yet, in my view, one of the signs that her proposal is problematic is precisely that it does not fit well into any of them.Rather, it seems to me that she presents her proposal as if it could keep the most attractive aspects of both conceptualism and foundationalism, without the drawbacks of either.However, as I will try to show, the view that she actually puts forward incorporates elements that do not fit well with one another, with the result that it cannot actually provide an adequate solution to the grounding problem.Let us turn now, then, to the details of Sutton's proposal.

Sutton's general strategy
It will be convenient to distinguish two steps in Sutton's argument for her attempted solution to the grounding problem.In the first step, she distinguishes two different ways in which an object might be composed.An object may be either extrinsically composed, in case its parts compose it in virtue of some relations that those parts bear to some other things; or intrinsically composed, in case the object is composed, but not extrinsically composed (i.e. in case it is composed in virtue of its parts' intrinsic properties and their relations to one another).
In the second step, Sutton applies this general distinction to show that the grounding problem presents no real threat to colocationism.This would be so because, as she argues, all alleged cases of coincident objects are such that at most one of them is intrinsically composed, while the others are composed on the basis of their parts standing in different extrinsic relations to distinct objects, these different relations explaining, in turn, why the objects have different sortalish properties.This is the general scheme that Sutton applies to the different classes of coincident objects she considers, which she distinguishes according to the different sorts of extrinsic relations allegedly grounding the extrinsic composition of the objects involved.She distinguishes three different classes of cases: (1) The colocation of artefacts with lumps (or masses) of matter.
(2) The colocation of natural non-organic objects with lumps (or masses) of matter.
(3) The colocation of human beings (or, in general, organisms) with lumps of tissue -and, in some cases, with persons.This is how she argues that none of these cases poses a threat to the colocationist.
As regards class 1, she argues that the artefact (for instance, a statue) and the matter it is made of (for instance, a lump of clay) are both extrinsically composed.According to her account, while the statue comes to be composed because its material parts (at some level of decomposition) are related to human intentions concerning being a statue, in the case of the lump, its being composed depends on those same parts being related to human intentions concerning being a lump.These different relations that the parts stand in are then supposed to ground their different sortalish properties -and therefore explain why one of them is a statue while the other is a lump of clay.
Class 2 is exemplified by a certain diamond that coincides with a lump (or a mass) of carbon atoms.As she has already established (in her discussion of class 1 cases) that lumps are extrinsically composed, it follows that the grounding problem is already solved in this case.Even if the diamond turned out to be intrinsically composed -which seems plausible to her, although she does not fully endorse the view -both objects would still have differences that could account for their distinct sortalish properties.
Finally, as regards class 3 cases, she argues that all the objects that might be taken to coincide in such situations are relationally composed.Lumps of tissue are composed of their parts because they are related to intentions concerning lumps; human beings (and other organisms) are composed of their parts because they are related to the species human being (or to other species in the case of other organisms); while persons come to exist as such because of the extrinsic relations they bear to the objects that they are aware of. 6his short discussion already shows that the key to understanding how material objects could coincide, and how coincident objects may acquire their different sortalish properties, lies in Sutton's explanation of extrinsic composition.Indeed, her account of intrinsic composition is so sketchy and tentative -and, as she mentions, she is not even committed to its having any instances -that in what follows I think we can safely ignore it. 7Instead, I will discuss in some detail her explanation of the different sorts of extrinsic composition that she makes use of in her account of these cases.8

The composition of artefacts
The case that she describes in most detail concerns the composition of artefacts, which she seems to take as the paradigmatic (or at least the clearest) example of extrinsic composition.According to her, composition takes place in such cases when the objects that will become, say, the statue's parts, satisfy our intentions concerning what it is to be a statue.In her words: Our intentions that create artefacts are intentions about the artefact kind itself.The composing parts or simples stand in a relation to our intention when those composing parts satisfy our criteria about what it is to be a thing of kind x. (2012: 709) Later she is more explicit -making clear, moreover, that she treats lumps in the same way as she treats artefacts: 9 Our intentions are about what it is to be a lump or what it is to be a statue.The things that are soon to be parts of the lump and parts of the 9 It is debatable whether lumps should be considered as artefacts (Sutton treats them as if they were artefacts, without actually claiming that they are), while Evnine, for instance, claims that they are not artefacts; see his 2016: 81).Perhaps a reasonable middle way would be to claim that some lumps could be artefacts (for instance, I might make a lump of clay by separating it from a larger quantity, in order to proceed to make a statue out of it), even if lump is not an artefact kind (one all of whose instances must be artefactual).Lump would then be like path, also a kind whose instances may or may not be artefacts.(Cf.Thomasson 2007a: 58  and Evnine 2016: 67 for discussion of the distinction between essentially and non-essentially artefactual kinds.)statue stand in the right relation to our intentions about lumps and statues.[…] The parts satisfy our intentions by being organized in the right way.(2012: 710) Now, Sutton's use of this account of the composition of artefacts to solve the grounding problem seems to me problematic for at least two reasons.On the one hand, it is not at all clear how she thinks that the relations to intentions she mentions might be distinguished so as to explain how distinct, coincident objects may in fact emerge as a result (let's call this "the distinction problem").On the other hand, it is not always clear which intentions she has in mind in her explanation, as she seems to refer to different things in different steps of her argument.As we'll see, the two problems are not unconnected.
