Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. Who’s Afraid of Non-Existent Manifestations? ABSTRACT – I shall defend in this paper the thesis that, if there are irreducible powers such as the power to produce a certain object (generative powers), then there are objects that do not exist and they are part of the fundamental level of the universe. Thus, generative powers come together with Meinongianism. After having clarified my argument, I shall examine and criticize Armstrong ’s attempt to reduce powers to other sorts of entities. Finally, I shall deal with five accounts of generative powers that are somehow alternative to Meinongianism and with some (more general) miscellaneous concerns about the truth of this doctrine. 1. Introduction According to many recent metaphysical accounts of powers1, powers are fundamental, irreducible entities2. Even if you do not believe that powers are the only sort of fundamental entities – or the only sort of fundamental properties -, as some philosophers such as Alexander Bird (2007) maintain, you could still believe that they are part of the basic ontological inventory of the universe, that God could not produce an exhaustive copy of our universe without reproducing powers and their instantiations. Moreover, it seems that powers are essentially individuated – among other – by their (possible) manifestations. Thus, the power to dissolve salt – that is seemingly possessed by water – is also essentially individuated by its (possible) manifestation, i.e., dissolving salt. That power is what it is – the power to dissolve salt – also in virtue of that (possible) manifestation. If it had had another (possible) manifestation, it would not have been that power, but the power to something else. If it had had no (possible) manifestation at all, it would not have been a power at all. Of course, some powers are associated with different (possible) manifestations in different (possible) circumstances. Yet, this does not imply that they are not essentially individuated also by those (possible) manifestations: perhaps, they are individuated by all those (possible) manifestations and all those (possible) circumstances. Anyway, I shall argue in this paper that, if you claim that powers are fundamental, irreducible entities and that they are essentially individuated also by their (possible) manifestations, you should also claim that there are non-existent objects. In this respect, I shall agree with David M. Armstrong on the thesis that a powers metaphysics is committed to the truth of Meinongianism, i.e., the doctrine according to which there are non-existent objects3. However, I shall disagree with him on the evaluation of this fact, since I do not believe that this constitutes a problem at all. In section 2 of this paper, I shall introduce my argument. In section 3, ) shall consider Armstrong’s analysis of seemingly true ascriptions of powers. In section 4, I shall examine five alternatives that are not committed to the truth of Meinongianism and, in section 5, I shall deal with some miscellaneous concerns about Meinongianism. ) take powers here as referring to powers, dispositions, capacities and propensities, because the distinctions between these sorts of entities are not relevant in this paper. See Choi, Fara (2012). 2 See, for example, Mumford (1998), (2004), Cartwright (1999), Ellis (2001), Molnar (2003), Bird (2007), Martin (2008), Mumford, Anjum (2011). 3 See, for example, Armstrong (1997: 79). 1 Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. 2. The Argument. Ascriptions of powers are often thought of as what is expressed by seemingly true statements such as (1) I can raise my arm (i.e., I have the power to raise my arm). If you think that powers are part of the basic ontological inventory of the universe and if you accept that my power to raise my arm cannot be reduced to other powers (possessed by me or by other entities), you are committed to the existence of my power to raise my arm or, more generally, of the power to raise one’s arm. Moreover, powers are often considered properties – particular or universal. Thus, if I have the power to raise my arm, it is either the case that there is a particular property such as my power to raise my arm – that characterizes or constitutes me4 – or that there is a universal property such the power to raise one’s arm – that is instantiated by me. The possession of a power does not imply its activation. It could be true that (1) and false that (2) I raise my arm. In principle, powers could remain unmanifested. Finally, I assume here that all powers are essentially individuated also by their (possible) manifestations. Together with the thesis that some powers could remain unmanifested, this implies that, nevertheless, those powers are essentially individuated by their merely possible manifestations, i.e., by manifestations that never exist or occur, even if they could have existed or occurred. The move towards Meinongianism is quite easy. I shall consider one particular case: the case of generative powers. Generative powers are powers to generate something. For example, a certain existing cell let me name it Cell- power to generate by mitosis two different cells (let me call them Cell- has the generative and Cell- . Perhaps, this power is never activated. Yet, Cell-1 could still possess this power, even without its activation. Of course, Cell-2 and Cell-3 cannot exist before the activation of that power, nor can they come into existence if that power is not activated. My argument runs as follows: (a) generative powers are fundamental, irreducible entities; (b) if it is true that (a), then there are non-existent objects and such objects are fundamental; (c) thus: there are non-existent objects and such objects are fundamental (from (a) and (b), by MP). 