Let's consider, to present what I called the distinction problem, a slightly more detailed account of her solution to the grounding problem, as applied to the specific case of a statue and a lump of clay: [A]ll that matters to solving the grounding problem is to show that for any group of colocated objects, at most one is intrinsically composed.Any number of extrinsically composed objects could be composed of the same parts, for each object comes to exist in virtue of a different extrinsic relation that the parts have to something or things external.[…] The statue and the lump […] differ in their kinds, which are grounded in relations that the parts stand in to (1) human intentions about statues and (2) human relations about lumps, respectively.(2012: 709) On the face of it, Sutton seems to refer here to a situation in which some parts -those that will constitute the statue -are related to a certain intention concerning statues, while those same parts, which will also constitute the lump, are also related to a certain intention concerning lumps.But it is not at all clear how we should conceive of these relations, so that they may give rise to distinct objects.Her description of these relations as constituting the "supervenience base" of the resulting artefacts may suggest that she intends her account to be of a foundationalist sort, namely, one according to which the distinct objective properties and relations that the parts that will make up the objects instantiate ground, just by themselves, the objects' further (and distinct) sortalish properties.Now, this seems problematic.In effect, given that the parts of the statue and the parts of the lump are (at least at a certain level of decomposition) exactly the same, they cannot fail to instantiate the same properties and relations, and, in particular, to be related to both of the intentions mentioned in the example.So, given that their parts necessarily share all the relations they stand in, being involved in those relations cannot be what explains why one of the objects that the parts compose is, say, a lump (and not a statue) while the other is a statue (and not a lump). 10is cannot be, then, the whole of Sutton's proposal.More needs to be said if it is to provide a satisfactory account.Now, when trying to figure out in which direction the view might be developed, it becomes relevant to notice that her talk of intentions seems rather equivocal.She seems to waver indeed between two distinct notions, that suggest two importantly different precisifications of her proposal.In this connection, it is important to notice, in the first place, that it is not clear whose intentions are invoked in this context, or what exactly they are directed at.And, besides, it is also unclear whether she is referring to (aspects of) mental acts or states of some specific agents (for instance, the authors of the artefacts), or rather, as she sometimes seems to do, to the content of some more generally valid, conceptual norms.
The latter understanding is suggested by the fact that she often talks, rather vaguely, of "our", or more generally of "human", intentions, which seems besides to fit well with the idea, also expressed in this connection, that these intentions are directed towards the kinds to which the artefactual objects belong (and not to specific, particular artefacts).The idea seems to be that our interests and purposes make it reasonable for us to produce and use objects of certain kinds, which are adapted to serve those interests and purposes, and which eventually give rise to conventions concerning the conditions under which objects satisfying those purposes (and perhaps some further properties) may be said to exist and to belong to the associated kind (no matter whether they have been manufactured specifically to fulfil those purposes).Finally, this understanding of intentions seems also connected with the shift she later makes from talk about intentions to talk about conventions -for instance, when she claims that "lump is conventional too" (the title of her section 2.5, p. 712).Let us call intentions in this sense, even though probably a misnomer, C-intentions -where "C" stands for community.Now, her discussion sometimes also seems to point at another sort of connection between artefacts and intentions, the one most often made between these notions, namely, that what is usually required for an object to be an artefactual object is that there should be, not just general (C-)intentions directed towards some kind (and its associated purpose), but also intentions to the effect that some specific parts of matter should compose an object of such-and-such kind; that is, that there should be some intentional, concrete mental events or states that actually guide the agent's work on those parts in order to produce the artefact as a result -and which we may call M-intentions ("M" for maker). 11The case of lumps precisely provides an example that could help us to see the difference between these two ways in which intentions may be present in the production of objects.On the one hand, it may be granted that we (say, a certain community) have interests that manifest themselves in C-intentions concerning the kind lump, which then give rise to conventions regarding the conditions under which something belongs to it.But, on the other hand, we also see that, at least for some lumps, the further M-intention oriented towards, and guiding the production of, a specific object may be lacking, with the result that the lumps produced in such circumstances would not be artefacts, because they are not made with this M-intention that their parts should compose lumps.We obtain the same result if we consider that the distinction we make between, say, 11 Cf.for instance the view of Hilpinen (1993: 156-7) who, after saying that "[a]n object o is an artifact if and only if o has an author", affirms: "I take a person A to be an author of an object o only if it is intentionally produced by A under some description of the object."Even though I refer in the main text to "some specific parts of matter", I do not mean to refer rigidly to the quantity of matter that actually makes up the relevant artefact, but rather to any quantity of matter that is worked on in the act of production.(I tend to agree with Evnine's view (2016: 86-96), according to which it is the act by which it is made, and not the matter from which it is made, that is essential to being a certain artefact.)boulders and pebbles seems very much conventional and dependent on our interests (there are C-intentions in relation to them), without this entailing that boulders and pebbles are in general artefactual (they are not produced as a consequence of M-intentions). 12ow, these remarks are relevant in the present context because they suggest two different ways of developing Sutton's proposal so as to provide a solution to the distinction problem -both of which are in my view, and for different reasons, flawed.On the one hand, if we focus on C-intentions, we may take her view as expressing a conceptualist account of composition and the determination of sortalish properties.According to this precisification of the proposal, we would have two distinct but coincident objects whenever the composing parts of both are related to two different C-intentions.This would allow us to solve the distinction problem, because these intentions are just conceptual norms according to which objects of a certain kind exist in a region in so far as the application and coapplication conditions associated to the corresponding concept are satisfied in that region (as it is represented by us).
12 I take the distinction between dependence on M-intentions and dependence on C-intentions, as applied to this case, to be very similar to a distinction that has been made in the moral realism literature between mind-dependence and attitude-dependence (cf.for instance Street 2006, fn. 1).The only difference is that, while mind-dependence is a relatively broad notion, that includes any sort of dependence on some mental act or event (and so is broader than dependence on M-intentions), attitude-dependence requires specifically dependence on the perspective from which any object or event (of any kind: mental or non-mental) is represented in a specific situation by a subject of experience (which is equivalent to our C-intentions).
Besides, there seems no reason why two such norms might not be satisfied simultaneously by the same parts, giving therefore rise to two objects in virtue of their satisfying two distinct C-intentions. 13t is worth noticing that some of Sutton's remarks seem to endorse this interpretation of her account of composition, as when she says, for instance: [A]rtefacts such as statues come into existence with and because of those sortal properties that depend on a relation between the human mind and composing parts.(2012: 718) 14 Applied to this particular case, the idea would then be that the statue and the lump are two coincident objects composed of the same material parts (at the relevant level of decomposition) in virtue of the fact that those parts satisfy the conditions encoded both in the concept of a statue and the concept of a lump.These relations to distinct concepts allow us then to explain how it is that we end up representing 13 This is how Thomasson (2007b: 38-44) accounts for the truth of claims asserting the existence of ordinary objects, namely, in terms of the application conditions of the sortal concepts (or terms) involved being satisfied; these concepts are then further associated with co-application conditions, that state the conditions under which a member of the sort counts as the same individual. 14In the same vein, she says on p. 707 that her view of artefacts "overlaps" Thomasson's view, but that her own take on other kinds of objects, such as organisms and persons, "avoids a conventionalism or conceptualism that seems appropriate in the case of artefacts" (ibid.), which suggests that she does regard Thomasson's (and, therefore, her own) view of artefacts as "conventionalist or conceptualist", at least to a certain extent.that region as including two distinct, though coincident, objects. 