4 A particular property characterizes something iff it is one of its modes (see Lowe (2006)). Modes ontologically depend on their bearers . On the other hand, a particular property constitutes something iff it is one of the tropes that constitute that thing. In fact, tropes are particular properties that are more fundamental than ordinary objects and that constitute such objects in appropriate conditions (see Maurin (2013)). Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. I shall prove (a) dialectically in the next sections. However, let me consider the case in which Cell-1 generates Cell-2. If Cell-1 generates Cell-2, then it was at least metaphysically possible that Cell-1 generated Cell-25. If it was at least metaphysically possible that Cell-1 generated Cell-2, then it was either the case that Cell-1 had the power to generate Cell-2, or that it had the power to acquire the power to generate Cell-2 in relevant circumstances6. Thus, Cell-1 either had the power to generate Cell-2, or it had the power to acquire the power to generate Cell-2. In both cases, there was a power essentially involving a reference to Cell-2, even before Cell- ’s starting to exist. Moreover, it seems that Cell-1 actually had the power to generate Cell-2: that power only needed to be activated. Thus, there was a generative power, such as the power to generate Cell-2. Is this power ontologically fundamental? Can it be eliminated or reduced to anything else? I do not think that it can. In fact, as I shall argue in the next sections, such eliminations or reductions are inefficacious, since they do not preserve the truth of some ascriptions of powers across actual and possible situations or they do not succeed in getting rid of nonexistent objects, or they commit us to entities that are more problematic than non-existent objects. The truth of premise (b) can be demonstrated as follows. I shall work under the hypothesis that generative powers are fundamental, irreducible entities, i.e., the antecedent of the conditional that I aim at demonstrating. Powers are essentially individuated by their (possible) manifestations and – in turn – such manifestations need to be individuated. The existence of something that still does not exist7 – such as the existence of Cell-2 – is the (possible) manifestation of generative powers (such as the power to generate Cell-2). Yet, this (possible) manifestation is individuated only if something that still does not exist is individuated. Thus, generative powers are individuated only if their (possible) manifestations are individuated and such manifestations are individuated only if something that still does not exist is individuated. Thus, if generative powers are fundamental and irreducible entities (that obviously need to be individuated), there is something that still does not exist and that nevertheless contributes to the individuation of generative powers. Moreover, if generative powers are fundamental, irreducible entities, non-existent objects are fundamental too. In fact, it seems reasonable to claim that whatever contributes to the individuation of fundamental entities is fundamental too. Generative powers are fundamental entities. Thus, nonexistent objects are fundamental too. Thus, if generative powers are fundamental, irreducible entities, there are non-existent objects and such objects are fundamental too. I anticipate an objection here: if generative powers are fundamental entities, can they depend for their individuation on anything else, i.e., on non-existent objects? Either they depend on something else for their individuation, or they are fundamental, and there is nothing which is both fundamental and depends on something else for its individuation. Yet, I think that one could still maintain that there are 5 Of course, this metaphysical possibility could still be reduced to something else, not involving Cell-2, by those who claim that there are no non-existent objects such as Cell-2 before their coming to existence. 6 See Molnar (2003): 100-101. 7 And that perhaps will never exist. Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. certain sorts of fundamental entities that also, but not only depend on something else for their individuation. Of course, if God aimed at creating an exhaustive copy of our universe, He would have to copy those entities too – not only the entities on which they partially depend for their individuation. Powers belong to this category of fundamental entities. In fact, they also, but not only depend for their individuation on their (possible) manifestations – at least if we assume that being a power is a primitive and irreducible feature of powers. Thus, the generative power to generate Cell-2 also, but not only depends on Cell-2 for its individuation. In addition, its being a power does not depend on anything else. My argument can be criticized in four major ways. Either you demonstrate that (i) generative powers are not fundamental entities, or that (ii) they are fundamental entities which do not depend on nonexistent objects for their individuation, or that (iii) those non-existent objects on which they depend for their individuation wholly depend in turn on existent objects or properties, or that (iv) those nonexistent objects actually exist. Yet, before turning to these criticisms, ) shall consider Armstrong’s attempt to get rid of powers qua fundamental entities. 