15It is important to notice, though, that we may obtain here two distinct composite objects because the relations their parts bear to the two concepts are relations in a sense different from the one in which they might be related, in a "worldly" sense, so to speak, to some actual state or event (which is the sense discussed above in which the parts cannot fail to be in exactly the same relations); in this case, the parts are related to two distinct ways of representing them as composed and, therefore, the resulting composite objects appear as distinct only from the perspective of someone that represents them that way.Now, although, in my view, there is much to be said in favour of an approach to the composition of ordinary objects along these lines, it would be problematic to suppose that this is the view that Sutton is advocating in her paper.The main problem seems to be that, if the validity of some relevant conceptual norms is sufficient to ground composition and kind-membership facts for artefacts, and if such norms are appealed to on the basis of the usual reasons given by conceptualists, namely, that objects in themselves don't have "built-in persistence criteria", 16 it is not clear why we should restrict the account just to artefacts instead of applying it across the board, for all sorts of ordinary objects.Indeed, this conclusion seems also to follow from the fact that the intentions that are specifically relevant for the composition of artefacts, M-intentions, do not play any role in such an explanation (over and above any role they might have as involved in the relevant conceptual norms).Whatever reason we might have thought we had to restrict the account to artefacts seems to depend on a confusion between the two kinds of intention discussed, which seems also behind Sutton's inclusion of a non-artefact, a lump, in the relevant class of objects, misleadingly describing the class as one of "artefacts".Finally, we may also notice that, if adopting a conceptualist view were the only way to solve the distinction problem, then it would follow that the non-conceptualist solutions to this problem for organisms and persons, presented later in the paper, would also be flawed.It would seem worthwhile, therefore, to see whether a nonconceptualist interpretation of Sutton's remarks is workable.Now, if, on the other hand, we settle on M-intentions as those relevant for the composition of artefacts, and we understand them (as we should) as (involved in) particular mental states in the minds of the relevant agents, we are again stuck with the fact that it seems impossible for the very same parts to be related to one such state without also being related to the other one.The only way I can think of to get around this problem would be to adopt a primitivist stance in relation to which intention is relevant for the composition of which object: one should take as a primitive fact, for instance, that the supervenience base for the composition of the statue involves the relation of its parts to the intention concerning statues without including their relation to the intention concerning lumps (and the same should be the case, mutatis mutandis, for the case of the lump).Now, I don't have a knock-down argument against the view one gets when one interprets Sutton's remarks in this way, but it does seem to me unsatisfactory.On the one hand, it seems to incur the cost of assuming primitive facts where one would like to find some sort of explanation -indeed, Sutton's own appeal to intentions in this context seems to express this desire for further explanation.On the other hand, it is not clear what the specific mention of intentions actually contributes to such a primitivist proposal.In fact, it seems that we could just as well claim (and that would be more economical) that it is a primitive fact that these parts compose both a and a lump, or maybe we could take them to compose one or the other according to whether they are (primitively) related to this or that function or form, so that the mention of intentions seems indeed to be idle.
It seems to me to follow from the discussion in this section, then, that Sutton's proposal for solving the grounding problem for artefacts is unconvincing: if we take her talk of intentions as referring to the validity of some conceptual norms, then it seems unclear why we should restrict the proposal to artefacts and other social kinds, which in the context of her paper, moreover, risks making her later efforts to give non-conceptualist explanations of composition pointless; if, on the other hand, we take it as referring to states involved in the concrete production of artefacts, then they cannot play any explanatory role.Either way, we don't get a convincing explanation of how the differences of the sortalish properties of artefacts are grounded.

The composition of organisms
Let us now consider how Sutton deals with the other alleged cases of extrinsic composition, beginning with the case of organisms.