3. Armstrong vs. Powers. Armstrong’s concerns on the fundamentality of powers are also based on his refusal on Meinongianism. Anyway, he thinks that powers ascriptions in general should be reduced to something else. In other terms, powers ascriptions should be analysed into ascriptions of something else to objects and/or to properties. Here is the form of such an analysis: (powers) as a matter of metaphysical necessity, for every object, that object has a certain power p1 iff Φ, where Φ should substituted by the analysans – that does not have to mention powers. The left side of the equivalence that object has a certain power p1 is the analysandum. Metaphysical necessity is invoked in order to distinguish appropriate analyses of powers ascriptions from accidental regularities. Let me now assume that p1 stands for the power to produce an instantiation of a certain property ( (e.g., the property of being identical with Cell-2). Thus, PH will stand for the property of having the power to produce an instance of a certain property H. If Cell-1 does not actually produce Cell-2, such a power is unmanifested. Armstrong’s analysis primarily deals with unmanifested powers. In fact, unmanifested powers are more problematic than manifested ones, since they seemingly introduce in the realm of existence mere possibilities. On the other hand, manifested powers point towards existent manifestations: they are not fundamental entities and their analysis can be easily performed in terms of existent entities. Moreover, I shall assume that F is a variable ranging over properties that are not powers and that constitute the microstructures of objects having p1 – as long as their microstructures reveal what those objects are. G is a variable ranging over properties whose instantiation – together with the absence of the instantiation of other properties, at least in some cases – is nomologically sufficient to produce the Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. instantiation of (. J , K , etc. stand for the properties whose instantiation should be excluded in order for G to be nomologically sufficient to produce the instantiation of H. N stands for the relation of nomological necessitation between properties. Finally, ) shall assume that the variable x ranges over existing objects, ◊N and ⧠M are two modal operators, that respectively represent nomological possibility ( given certain laws of nature, it is possible that necessity. and the aforementioned metaphysical Armstrong (1997: 81-82) offers this analysis of powers ascriptions8: (arm.powers) ⧠M ∀x(PHx ↔ ∃F(Fx & ∃G N(F & G (& ∼J & ∼K & … ( & ◊NGx)) Informally: as a matter of metaphysical necessity, any object has the power to produce an instance of H iff there is a microstructural property F instantiated by that object and F and G (and, in case, the absence of J, of K, etc.) nomologically necessitate H and it is nomologically possible that that object instantiates G. The first conjunct of the analysans is the instantiation of a microstructural property, the second is a certain law of nature laws of natures are relations between universals, from Armstrong’s perspective9), the third is the nomological possibility of G’s instantiation. Here are some problems with this analysis10. Firstly, it seems that some powers can be associated with different microstructural properties: fragility, for example, is realized by vases, glasses, and so on, i.e., by objects having different microstructures. Thus, either one substitutes F with a disjunction of microstructural properties or s/he claims that there is a different power for each microstructural property. Yet, in the former case, disjunctions (or disjunctive properties) would turn out to be fundamental – or more fundamental than powers ascriptions –, while, in the latter case, one should abandon the idea that there is a universal power ascription – even if it seems that all the objects that have p1 with different microstructures have the same power. Secondly, different properties G could be associated with one and the same power. Moreover, such properties could in turn be associated with different negative clauses, excluding the instantiation of certain properties. Thus, the law of nature in the second conjunct could become much more complex, including further conjunctions and disjunctions of properties: N(F & ((G1 & ∼J) V (G2 & ∼K V … H. Thirdly, negative clauses are problematic, since they are identical with negations of properties. Yet, what is the negation of a property? Within a law of nature, the negation of a property cannot be the noninstantiation of that property in a certain situation, since laws of nature are relations between properties, and not between instances of properties. Thus, either the negation of a property is a negative 8 In this text, Armstrong does not actually talk of microstructural properties. This terminology is introduced, for example, in Armstrong, Martin, Place (2002: 41), by preserving the analysis of unmanifested powers suggested five years before. Moreover, Armstrong (in Armstrong, Martin, Place (2002: 39)) affirms that a disposition is a microstructure picked out via its causal role. 9 See Armstrong (1982). Moreover, Armstrong (2005) claims that laws of nature are relations between types of states of affairs. Anyway, for the sake of simplicity, I shall maintain here that they are relations between universals, since types of states of affairs are, in turn, universals. 10 For some of the difficulties that I shall briefly examine here, see Bird (2007: 18-42). Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. property of a property (e.g., the property J is such that it is not instantiated) or it is a negative property (e.g., the property non-J). As it is well known, both disjunctive and negative properties are problematic entities for Armstrong: they do not exist or, at best, they are not fundamental11. Thus, a different analysis of powers ascriptions should be provided without invoking them. However, even if we were inclined to accept such properties, some difficulties would still remain. First of all, there are some powers (finkish powers12) that can be lost by objects after the instantiation of the relevant G, so that H is not instantiated (i.e., the manifestation does not occur). Such powers should be analysed by including within the relevant law of nature in the second conjunct of the analysans the negations of all the conditions that could produce their loss. This means that further negations of properties – constituting further negative clauses – should be added13. Yet, such negative clauses are introduced only because their corresponding positive properties produce the loss of the finkish power. Thus, each negative clause is there only because it is compatible with the existence of p1 after the instantiation of G and before the instantiation of H, while the corresponding positive property is such that is incompatible with the existence of p1 after the instantiation of G. This explanation of negative clauses provides a bad analysis for powers ascriptions, since it reintroduces powers in the analysans and it makes it the case that the analysans is what it is only in virtue of certain facts involving the powers to be analysed. Secondly, there is a more general problem with conjunctive properties. Armstrong (1978: 30-42) accepts conjunctive universals within his ontology. Thus, the first relatum of the nomological necessitation relation in the second conjunct of the analysans is a really complex conjunctive property. Yet, it is either the case that every conjunction of properties gives rise to a complex conjunctive property that should be included within our ontology and/or that should be invoked as a fundamental entity, or that only some conjunctions of properties give rise to complex conjunctive properties to be included within our ontology and/or to be invoked as fundamental entities. However, accepting the first horn of this dilemma, too many complex conjunctive properties turn out to be included within our ontology and/or to be fundamental, and such properties cannot be dispensed with by only accepting their conjuncts. In fact, only conjunctive properties stand in the nomological necessitation relation with H – and not their conjuncts. Yet, following the second horn of the dilemma, one still has to explain why certain conjunctions of properties give rise to conjunctive properties that figure in laws of nature about powers ascriptions, while other conjunctions of properties do not give rise to conjunctive properties. This explanation cannot mention p1. Yet, it seems that only p1 provides an adequate explanation: complex conjunctive properties are there only because their conjuncts – put together – are somehow 11 See Armstrong (1978: 19-29) and (2004: 54-67). See Martin (2008: 12-23). 13 Thus, there will be two different kinds of negative clauses: those that only prevent the non-obtaining of the manifestation and those that prevent the non-obtaining of the manifestation by preventing the loss of the relevant power after the instantiation of G. 12 Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. associated with p1 and, more precisely, with the activation of p1. Of course, one could accept the second horn and give no explanation: it is a primitive truth that only some conjunctions of properties give rise to conjunctive properties. Yet, this would imply the substitution of something (a power) with something else which is more complex and perhaps problematic a conjunctive property within one’s ontology and/or at the fundamental level of the universe. In sum, Armstrong’s project – as it is expressed by (arm.powers) – seemingly fails. Perhaps, there is something unanalysable about powers ascriptions, so that powers should be part of our ontological inventory and/or of our fundamental ontological inventory14. 