The case she discusses in more detail in this connection is that concerning the composition of human beings, but it is clear from what she says that she intends her discussion to apply to all kinds of organisms (on p. 724, for instance, she takes her views to be about "biological-kind membership" in general, and she also briefly discusses, on p. 725, the property of being a dog).She also seems to intend her account, even more clearly than the account of artefacts discussed above, to be of a relational foundationalist sort, except that in this case the relation allegedly grounding the fact that some parts compose an object belonging to a specific biological kind is the relation that those parts bear to the relevant species.For instance, concerning the case of a dog she says: To be a dog, the composing parts [i.e.those that will compose the dog] must stand in the appropriate relation to members of the Canis familiaris species.(2012: 724) Correspondingly, she construes the kind-membership properties of organisms, such as being a human being, as extrinsic properties of them: The property of being a human being is an extrinsic property because the property arises from a relation between an object, such as Aristotle's body,17 and a species, in this case Homo sapiens, which depends at least in part on something outside Aristotle's body.(2012: 724)   The proposal then is that something is an organism of a certain kind because its parts stand in a certain relation to the species (or, more precisely, in the case of the species-asindividuals view that she favours, to all the other members of the species besides the specific organism at issue). 18Can this account explain the composition and kind-membership facts of organisms, and thus help to solve the grounding problem for them?Notice, first of all, that the main difficulty that we faced in connection with Sutton's account of artefacts also affects her view of organisms, and in particular, her view of how they would be able to coincide with lumps of tissue.Given that the same parts that are supposed to compose the lump of tissue are also supposed to compose a certain organism that belongs to a biological kind in virtue of their relation to a species, it is clear that the parts of one of them cannot fail to instantiate the same properties, and stand in the same relations to everything else, as the parts of the other.It seems to follow that we could not end up grounding the composition of two different composite objects unless we found some way of restricting the relevant supervenience base, so as to include only one of the relevant relations in each case.Besides, contrary to what happened in the previous case involving artefacts, it doesn't even seem initially plausible to suppose Sutton tries to solve this problem along conceptualist lines, as she doesn't mention intentions in her explanation.It seems then that we should settle for an interpretation according to which it is just primitive that some relations are involved in the explanation 18 At some point in her paper, she considers how her proposal would go if we accepted different views on the nature of species.In what follows, and mainly for reasons of space, I will discuss only the problems that arise if we assume what is in fact her preferred view of species, according to which, following Hull's (1978)  proposal, they should be treated as individuals.
of the organisms' constitution, which is as unsatisfactory as in the previous case.
Further problems arise when we try to figure out precisely how the relation to a species might ground the composition facts.A first point to notice is that Sutton never describes in any detail what this relation would consist in, besides claiming that it is the relation that accounts for the organism's composition.Indeed, once we come to think of it, it is not clear how a relation to some other purely material object (namely, one with no intentional or representational properties) might help us to explain the kind properties and persistence conditions that an organism acquires.In any case, the proposal seems to face more specific problems.Remember that one of its main ideas is that the organism's parts' relations to the species would determine the organism's compositional facts (and the ensuing identity conditions).Now, if we take a look at what philosophers of biology have to say about species considered as individuals, it is clear that for them species face a similar problem concerning what their own identity conditions are, that is, how their own boundaries are to be drawn.For instance, according to Hull, species should be considered as populations, that is, as "segments of the phylogenetic tree" (Hull 1978: 349); but then we face, in relation to them, the same individuation problems as with the organisms themselves: If species are historical entities, then the same sort of considerations which apply in the individuation of organisms should also apply to them, and they do […].[H]ow can a gradually evolving lineage be divided in an objective, nonarbitrary way? (1978: 346-7)  It would be beyond the scope of this paper to deal with this issue in any detail, but it seems clear that, if we adopt Sutton's preferred view according to which species are individuals, that is, historical entities in the very same sense in which organisms are individuals, then it is only to be expected that individuating them faces exactly the same difficulties that we face when trying to individuate (or account for the composition or kind-membership facts of) individual organisms.To the extent, then, that species themselves must have determinate identity conditions in order to account for the corresponding organisms' identity conditions (as seems reasonable to assume, and as Sutton herself seems to assume, given her view of species), species thus conceived would therefore be ill-suited to explain the organisms' composition facts.Related circularity worries also appear if we consider that, if species are indeed historical populations made up of organisms, then it would seem that the existence and individuation conditions of those populations cannot be established independently of, and prior to, the existence and individuation conditions of the individual organisms that make them up.And that would put in jeopardy Sutton's attempt to ground the individuation of organisms on the basis of the individuation of their species, understood as populations.
For the various reasons just given, then, Sutton does not seem to have solved the grounding problem as it arises for organisms.