4. Getting Rid of Non-Existent Manifestations. I shall now analyse five attempts to get rid of non-existent objects as manifestations (or as part of the manifestations) of generative powers. Such attempts are based on four strategies, that I have mentioned in section 2, even if there is no strict correspondence between attempts and strategies: (i) generative powers are not fundamental entities; (ii) they are fundamental entities which do not depend on nonexistent objects for their individuation; (iii) those non-existent objects on which they depend for their individuation wholly depend in turn on existent objects or properties; (iv) those non-existent objects actually exist. Stephen Mumford (2004: 194-195) claims that powers are directed towards the manifestation of certain universal properties and they do not require particulars within their essences. Thus, the fundamentality of generative powers would not require non-existent objects as fundamental entities. He adds in the same place that first, a power is not typically a power to manifest a universal in some very precise way, at a precise time and place. A power might be a power to dissolve, when and wherever, and rarely a power to dissolve at spatiotemporal location p1t1. Second, because a universal is fully present in its instances, we can note that the thing to which the power is directed will indeed be present whenever we have an actual and specific instantiation. Our universal F is present in the specific manifestation F(p1,t1) and is the part of F(p1,t1 for which the power was a power . Mumford suggests that, if powers were powers to produce certain particulars, then they would be essentially individuated not only by that particular, but also by certain spatiotemporal locations. He seemingly thinks of something similar to Kimian events as the particular manifestations of such powers. In fact, Kimian events are essentially individuated by the objects and the n-adic properties that are involved in those events and by the time at which they occur. Anyway, one could reply that generative powers are directed either to non-existent objects, or to facts involving non-existent objects (e.g., the fact that a certain non-existent object starts to exist). Non-existent objects and facts are such that they are not essentially individuated by certain spatio-temporal locations: it is metaphysically contingent 14 Armstrong (2004: 137-138) gives an analysis of powers ascriptions in terms of conditionals and counterfactuals. Anyway, that analysis seems to be affected by the some problems that I have examined here. Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. that a certain object starts to exist at a certain spatio-temporal location and whatever is metaphysically contingent for something is not part of its essence. The second remark suggests that universals are more fundamental than particular manifestations, since those manifestations exist only because universals are instantiated. Besides reaffirming that it is not necessary that manifestations are essentially individuated by spatio-temporal locations, it is worth noticing that universals are instantiated only because they are instantiated by objects, so that objects turn out to be more fundamental than universals’ instantiations. Strategies (i)-(iii) are compatible with the following reduction scheme of the power to produce Cell-2, which is a paradigmatic case of generative power: (gen.powers) as a matter of metaphysical necessity, for every object, that object has the power to produce Cell-2 iff Ψ, where Ψ can be substituted by other powers that are not essentially individuated by non-existent objects or by entities that are not powers. Anyway, ) have already criticized Armstrong’s reduction of powers to other entities. It is now time to examine the former alternative. At first, one could consider powers towards the instantiation of properties, i.e., powers towards the fact that a certain property P is instantiated – or that it starts to exist, or that it acquires some feature (such alternatives are introduced in order to deal with properties that are not universal, as we will see). Here are some possibilities: Ψ Ψ = that object has the power to produce an instantiation of the property of being identical with Ψ Ψ = that object has the power to produce an instantiation of a conjunctive property PC (that Ψ Ψ = that object has the power to produce the existence of a certain mode (i.e., of a certain particular Cell-2; uniquely individuates Cell-2)15; property, that essentially depends on its bearer whatever else; 16), such as Cell- ’s existence or Cell- ’s being a cell or Ψ Ψ = that object has the power to produce the existence of a certain aggregation of tropes i.e., of a Ψ Ψ = there is a certain internal relation17 between the Platonic universal of having that power and certain aggregation of particular properties that do not essentially depend on their bearers ); some other Platonic Universal UC – that presumably uniquely individuates Cell-218. 15 Armstrong (1995: 619) accepts that non-existents can be reduced to combinations of properties. See Lowe (2006). 17 An internal relation is a relation that wholly depends on the existence and/or on the essence and/or on the intrinsic properties of its relata. 18 This solution is partly inspired by Tugby (2013). 16 Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. Following Ψ - Ψ , one could either claim that i generative powers are not fundamental entities, since they are reduced to other powers, or that (ii) they are fundamental entities which do not depend on non-existent objects for their individuation, but only on existent properties, or that (iii) those nonexistent objects on which they depend for their individuation wholly depend in turn on existent objects or properties, so that Cell-2 depends, for example, on tropes – at least according to Ψ . Ψ is hardly acceptable. In fact, it seems that the property of being identical with Cell-2 is essentially individuated by Cell-2, and not the opposite, at least if you do not wish to claim that there are haecceities that are more fundamental than objects – provided that they ground their individuation. Ψ3) should be dismissed for similar reasons, since it explicitly claims that the mode which