The composition of persons
The last case of extrinsic composition that Sutton considers is that of persons.She argues that they are extrinsically composed because, according to Baker's account, which she follows here, being a person seems to require having a first-person perspective, which in turn requires being a conscious subject that is aware of something other than itself: Are persons extrinsically composed?Baker argues convincingly that, if to be a person is to have a first-person perspective, then personhood is relational, and essentially so.(2012: 721)   After referring to a passage in Baker's book in which she presents this view (the reference is to Baker 2000: 69, 72-9), Sutton adds: In this case, the person is extrinsically composed, and we have an explanation of how the properties of the person are grounded in more than just the parts of the person.(2012: 721; my emphasis) Now, in my view, there is an important gap between what is said in the first of these passages and the conclusion that is drawn in the second, according to which "the person is extrinsically composed".This is so because, although some of the properties of persons may be grounded in these extrinsic relations, it does not follow that all of them (as Sutton seems to assume by saying in this context "the properties of the person"), in particular their composition and identity conditions, are so grounded.In other words: what makes a state or an event one of a personal character is not necessarily the same as what determines the composition facts in which it's involved, namely, those that establish which personal states belong to which self or person.It seems to me, though, that the relational facts that Baker appeals to in her explanation are only relevant to the issue of what makes a state or an event a personal one in the first place, while they leave untouched the individuation facts in which those states may be involved.Here again is Sutton reporting on Baker: Baker claims that having a first-person perspective requires a contrast class or individual.In order to have the perspective of 'I', there must exist something that I can distinguish from myself.(2012: 721) As I suggested above, it does indeed seem plausible that for something to have a first-person perspective (and to be selfconscious, something that Baker takes to be "the key to being a person"; 2000: 60) it is necessary for there to be something else of which a subject should be aware.But, again, it is unclear how this could help to explain the composition of persons.There are a couple of points that seem to support this sceptical stance.On the one hand, it is important to notice that, while Sutton's accounts of the composition of artefacts and organisms relied on their relations to some specific kinds of entities (intentions or species), which in turn had at least some initial plausibility for determining the object's identity conditions, the present account of persons appeals to relations that might involve anything whatsoever, with the only restriction being that it should be distinct from (the parts that will constitute) the person.This seems to be an important disanalogy, as in the previous cases it was the nature of the object involved in the relation that was supposed to ground the composition facts.It is in fact unclear just how a relation to anything whatsoever could be expected to ground a person's identity conditions.On the other hand, it should also be stressed that Baker's account of personhood, which only requires being aware of something distinct, seems perfectly compatible with different views of personal identity -i.e.concerning the conditions under which a pair of experiences of a personal character belong, or fail to belong, to the very same person.For instance, the account of personal states or events just given is compatible not only with a standard psychological continuity view, according to which someone counts as the same person as a child forty years earlier if she retains some memories of that time, but also with an animalist view according to which some experiences belong to the same person just in case they take place within the same organism, and also with a revisionary view such as Galen Strawson's, according to which selves are much more short-lived than the human beings to which they're connected (cf.Strawson  2009: 9).In brief, it seems that conditions for something to have a first-person perspective, and therefore for having a personal character (which should be understood as application conditions for the concept of a person), do not determine the individuation and persistence conditions of persons (that is, the concept's coapplication conditions). 19But this is what we need if we are to solve the grounding problem for a case involving persons and human bodies, because this is what is required in order to explain the difference between the coincident objects, and this precisely what, if my above comments are correct, Sutton's account does not manage to supply. 2019 As already mentioned, the idea that sortal terms are semantically associated with application and coapplication conditions has been taken from Thomasson 2007b. 20 Given that something is extrinsically composed if and only if its composition is grounded at least in part in relations that the object's parts bear to some other things that do not overlap them, it may be argued that this relation to something else that something should have in order to be a person may still contribute to the person's composition facts, and that, therefore, my above

Conclusion
If my above remarks are on the right track, then, Sutton's attempts to solve the grounding problem for the different cases she considers are flawed, so that we can't find in her paper, in fact, any satisfactory solution to it.But the interest of the discussion seems to me to go beyond just finding out this, as there are at least two lessons that seem to follow from it.On the one hand, it seems that Bennett (2004) was right when she pointed out that relations to concrete objects and events are ill suited to ground composition facts.On the other hand, it would also seem to follow that, if we introduce concepts or other sorts of contentful states to provide the required individuation conditions, then composition and kind membership might take place, but only, as the conceptualist would have it, when considered from the perspective of someone for whom the relevant conceptual norms are valid.I take it, therefore, that the discussion above clearly suggests adopting a conceptualist view of composition.Developing in more detail such a view should anyway be left for another occasion. 21guments do not undermine Sutton's views after all.But it seems clear to me, in any case, that these awareness relations to some external things, whatever else they ground in the character of the resulting persons, do not in any way contribute to setting their spatio-temporal and modal boundaries.
21 I would like to thank José Tomás Alvarado Marambio, Sebastián Briceño, Santiago Erpen, Gonzalo Rodriguez-Pereyra, Carlo Rossi, Javier Vidal, the two referees for Manuscrito and, especially, Marta Campdelacreu and Catherine Sutton for comments on earlier versions of this paper.The research leading to this paper has been partly funded by the Ministry of Science and Innovation (Government of Spain) and the European Union (grant number PID2019-106420GA-100/AEI/10.13039/501100011033).