starts to exist is also essentially individuated by Cell-2. Anyway, if one does not think of haecceities in terms of modes, Ψ1) turns out to be more acceptable than Ψ3). Ψ ) has two major problems. Firstly, it does not respect the intuition that there are metaphysically possible worlds in which Cell-2 has different properties, since it claims that Cell-2 is uniquely individuated by a certain, really complex conjunctive property PC. In this respect, you should accept the questionable assumption that every property within PC is essential to Cell-2, so that conjunctive properties that are slightly different from PC in other possible worlds individuate objects that are different from Cell-2. In other terms, Cell-2 does not exist in those worlds in which PC is not instantiated, even if a conjunctive property which is slightly different from PC is instantiated there. Secondly, there is a problem with PC that is analogous to the problem presented in section 3: either each conjunction of properties gives rise to a conjunctive property such as PC (and perhaps to a power), or only certain conjunctions of properties give rise to conjunctive properties such as PC. Both alternatives are problematic, as we have already noticed19. Ψ share these problems, as long as this solution only specifies the Platonic nature of the properties involved. Anyway, Ψ has the advantage of admitting UC’s Platonic existence even before its instantiation by Cell-2, so that UC can contribute to the generative power’s individuation even before that power’s activation. Finally, Ψ ) denies the intuition that there are metaphysically possible worlds in which Cell-2 is constituted by different tropes. Moreover, it also denies that those tropes that participate in grounding Cell- ’s individuation in the actual world can live different lives in other metaphysically possible worlds. Alexander Bird (2007: 112) claims that unrealized possibilities (i.e., mere possibilia) exist, even if they are only contingently abstract. Existence is here substituted by concreteness and he seemingly accepts that, while every object exists, not every object is concrete. The disagreement with Meinongians could be merely terminological. Anyway, concreteness implies a certain characterization of existence in terms of having a spatio-temporal location. Meinongians are not necessarily committed to the 19 In the latter case, what turns out to ground the existence and/or the fundamentality of a certain conjunctive property such as PC is a non-existent object. Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. acceptance of this notion of existence: they could accept other, non-equivalent notions. Moreover, Meinongians typically claim that there is no universal feature of existing – or of having being – that could be legitimately attributed to every entity. Their doctrine is not expressed by the thesis that there are entities that have being, even if they do not exist. According to Meinongians, there are entities that do not exist – full stop20. The two remaining solutions are more radical. According to the first – the nothing new under the Sun solution – there is no new object such as Cell-2, regardless of its existence or non-existence. Seemingly new objects are only rearrangements of existing, more basic objects – such as sub-atomic particles. I am not inclined to eliminate cells from my ontological inventory. Anyway, in order for this solution to be a coherent alternative to Meinongianism, it should not only deny that cells start to exist as new objects, but also that new objects at the fundamental micro-physical level of the universe start to exist. In fact, if there were generative powers at that level, they would be essentially individuated by nonexistent micro-physical objects. Finally, non-Meinongians could assert that there are no objects in the universe (or that objects are not fundamental entities), that the only existent entities (or the only fundamental entities) are powers themselves. However, what would powers be within this perspective? How would they provide a satisfactory reduction of generative powers or, more precisely, of seemingly true ascriptions of generative powers? If powers were something like properties (universal or particular), then such a solution would have the same troubles that characterize Ψ - Ψ . 5. Some Miscellaneous Concerns about Meinongianism. In this final section I shall briefly deal with four major concerns about Meinongianism. Following Quine (1948), it can be held that non-existent objects do not have definite identity conditions, that they are somehow indeterminate, so that they should not be accepted within our ontological inventory. Indeterminate objects are not objects at all. Thus, Cell-2 is not a non-existent object, because it is indeterminate, i.e., because it is not an object at all. The generative power to produce Cell-2 cannot be essentially individuated (among other) by Cell-2, since Cell-2 does not have definite identity conditions and it cannot help that power with its individuation. Yet, it is perhaps the case that the indeterminacy of Cell-2 is only epistemic, that we cannot define what is for something to be Cell-2, rather than Cell-3, even if there is a fact of the matter about their distinction even before their coming to existence. In addition, given Cell- ’s features, given the identity conditions for cells and given the ways in which cells generate other cells, it is legitimate to claim that Cell-1 can only generate a certain number of cells with certain features. Perhaps we do not know their exact number and we do not know all their features. Yet, this does not affect their real possibility , i.e., their being a definite number of distinct, non-existent-yet-possibly-existent objects. Moreover, both Cell-2, 20 Of course, entities do not only refer to existents. Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. Cell-3 and all the cells that can be generated by Cell-1 can still be numerically distinct from one each other. Thus, in my perspective, Cell-1 does not simply have the generative power to produce a cell, but it has the generative powers to produce Cell-2, Cell-3, and so on. Does this imply an overabundance of powers and/or of objects at the fundamental level of the universe? Put in these terms, the question turns out to be rhetorical. There cannot be overabundance of entities at the fundamental level of the universe: that level comprehends all and only the entities that it has to comprehend, regardless of our economical evaluation. Thus, if generative powers and non-existent objects such as Cell-2 turn out to be irreducible to other entities, they have the right to be part of that level, regardless of their being too many . Furthermore, attempts to analyse seemingly true ascriptions of generative powers turn out to commit non-Meinongians to an indefinite number of other entities: conjunctive properties, aggregates of tropes, haecceities, and so on. Thus, why should we accept an indefinite number of such entities and not accept an indefinite number of non-existent objects? Two further concerns remain. Firstly, it seems that what exists cannot be grounded, for its existence and/or for its features, on what does not exist. What exists is somehow more fundamental than what does not exist. Thus, Cell-2 cannot ground the nature of an instantiated generative power and it cannot ground one of the features of Cell-2 (its having that power). However, Meinongians could still preserve the primacy of what exists by claiming that all and only existents have irreducible causal powers, so that all and only existents can explain, by their causings, what happens in the universe. Finally, you cannot get rid of non-existent objects by simply claiming that such objects would both exist (since there are, i.e., there exist such objects) and do not exist. The argument that concludes from Meinongianism to this paradox of non-existence is question-begging, since it assumes that there is an object and there exists an object have the same meaning – and this is precisely what Meinongians deny! I shall conclude this paper with one final suggestion. Tugby (2013) claims that, if we accept that there are irreducible powers towards something, we should also accept a Platonic conception of properties. In fact, we should accept that there are properties that are not instantiated – or properties that are still not instantiated. Yet, within the Meinongian perspective that I have defended here, Armstrong’s Aristotelianism might be vindicated: seemingly non-instantiated properties could turn out to be properties that are actually instantiated (perhaps in certain peculiar ways) by objects that do not exist. Michele Paolini Paoletti Università degli Studi di Macerata michele.paolinip@gmail.com Refences. Armstrong, David Malet (1978). Universals and Scientific Realism. Volume II: A Theory of Universals. Cambridge: Cambridge University Press Michele Paolini Paoletti (Università degli Studi di Macerata) – michele.paolinip@gmail.com In F. F. Calemi (ed.) (2016), Metaphysics and Scientific Realism: Essays in Honour of David Malet Armstrong. Berlin: De Gruyter: 193-206. Please quote only from published version. ______ (1982). What is a Law of Nature? Cambridge: Cambridge University Press ______ . Reacting to Meinong . Grazer Philosophische Studien, 50: 615-627 ______ (1997). A World of States of Affairs. Cambridge: Cambridge University Press ______ (2004). Truth and Truthmakers. Cambridge: Cambridge University Press ______ . Four Disputes about Properties . Synthese, 144 (3): 309-320 Armstrong, David Malet, Martin, Charles Burton, Place, Ullin T. (2002). Dispositions. A Debate. New York: Taylor & Francis Bird, Alexander (2007). Nature’s Metaphysics. Laws and Properties. Oxford: Oxford University Press Cartwright, Nancy (1999). The Dapple World. A Study of the Boundaries of Science. Cambridge: Cambridge University Press Choi, Sungo, Fara, Michael . Dispositions . In: Stanford Encyclopedia of Philosophy Online (http://plato.stanford.edu/entries/dispositions) Ellis, Brian (2001). Scientific Essentialism. Cambridge: Cambridge University Press Lowe, E. Jonathan (2006). The Four-Category Ontology. A Metaphysical Foundation for Natural Science. Oxford: Oxford University Press Martin, Charles Burton (2008). The Mind in Nature. Oxford: Oxford University Press Maurin, . Anna-Sofia Tropes . (http://plato.stanford.edu/entries/tropes) )n: Stanford Encyclopedia of Philosophy Online Molnar, George (2003). Powers. A Study in Metaphysics. Oxford: Oxford University Press Mumford, Stephen (1998). Dispositions. Oxford: Oxford University Press ______ (2004). Laws in Nature. New York: Routledge Mumford, Stephen, Anjum, Rani Lill (2011). Getting Causes from Powers. Oxford: Oxford University Press Quine, Willard van Orman Tugby, Matthew . On What There )s . Review of Metaphysics, 2 (5): 21-36 . Platonic Dispositionalism . Mind, 122 (486): 451-480